A blue luminescent MOF as a rapid turn-off/turn-on detector for H2O, O2 and CH2Cl2, MeCN: 3 ∞ [Ce(Im)3ImH]·ImH

Article abstract of DOI:10.1039/c4dt03578j

The blue emitting luminescent MOF 3∞[Ce(Im)₃ImH]·ImH forms a 3D-framework with Kagom net topology. The framework exhibits an intense blue luminescence which can be retained upon activation of the MOF with the formula 3∞[Ce(Im)₃ImH]. The luminescence is metal-based due to parity-allowed 5d-4f-transitions. Time-dependent investigations of the interaction with liquid and gas analytes show that the MOF-by utilising 5d-4f-transitions of Ce³⁺-can be used as a high-speed "turn-off" detector for water and oxygen in dry air. Other protic or polar solvents, like methanol, acetone or pyridine, which also show a "turn-off"-effect can be distinguished from water-detection either on a time scale (ranging up to 250000:1) or a shift of the chromaticity, the latter being pronounced for MeOH. The fast time-dependent decrease of the luminescence intensity for water arises from an extremely fast hydrolysis and is irreversible. Polar aprotic molecules like dichloromethane and acetonitrile can also result in a "turn-on"-effect of the luminescence intensity due to their behaviour as additional sensitizers for Ce³⁺-emission. We conclude that the cerium-MOF can be utilised in gas and liquid sensing applications as a detector material for water and oxygen in dry air. The luminescence is intense with good quantum yield between 55% (as-synthesised) and 36% (activated). This implies that only milligram amounts of the material are needed to detect the analyte species and is especially useful, as the MOF can be directly used as-synthesised for water detection in applications for which an irreversible signal change is desired, e.g. preventing a signal change upon unwanted re-drying.

Full text of DOI:10.1039/c4dt03578j

Dalton
Transactions
View Article Online
View Journal
PAPER
A blue luminescent MOF as a rapid turn-off/
turn-on detector for H O, O and CH Cl , MeCN:
2
2
2
2
Cite this: DOI: 10.1039/c4dt03578j
3
[
Ce(Im) ImH]·ImH†
∞
3
a
a
a
b
b
L. V. Meyer, F. Schönfeld, A. Zurawski, M. Mai, C. Feldmann and
a
K. Müller-Buschbaum*
3
∞
The blue emitting luminescent MOF
[Ce(Im)
3
ImH]·ImH forms a 3D-framework with Kagomé net topo-
logy. The framework exhibits an intense blue luminescence which can be retained upon activation of
3
∞
the MOF with the formula
3
[Ce(Im) ImH]. The luminescence is metal-based due to parity-allowed
5
d–4f-transitions. Time-dependent investigations of the interaction with liquid and gas analytes show
3
+
that the MOF – by utilising 5d–4f-transitions of Ce – can be used as a high-speed “turn-off” detector
for water and oxygen in dry air. Other protic or polar solvents, like methanol, acetone or pyridine, which
also show a “turn-off”-effect can be distinguished from water-detection either on a time scale (ranging
up to 250 000 : 1) or a shift of the chromaticity, the latter being pronounced for MeOH. The fast time-
dependent decrease of the luminescence intensity for water arises from an extremely fast hydrolysis and
is irreversible. Polar aprotic molecules like dichloromethane and acetonitrile can also result in a “turn-on”-
3
+
effect of the luminescence intensity due to their behaviour as additional sensitizers for Ce -emission.
We conclude that the cerium-MOF can be utilised in gas and liquid sensing applications as a detector
material for water and oxygen in dry air. The luminescence is intense with good quantum yield between
55% (as-synthesised) and 36% (activated). This implies that only milligram amounts of the material are
needed to detect the analyte species and is especially useful, as the MOF can be directly used as-
synthesised for water detection in applications for which an irreversible signal change is desired, e.g. pre-
venting a signal change upon unwanted re-drying.
Received 21st November 2014,
Accepted 14th January 2015
DOI: 10.1039/c4dt03578j
www.rsc.org/dalton
cules and synthetic approaches. A major focus in recent years
has been the development of MOFs for sensing applications by
Introduction
Metal organic frameworks (MOFs) have made a significant using changes in photoluminescence properties caused by
1
0
impact on the field of materials science since their initial host–guest interactions as interpretable signals. For such
1
–3
4
discovery.
Renowned for their microporosity, MOFs also applications, two different sensing principles have been
display a variety of important properties as their secondary reported. The first of these relies upon a distinct and large
5
6
characteristics, such as magnetism, catalytic activity and change in the luminescence intensity upon interaction with
7
–9
photoluminescence.
This is leading to a growing field of the analyte species, leading to a turn-off or turn-on effect for
1
1
study of materials with various potential applications that can the luminescence as an observable/measureable parameter.
be modified by variations of linkers, metals, templating mole- For example an efficient quenching of the luminescence can
be used for sensing if the effect is selective to a certain mole-
cule, ion, etc. Such effects have been recently shown for the
a
Institute of Inorganic Chemistry, Julius-Maximilians-University Würzburg,
sensing of
2
O -gas using the porosity of MOF UiO-67
Am Hubland, 97074 Würzburg, Germany.
E-mail: k.mueller-buschbaum@uni-wuerzburg.de; Fax: +49-931-3184785;
Tel: +49-931-3188724
2−
(
Zr O (OH) (bpdc) , bpdc = para-biphenyldicarboxylate) for
6
4
4
6
3
+
loading with phosphorescent Ir -complexes. Interaction of
these complexes with O leads to quenching, whereas no sensi-
tivity is observed for N2. Different intensity changes can also
be utilised to track different analyte species, as shown for the
turn-off sensing of different nitro-aromatics by fluorescent
b
Institute of Inorganic Chemistry, Karlsruhe Institute of Technology (KIT),
2
1
2
Engesserstr. 15, 76131 Karlsruhe, Germany
†
Electronic supplementary information (ESI) available: DTA/TG investigations,
experimental details, crystallographic and powder diffraction data, photo-
luminescence spectra showing turn-off and turn-on effects. CCDC 1027917. For
ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/
c4dt03578j
2
−
Zn-based MOFs like [Zn (bpdc) bpee] (bpdc = 4,40-biphenyl-
2
2
dicarboxylate; bpee = 1,2-bipyridylethene), meaning they are
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Paper
Dalton Transactions
1
3
capable of detecting explosives. Recently, the europium MOF
Results and discussion
Crystal structure, topology, thermal properties, porosity and
[
Eu (MFDA) (NO )(DMF) ] (H MFDA = 9,9-dimethylfluorene-
3 4 3 3 2
2
,7-dicarboxylic acid) was shown to have variable quenching
adsorption
for ten different nitro aromatic systems, with the highest
3
∞
quenching observed for 3,4-dinitrotoluene (3,4-DNT) and 1,2- The compound [Ce(Im) ImH]·ImH (1) crystallizes in the
3
1
4
dinitrobenzene (1,2-DNB). The contrasting process, a turn-on monoclinic crystal system with the space group P2
1
/n, and is
3
23
of luminescence, is also of interest and can arise from an isotypic to
∞ 3
[Pr(Im) ImH]·ImH. The four crystallographically
analyte that acts as a sensitizer for a luminescence process independent Ce-atoms each are surrounded by seven N-atoms.
that would not be otherwise observable. An example of this The coordination sphere of three of the Ce-atoms can be
3
−
was shown for [Eu(BTPCA)(H
1
2
O)], BTPCA = 1,1′,1″-(benzene- described as a single-capped trigonal prism. The coordination
2
+
,3,5-triyl)tripiperidine-4-carboxylate, in which binding to Zn
sphere of the fourth Ce-site can be described as a pentagonal
3
+
leads to
intensity.
a
remarkable increase of the Eu
emission bipyramid. In all polyhedra, six edges are occupied by
1
5
1
3
−
μ-1κN : 2κN -bridging imidazolate anions (Im ), while one
The second possibility for luminescence sensing by a MOF edge is occupied by a neutral end-on coordinating imidazole
is by a change of the luminescence colour upon interaction of molecule (ImH), resulting in trivalent Ce-atoms (Fig. 1,
1
0,11
the MOF with the analyte.
This occurs as a result of the bottom). Six imidazolate anions connect four cations resulting
analyte influencing relevant energy levels involved in the lumi- in a P O -analogous pseudo-tetrahedral Ce (Im) -unit (Fig. 1,
4
6
4
6
nescence process without quenching the luminescence. Such a top). Connecting them further via sides and corners yields a
process was successfully shown for different simple aromatic three dimensional network (Fig. 2, left). The resulting topology
ligands interacting with the MOF. A representative example is can be described as four two-dimensional Kagome-like 3,4-net-
the use of zinc containing IRMOFs to detect benzene and works (Fig. 2, right). Based on this 3,4-net parallel to the (110)-
some of its derivatives like Tol, MePh, PhCN or PhOMe. Inter- plane, three additional distorted Kagomé nets share common
action of the IRMOF leads to a bathochromic shift of the MOF vertices with the initial base net. These three nets are
fluorescence over the complete visible spectrum from benzoni- arranged in 126°, 115° and 112°, respectively, with respect to
16
trile to iodobenzene. In addition to the detection of gas and their common base net. The framework exhibits an extensive
2
+
liquid molecules, the sensing of metal ions besides Zn has channel system along all three cell axes.
also been shown for other transition metal ions and lantha- To indicate phase purity and high yields of 1, the results of
1
7
nides. It is evident that for sensing applications the sensor PXRD investigations were compared to single crystal data and
must be fast, sensitive and at best selective for only a few are shown in Fig. 3. Due to the higher Coulomb participation,
different analytes. This limits the current number of known the amide bonds between imidazolate anions and Ce-atoms
suitable luminescence sensors to a small number of MOFs. are about 8–14 pm shorter than the amine bonds of the end-
Thus the identification of new MOFs as potential sensors is an on imidazole ligands.
attractive idea.
In recent years, imidazolate MOFs have been highlighted
for use as ZIFs due to their capability of adopting zeolite struc-
tures if the metal centre is four-coordinate. In addition, metal
ions with large coordination numbers (CN) can lead to inter-
esting MOFs, e.g. when combined with lanthanide ions as they
offer multifunctionality constituted by functions like porosity
and luminescent properties. However, the high CN limits the
inner surface of the materials. We have previously shown that
reacting lanthanide metals with melts of heterocyclic
amines like azoles leads to a variety of photoluminescent
1
8–22
frameworks,
some of the most potent and efficient emit-
ters being imidazolates. Herein, we wish to elaborate on the
formation of a previously unknown cerium-based imidazolate
3
framework, ∞[Ce(Im) ImH], that shares structural character-
3
3
23
istics with the framework
[Ce(Im) ImH]·ImH (1) is obtained by reacting elemental
∞
[Pr(Im)
3
(ImH)]. The compound
3
∞
3
cerium with a self-consuming melt of the linker molecule 1H-
imidazole and exhibits an intense blue photoluminescence
originating from f–d-transitions. Further studies show micro-
porosity upon activation as well as capable function as an
analyte specific detector. This can be related to rapid changes
in the photoluminescence intensity upon exposure to various
analyte gases and liquids.
3
∞
Fig. 1 Coordination sphere of the Ce atoms in
bottom) with a probability level of the atoms of 50% and a Ce
unit in comparison with P (top).
[Ce(Im)
3
ImH]·ImH
(
4
(Im) -
6
4
O
6
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
3
∞ 3
Fig. 2 Crystal structure of [Ce(Im) ImH]·ImH (1) view along [010] (left),
with ImH occupied channels (red circles); one of the two-dimensional
kagome-like 3,4-networks (right).
Fig. 4 Adsorption isotherms according to BET on activated 1, indicating
2
microporosity together with a preference for CO adsorption.
excess imidazole). These findings corroborate the results of
the elemental analysis, which also show a slight reduction
(
3
calcd C, 37.73; N, 29.33; H, 3.59. Found C, 36.62; N, 28.24; H,
.42%). It is most likely that non-coordinating imidazole mole-
cules are released initially, before coordinating end-on imid-
azole molecules are released. The resulting compound with
the constitution [Ce(Im) ] is thermally stable up to 550 °C.
3
In order to prove the MOF character and potential micro-
porosity, 1 was activated at 225 °C at 10− mbar for 4.5 h.
Adsorption–desorption isotherms were recorded by BET inves-
tigations for N , Ar, CO , and Kr (−196 °C for N , and Ar, as
3
2
2
2
well as −75 °C for CO
2
, and −153 °C for Kr) in the relative
pressure range 10− to 1 atm. The results were compared to
the non-activated framework. While the non-activated frame-
work is dense simply due to the occupied pores, significant
amounts of all gases investigated can be incorporated into the
6
Fig. 3 Powder diffractogram of bulk 1 (black) compared to powder
diffractogram of simulated single crystal data of 1 (red).
In accordance with the lanthanide contraction, the Ce–N- activated MOF (see Fig. 4). In addition, the MOF shows the
distances are longer than the Pr–N-distances in the isotypic highest preference for CO adsorption giving an uptake of
2
3
23
3
−1
3
−1
compound
∞ 3
[Pr(Im) (ImH)]·ImH. The angles N–Ln–N are in 160 cm g followed by Ar adsorption (135 cm g ), whereas
the same range in both compounds. The Ce–N distances are in only 85 cm g of N and Kr could be adsorbed. This corres-
the range of 2521(3)–2657(3) pm and accord well with the ponds to a moderate surface area of SBET = 480 m g for CO2
3
−1
2
2
−1
1
∞
24
former published
[Ce(Btz)
3 2
(BtzH)] as well as [Ce(triRNTB) ]- adsorption at −75 °C. We also tried high-pressure CO adsorp-
2
(
CF SO ) (triRNTB
=
tris(N-alkylbenzimidazol-2-ylmethyl)- tion up to 45 bar at 50 °C, but decomposition stops the adsorp-
3
3 3
2
5
amine). The pores of 1 provide a maximum diameter of tion process for this CO pressure. The lower uptake Kr and N
2
2
7
60 pm with apertures of 580 pm when taking hydrogen can be explained by the atom size and the kinetic diameter,
2
6
van-der-Waals radii into consideration.
Simultaneous DTA/TG investigations on
respectively. The observed preference for CO may be a result
2
3
∞
[Ce(Im)
3
(ImH)]· of the oxophilicity of the lanthanide cerium in 1.
ImH (1) exhibit two endothermic signals at 175 and 245 °C
with a subsequent mass loss of 26% (see Fig. S1, ESI†). Given
that one equivalent of imidazole corresponds to 14% of the
overall mass, the mass loss of 26% can be attributed to the
Photoluminescence properties and sensing studies for liquids
and gases
loss of the two imidazole molecules: an initial loss of the inter- Both the as-synthesised and the activated compound 1 show
3
∞
calated imidazole equivalent having the formula
[Ce(Im)
3
-
an intense blue luminescence under UV-excitation (Fig. 5) that
(ImH)] and subsequently, to the loss of the coordinating imi- does not change upon activation. The identical shapes of the
dazole molecule and [Ce(Im) ]. The initial release step gives a photoluminescence spectra prove that the chemical surround-
3
value for the imidazole release of about 3% less than the ing of cerium is maintained upon activation and that no
expected value. This can be explained by an observed partial structural collapse occurs. This is very useful, especially, as
loss of imidazole during the workup processes (sublimation of PXRD is not possible.
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Paper
Dalton Transactions
3
∞
Table 1 Crystallographic data of
[Ce(Im)
3
ImH]·ImH (1). Standard devi-
ations are given in brackets
Empirical formula
15 17
C H N10Ce
M
r
477.51
Monoclinic
P2 /n
Crystal system
Space group
a/pm
b/pm
c/pm
1
2325.3(5)
1402.4(3)
2345.5(5)
107.7(1)
7288(3)
16
1.741
25.2
3760
β/°
V/10 ·pm
6
3
Z
D
calcd/g cm−
3
−1
µ(Mo-Kα)/cm
F (000)/e
hkl range
Refl. observed/unique/Rint
Parm. refined
R(F)/wR(F ) (all refls.)
−33 ≤ h ≤ 33; −19 ≤ k ≤ 20; −31 ≤ l ≤ 33
77 635/22 006/0.0623
1145
Fig. 5 Normalised excitation and emission spectra of as-synthesised
and activated 1 (left); visible blue emission of as-synthesised (bottom)
and activated 1 (top) under UV excitation (λExc = 366 nm) (right).
2
a
0.0529, 0.1167
1.40/−1.56
−3
Δρ (max/min)/e Å
a
2
2 2
2 2 1/2
R
1
= ∑[|F
o
| − |F
c
|]/∑|F
o
|, wR
2
= [∑w(|F
o
| − |F
c
| ) /∑w(|F
o
| ) ]
.
3
+
The luminescence of Ce is based on parity allowed 4f–5d
transitions for excitation and vice versa 5d–4f transitions for
emission. Typically, 5d–4f transitions of lanthanide ions
mainly occur in the UV region due to 5d states lying at high
Table 2 Photoluminescence data of as-synthesised and activated 1
2
+
3+
as-Synthesised
MOF
Activated
MOF
energy for most lanthanide ions. Besides Eu , Ce is an
exception among the trivalent lanthanide ions as it can
−1
provide 5d–4f transitions <33 000 cm by crystal and ligand- Max. emission (λExc = 366 nm)
Max. excitation (λ = 422/425 nm)
422 nm
367 nm
425 nm
380 nm
2
7
Em
field splitting. Depending on the chemical surrounding of
3
2
CIE coordinates
3
+
the Ce ions, these transitions can be shifted to lower ener-
x
y
0.150
0.059
55%
0.149
0.066
36%
3
+
gies. This is a prominent feature of the Y Al O :Ce laser
(
3
5 12
2
8
Absolute QY (λExc = 366 nm)
YAG:Ce) that emits in the yellow region. Due to the partici-
pation of d-states, and thus due to the strong influence of both
ligands and crystal fields, excitation and emission energies are
dependent on the chemical surrounding. This results in a blue activated MOF exhibit an additional minor emission
3
luminescence for as-synthesised
∞
[Ce(Im)
3
ImH]·ImH and acti- maximum at λ = 460 nm. We attribute this secondary band to
3
3+
vated [Ce(Im) ImH]. The parity allowed character of these multiple effects: Ce shows splitting of the ground states
transitions is responsible for a high transition probability and ( F5/2 and F7/2 expected value = 2300 cm according to ref.
thus intense luminescence of 1. The excitation of Ce occurs 27). As the observed value 1690 cm is at the border of the
∞
3
2
2
−1
3+
−1
1
0
1
via promotion of the 4f ground state into the 4f 5d -level. reported energy gap even if a large literature deviation of
Due to the nephelauxetic effect, 5d-levels are lowered in energy ±600 cm− is considered, we also looked for additional non-
1
3
+
and split (crystal field splitting) in the coordinated sphere metal based effects. In addition to the Ce -ground state split-
compared to free Ce . The broad emission occurs upon relax- ting, an emission of the deprotonated ligand imidazolate Im
ation from 4f 5d -levels into 4f -levels ( F5/2 and F7/2) that are immobilised in the rigid framework may be another origin of
split by spin orbit coupling.
the emission shoulder. Additional energy pathways, which are
The excitation spectra of 1 exhibit a maximum at λ = not observed for the free ligand, have been previously reported
3
+
−
0
1
1
2
2
3
0,31
3
67 nm and a weaker second maximum at λ = 276 nm. The for ligands immobilised in the MOF structures.
As 1H-
weaker additional maximum can be addressed to an excited imidazole does not show the related band, sodium imidazolate
1
9,29
*
S-state of the ligand.
The emission maximum is located was prepared as a spectroscopic reference for the imidazolate
at λ = 422 nm. The absolute quantum yield was determined to anion (see Fig. S2, ESI†). Sodium imidazolate (NaIm) also
be QY = 55% (λExc = 366 nm) and shows that a high amount of shows the corresponding emission band without the ground-
3
+
absorbed light is reemitted. For the activated MOF, the first state splitting of Ce . Accordingly, comparison between the
3
∞
and second excitation maxima are at λ = 380 nm and λ = emission maxima of
81 nm, the emission maximum is λ = 425 nm. In comparison firms that the minor maximum originates from the emission
with the as-synthesised MOF, the absolute quantum yield of anionic imidazolate in addition to the ground state splitting
[Ce(Im) ImH]·ImH (1) and NaIm con-
3
2
3
+
decreases to QY = 36% (λExc = 366 nm), indicating a sensitizer of Ce .
effect of the intercalated imidazole in the as-synthesised MOF.
The photoluminescence of 1 can be influenced by liquids
In Table 2 the significant luminescence data for as-synthesised like classic solvents and gases, thus providing potential for
and activated 1 are reported. Both the non-activated and the detection of the interacting species as analyte. Altogether,
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
eight different solvents were investigated by addition of the
solvent to the MOF: water, methanol, acetone, pyridine, aceto-
nitrile, dichloromethane, toluene and hexane under inert con-
ditions in argon. Time dependent excitation and emission
spectra were recorded subsequent to the addition of the
respective solvent. They were then compared to the state of the
pure MOF and the influence of the solvent on the lumine-
scence was monitored.
In addition, the behaviour of the MOF in the presence of
four different gases, namely Ar, N , CO and O , was investi-
2 2 2
gated. A sample cell filled with the compound that was pre-
viously activated in vacuo was flooded with the gas. Again, time
dependent excitation and emission spectra were compared to
the spectra of 1 to monitor the potential influence of the
analyte gas.
The time-dependent luminescence investigations for sol-
vents can be beneficially carried out with the purified as-syn-
thesised MOF, sparing the full activation process. During our
studies, it became evident that the time-dependence can be
positively influenced by
a molecule exchange process.
Exchange processes elongate the time scale for solvents that
cannot easily replace imidazole from the as-synthesised MOF.
Therefore, a time-dependent selectivity is available. Moreover,
sparing the full activation of the MOF simplifies potential
application as a detector/sensor.
The investigation of the solvents shows a variable influence
on the MOF photoluminescence depending on the character
of the solvent: protic, polar, aprotic and non-polar. This results
Fig. 6 Emission spectra of 1 as time dependent monitoring of the
influence of H O (top) and MeOH on the emission intensity.
2
in intensity changes either being “turn-on” or “turn-off” tion could easily happen as such products are not continu-
effects. Up to now, contact times in hours, rather than time ously monitored during storage. Accordingly, this kind of
scale of detection have been reported in studies on lumine- detection has to be irreversible.
scence sensing with MOFs. For this reason, we carried out
In comparison with water, the protic solvent MeOH also
time-dependent investigations on sensing with 1. In order to gives a reduction of the emission intensity, but on a longer
avoid a concentration effect, the time-dependent investigations time scale. It takes 10 days to reduce the luminescence inten-
were carried out by direct contact with the analyte.
sity by the same amount as water does instantaneously. Thus,
The luminescence is immediately and fully quenched upon water and MeOH as an alcohol can be clearly distinguished
addition of water in less than 3 s. The quenching, reaching by a ratio of complete “turn-off” luminescence effect of
λmax at 400 nm half-width <3 s, is faster than recording an >250 000 : 1. For methanol, 65% of the intensity is still present
emission spectrum, as proven by spectrometer parameters: after 48 h and 15% after 10 days (see Fig. 6 and S3, ESI†).
range 380–650 nm, scan time 0.1 s nm−1, 1 s integration time
The aprotic polar solvents acetone and pyridine (py) also
for the overall spectrum, giving an overall recording time of show this gradual loss of the excitation and emission intensi-
8 s. Longer interaction with water does not further change ties; acetone taking one hour and pyridine six hours for an
2
the photoluminescence behaviour (see Fig. 6). This “turn-off” almost complete “turn-off” effect (see Fig. S4, S5, ESI†). This
effect arises from the hydrolysis of the MOF and thus allows gives a time scale ratio of 1000 : 1 (acetone) and 6000 : 1 (py)
an irreversible detection of water. In contrast to other water- when compared to water detection. For this reason, the time-
sensitive MOFs, the high speed of the process combined with dependence of an intensity-change can be used to increase the
the observable parameter of luminescence intensity loss selectivity of the “turn-off” detection process.
renders 1 a suitable detector for water.
Irreversible detection of water can, e.g. be especially useful tinguish the different “turn-off”-effects and thereby the ana-
Aside from the time-scale, another method used to dis-
3
2
for detection systems that do not allow a reversible change of lytes was the chromaticity of the emission according to CIE.
the detection signal, but instead rely on maintaining the detec- These were determined as colour points after addition of each
tion grade for longer times. Such examples include water-sen- solvent. The colour point was determined when a stable value
sitive pharmaceutical formulations, for which we are currently was reached and remains almost unchanged with the excep-
developing such sensors. For such pharmaceuticals reversible tion of the addition of methanol. Methanol addition shows a
processes would enable possible unwanted re-drying and clear shift (see Fig. 7) compared to 1 or 1 with MeCN or pyri-
therefore a wrong detection signal. A wrong outcome of detec- dine (equal to a bathochromic shift of >20 nm). MeOH can
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Fig. 7 Excerpt of the chromaticity diagram according to CIE showing
the colour points of 1 after the addition of solvents.
therefore also be distinguished e.g. from MeCN and py; the
time-dependence measurements do not allow for differen-
tiation between these three solvents. Such irreversible “turn-
off” effects can arise from dissolution of the MOF by coordi-
nation of the analyte, as e.g. pyridine complexes of the lantha-
3
3
nide are known. The formation of Ln–pyridine coordination
could occur due to acid–base adduct formation between the
3
+
Lewis-base pyridine and the Lewis-acid Ce . An analogous
process can of course also occur for protic solvents like MeOH
and water, for which the fast irreversible hydrolysis are also Fig. 8 Time dependent excitation (top) and emission (bottom) spectra
observed. To identify potential products and residues of the of 1 in dichloromethane (CH
2 2
Cl ) indicating “turn-on”-effect of 30%.
MOF framework suspended in each solvent, powder diffraction
experiments were carried out after evaporation of respective
solvents, but gave only X-ray amorphous patterns. show no change if it is dry and kept under inert conditions. As
Therefore, the time-dependent loss in intensities of excitation expected, this extends also to chromaticity (see Fig. 7).
and emission after addition of the respective solvent is a hint
of an irreversible reaction of the analytes with the MOF carried out for the typical gas components of air: N
compound 1.
and CO . Therefore, a sample cell filled with compound 1,
Time dependent gas-sensing investigations of 1 were also
, O , Ar,
2
2
2
For the two aprotic, polar solvents acetonitrile and dichloro- previously activated in vacuo, was flooded with the analyte gas.
methane, interestingly a reversible “turn-on”-effect is detected An initial photoluminescence spectrum was recorded under
for the excitation and emission intensities of 1 (see Fig. 8). vacuum conditions (10− mbar) prior to flooding of the
This “turn-on”-effect for dichloromethane can be identified sample cell up to a pressure of 99 992 Pa (750 mmHg) of the
5
1
3
0 minutes after solvent addition and yields an increase of gas. Luminescence spectra were recorded every 300 s for one
0% in overall intensity, reached after 1 h. For acetonitrile, a hour and compared to the spectra prior to gas application. The
moderate intensity increase of 15%, reached after 2 h, was results are again dependent on the analyte indicating potential
recorded after the addition of the solvent (see Fig. S6, ESI†). “turn-off” sensing for O and to a lesser extent for CO rep-
Both solvents can obviously act as additional sensitizers for resented by a decrease in the luminescence intensity. No
2
2
3
+
the photoluminescence of the MOF feeding the Ce emission change or shift was observed in signal shape or position of the
by an energy transfer. These additional antenna effects are intensity maximum, respectively. Subsequently, all spectra
reasonable, as e.g. nitriles have already been shown to act were integrated over the full width of the observed range to
as sensitizers for MOF-luminescence in the dinitriles obtain the absolute intensity decrease (380 nm to 650 nm, see
3
∞
34
[
LnCl (1,3-Ph(CN) ], Ln = Eu, Tb. However, the amount of Fig. 9). The integrated intensities were normalised to the refer-
3 2
intensity increase is quite surprising for CH Cl (see Fig. 8, S6, ence spectrum recorded for the evacuated sample cell.
2
2
ESI,† for MeCN).
Intensity fluctuations with a magnitude of up to 4% were
exposures at room temperature
The time-scale observed for the reversible “turn-on” pro- observed for both Ar and N
2
cesses can also be attributed to the soaking process in which but these intensity changes are not thought to be significant.
the analyte enters the MOF pores. This is a result of the inter- The deviations may be explained by a variety of minor error
action of the analyte molecules and the MOF influencing both sources such as deviations occurring between flooding the
time-dependence and absolute intensity change. Furthermore, sample cell with analyte gas from vacuum to 99 991.8 Pa
photoluminescence spectra were also recorded for 1 in the (750 mmHg), mobility of the finely ground powder particles or
aprotic, non-polar solvents toluene and hexane. The spectra fluctuations of the excitation beam intensity. It can be con-
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
With regard to the ratio of O
2
and CO
2
in air, 1 can be used
as a detector for oxygen in dry air, as the low amount of CO in
2
air hardly influences the result giving a deviation/error of <1%
of the detection signal.
Conclusions
3
∞
A novel blue luminescent MOF
[Ce(Im)
3
(ImH)]·ImH was
achieved via solvent-free redox reaction of the linker 1H-imida-
zole and elemental cerium. The as-synthesised MOF can be
3
∞
activated to give the microporous constitution
[Ce(Im)
3
-
(
ImH)]. The intense blue luminescence results from parity
allowed 5d–4f-transitions and is retained upon activation.
Absolute quantum yields were determined for the as-syn-
thesised and the activated form of the MOF reaching 55% for
the non-activated due to a sensitization effect of imidazole in
Fig. 9 Time-dependent integrated relative emission intensities of 1 for
different gas exposures.
cluded that the MOF luminescence is not influenced by the the pores. Removal of these molecules reduces the quantum
gases Ar and N within these deviations.
yield to 36%. The luminescent MOF can be utilised as a poten-
leads to a very sharp drop in tial sensor/detector for O in dry air and H O by a significant
2
Exposure of the MOF to O
2
2
2
intensity of 70% within seconds, and about 85% in 5 min. “turn-off”-effect of the luminescence. Another application lies
This behaviour follows an asymptotic function and may indi- in the detection of CO in the absence of the other two ana-
2
cate a direct interaction of the oxophilic luminescent centres lytes. In general, “turn-off”-effects can be observed for protic
with O2 and quenching of the luminescence processes by and polar solvents. Time-dependent investigations show a
3
+
oxygen-presence within the coordination sphere of the Ce
high-speed detection (<3 s) for water only, whereas MeOH,
centres. A weaker “turn-off”-behaviour can be observed for the acetone, pyridine show a much longer period for the time-
exposure of 1 to CO (99 992 Pa, 750 mmHg; see also Fig. 9), dependent “turn-off”. Selectivity can thereby be achieved by
2
but reaches an intensity reduction to only 45% of the original the high ratio of the time-dependent signal change, which
intensity within an hour, and almost no change within the ranges up to 250 000 : 1 for H O–MeOH.
2
first minute. Both gases therefore show a “turn-off” effect that
is clearly distinguishable by their time-/intensity-dependence, detector with good time-dependent selectivity, as the exchange
with O being much more pronounced and spontaneous.
process of intercalated molecules in the pore system enhances
The purified as-synthesised MOF can be directly used as a
2
To identify potential chemical processes that might be the time scale. The complex activation process can be omitted,
responsible for the decrease of the emission intensities, sub- whereas gas detection depends on the accessibility of the pore
sequent to flooding and evacuating of the sample cell, powder system. The “turn-off”-effects of the analytes MeOH, py and
diffraction investigations were carried out. In both powder acetone can be further distinguished by a shift in the chroma-
diffractograms, the main reflections of the MOF (1) are ticity, especially for MeOH. The chromaticity of the lumine-
present, giving evidence that the MOF structure is maintained scence shifts to higher wavelengths with increasing polarity of
(
see Fig. S7, ESI†). Monitoring the CO -pressure within the the analyte. In contrast to “turn-off”, the analyte molecules
2
sample cell during the exposure phase shows a loss of 1639.9 acetonitrile and dichloromethane provoke a “turn-on”-effect of
Pa (12.3 mmHg) within the first 60 minutes of the experiment the luminescence by acting as sensitizers. For the gases consti-
and a loss of 5732.9 Pa (43.0 mm Hg) after an exposure of tuting dry air, PXRD shows that the MOF structure is main-
2
4 hours. This is consistent with the equilibration times tained, whereas the speed of the water detection is attributed
observed in the BET experiment, which may be the explanation to an extremely rapid hydrolysis. This is useful for all processes
for the much slower detection/luminescence change of/by in which an irreversible signal change is desired (viz. that
CO
2
. In order to check for a reversible sensing, the sample cell should not be changed upon re-drying).
Due to the spontaneous character and impact strength of
was evacuated for an extended period of time (e.g. 72 h) and a
final spectrum was recorded. However, the intensity observed the “turn-off”-effect, the novel cerium-MOF can be used as a
remained at the reduced intensity values resulting from the rapid luminescence detector for water and oxygen. This is
exposure to CO and O . As the crystal structure does not especially the case for water detection as the as-synthesised
2 2
change no chemical reaction is taking place, different from MOF can be directly utilised. To the best of our knowledge,
exposure to water. This indicates the occurrence of a chemical this is the first example utilising 5d–4f-transitions for detec-
sorption that cannot be reversed by simple evacuation. As the tion with MOFs. Altogether, this report distinguishes and com-
lanthanides are highly oxophilic this observation also hints pares irreversible detection from reversible sensing for the
3
+
that when the gases enter the proximity of the Ce -ions, they different “turn-off” and “turn-on” effects reported including
act as quenchers. This quenching is not possible by N
2
and Ar. chemical reactions like hydrolysis, chemisorption and revers-
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Paper
Dalton Transactions
ible interaction, as all of them can be useful depending on the Photoluminescence spectroscopy & quantum yield
sensing/detection requirements.
determination
Excitation and emission spectra were recorded with a Horiba
Jobin Yvon Fluorolog 3 photoluminescence spectrometer with
4
50 W Xe lamp, an integration sphere, Czerny–Turner double
Experimental
Synthesis of ∞[Ce(Im) ImH]·ImH (1)
grating (1200 grooves per mm) excitation and emission mono-
chromators and an FL-1073 PMT detector.
3
3
Excitation spectra were recorded from 250 to 410 nm and
corrected for the spectral distribution of the lamp intensity
using a photodiode reference detector. Emission spectra were
recorded from 380 to 650 nm and corrected for the spherical
response of the monochromators and the detector using
typical correction spectra provided by the manufacturer.
The quantum yield was also recorded using a photo-
luminescence spectrometer Horiba Jobin Yvon Spex Fluorolog
3
The compound
∞
[Ce(Im)
3
ImH]·ImH (1) was synthesized via a
solvent-free reaction in a self-consuming organic melt of 1H-
imidazole. All reagents were used as purchased. Freshly filed
Ce-metal (99.9%, Smart elements, 1 mmol, 140 mg) and 1H-
imidazole (99.5%, Sigma Aldrich, 5 mmol C N H , 340 mg)
3
2
4
were sealed in an evacuated DURAN® glass ampoule (1 × 10−
3
mbar). The reaction mixture was heated in a resistance tube
oven from room temperature to 190 °C within 3 hours. The
temperature was kept constant for 72 h and then cooled to
room temperature within 12 h. To achieve complete reaction,
the mixture was ground with further 1H-imidazole (1 mmol,
3
, equipped with a 450 W xenon lamp, double monochroma-
tors for excitation and emission, an integrating sphere
Ulbricht sphere) and a photomultiplier as the detector. For
excitation wavelengths >330 nm, the absolute quantum yield
(
6
8 mg), sealed and heated again, using the heating program
3
8
was determined according to Friend. First the diffuse reflec-
tion of the sample was determined at a certain excitation wave-
length. Second, the emission was measured under excitation
at the same wavelength. Integration of the reflected and
emitted photons by use of the Ulbricht sphere resulted in the
absolute quantum yield. Corrections were made accounting
for the spectral power of the excitation source, the reflection
behaviour of the Ulbricht sphere, and the sensitivity of the
detector. For excitation wavelengths <330 nm, the reflectivity of
the Ulbricht sphere starts to be perceptibly below 100%, result-
ing in a less precise determination of the absolute quantum
yield. Therefore, the relative quantum yield was determined by
comparison with a fluorescent lamp phosphor with a defined
described above. Excess 1H-imidazole was removed via subli-
mation. The remaining product was a colourless fine powder
of 1. Yield: 353 mg = 74% Anal. calcd for C15
H
17
N
10Ce (M
r
=
1
4
2
2
77.51 g mol− ): C, 37.73; N, 29.33; H, 3.59. Found C, 36.62; N,
8.24; H, 3.42%. MIR (KBr): (3358 m, 3124 m, 3021 m, 2918 m,
827 m, 1677 w, 1596 w, 1528 w, 1456 s, 1419 m, 1321 w, 1304
w, 1252 m, 1231 m, 1141 m, 1078 bs, 954 w, 933 s, 836 m,
−
1
7
1
54 m, 675 s, 661 m, 621 w, 595 m) cm . FIR (PE): (192 w,
−
1
57 w, 142 w, 130 w, 116 vs) cm
.
X-Ray diffraction
3
∞
A suitable crystal of
[Ce(Im)
3
ImH]·ImH was selected for the
2+
2+
quantum yield. Here, BaMgAl O :Eu (BAM:Eu ) (Philips)
1
0 17
single crystal X-ray analysis in a glovebox and sealed in a glass
capillary. The data collection was carried out on a STOE IPDS-I
39
was used as a reference.
diffractometer (MoK
structure was determined using direct methods. The non-H
atoms were refined anisotropically by least squares tech- For the potential luminescence sensing of solvents, [Ce(Im)3-
α
radiation λ = 0.7107 Å) at 140(1) K. The
3
5
Sensing/detection investigations
3
∞
3
6
niques. The positions of the hydrogen atoms were refined (ImH)]·ImH (1) and a mini agitator were placed under inert
isotropically for all framework H-atoms and derived from pre- conditions in a quartz cuvette, which was sealed with a
calculated positions using a riding model for the intercalated septum and additional wrapping with Parafilm®. Directly after
imidazole molecule. The compound crystallizes in the mono- the addition of the relative solvent excitation and emission
clinic crystal system in the centro symmetric space group spectra were recorded. The used solvents were dried under
3
7
P2 /n. Integrity of the symmetry and geometry were checked.
standard conditions (except water and acetone). The sample
Crystallographic data for the crystal X-ray determination are cell was coupled with a Horiba Fluorolog-3 photoluminescence
summarized in Table 1. spectrometer via a fibre-optic cable. This assembly allows
1
Powder samples were prepared in Lindemann glass capil- photoluminescence experiments with solid/liquid mixtures
laries (Ø = 0.3 mm) under an inert gas atmosphere. The under reflection conditions, while the Ce-imidazolate was
powder diffractograms were recorded on a Bruker D8 Discover stirred in order to provide contact between liquid phase and
with Da Vinci design and LynxEye detector in transmission the framework particles.
geometry with a focusing Göbel mirror with CuKα-radiation
The used eight solvents were purified and dried under stan-
(
λ = 1.54056 Å). For investigation of the residues of the dard conditions (except water which was used as distilled
liquid sensing photoluminescence spectroscopy, the respective water and acetone which was used as purchased): dried metha-
solvent was evaporated under vacuum and the powder nol: pre-dried using magnesium chips and iodine followed by
diffractograms were recorded on a Si-waver in reflection geo- distillation; pyridine: left over molecular sieves 4 Å for several
metry with CuKα-radiation (λ = 1.54056 Å).
days; acetonitrile: distillation over P O ; dichloromethane: left
2 5
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Dalton Transactions
Paper
2 2 5
over CaCl overnight and then distilled over P O ; toluene: dis-
Notes and references
tillation over CaH ; hexane: left over 24 hours over CaCl and
2
2
then distilled over P
2
O
5
.
1 (a) B. F. Hoskins and R. Robson, J. Am. Chem. Soc., 1989,
111, 5962; (b) B. F. Hoskins and R. Robson, J. Am. Chem.
Soc., 1990, 112, 1546.
2 (a) G. Ferey, Chem. Soc. Rev., 2008, 37, 191; (b) J. R. Long
and O. M. Yaghi, Chem. Soc. Rev., 2009, 38, 1213;
(c) L. J. Murray, M. Dincă and J. R. Long, Chem. Soc. Rev.,
2009, 38, 1294; (d) C. Janiak and J. K. Vieth, New J. Chem.,
2010, 34, 2366.
3 (a) O. M. Yaghi, H. Li, C. Davis, T. Richardson and
T. L. Groy, Acc. Chem. Res., 1998, 31, 474; (b) S. Kitagawa
and M. Kondo, Bull. Chem. Soc. Jpn., 1998, 71, 1739;
(c) H. Le, M. Eddaoudi, M. O’Keefe and O. M. Yaghi,
Nature, 1999, 402, 276; (d) G. Ferey, Chem. Mater., 2001, 13,
3084; (e) C. Janiak, Dalton Trans., 2003, 2781;
(f) S. Kitagawa and K. Uemura, Chem. Soc. Rev., 2005, 34,
109; (g) S. Kaskel, Nachr. Chem., 2005, 53, 394;
For potential gas detection, 1 was placed under inert con-
ditions in a specially designed L-shaped sorption cell of a
Quantachrome Autosorb AS-1c. The glass ceiling of the sample
cell was made from Suprasil glass, which is suitably transpar-
ent for photoluminescence experiments including the UV-VIS
region. The sample cell was then coupled with a Horiba Fluoro-
log-3 photoluminescence spectrometer via a fibre-optic cable,
allowing photoluminescence experiments under reflection
conditions, while the sample cell was flooded with gas
handled by a Quantachrome Autosorb AS-1c. Experiments were
carried out for Ar (Fa. Linde Gas, purity 5.0), N2 (Fa. Linde
Gas, purity 5.0) and CO
2
(Fa. Linde gas, purity 5.0). The
spectra were obtained at an excitation wavelength of 363 nm
and emission was measured between 380 nm and 650 nm with
incremental steps of 1 nm.
(
h) A. K. Cheetham and C. N. R. Rao, Science, 2007, 318, 58.
S. Kitagawa, R. Kitaura and S. Noro, Angew. Chem., Int. Ed.,
004, 43, 2334.
M. Kurmoo, Chem. Soc. Rev., 2009, 38, 1353.
M. Yoon, R. Srirambalaji and K. Kim, Chem. Rev., 2012,
4
Thermal and elemental analysis
2
For thermal analysis the bulk material of the synthesis of
5
6
3
∞
[
Ce(Im) ImH]·ImH was purified by evaporation of excess 1H-
3
imidazole at 110 °C under vacuum. The thermal properties
were investigated by simultaneous DTA/TG (Netzsch STA 409,
Proteus Software), using 34 mg of sample. The sample was
112, 1196.
7
(a) D. F. Sava, L. E. S. Rohwer, M. A. Rodriguez and
T. M. Nenoff, J. Am. Chem. Soc., 2012, 134, 3983;
kept under an inert gas atmosphere (50% Ar and 50% N
2
) and
(
b) A. Ablet, S.-M. Li, W. Cao, X.-J. Zheng, W.-T. Wong and
heated from 25 to 950 °C at a heating rate of 10 °C min−
1
.
L.-P. Jin, Chem. – Asian J., 2013, 8, 95; (c) P. Falcaro and
S. Furukawa, Angew. Chem., Int. Ed., 2012, 51, 8431;
Elemental analysis for C, H and N was performed with a Euro
Ea CHNSO Analyzer (HEKAtech GmbH).
(
d) M. D. Allendorf, C. A. Bauer, R. K. Bhakta and
R. J. T. Houk, Chem. Soc. Rev., 2009, 38, 1330.
Gas adsorption properties
8
9
P. R. Matthes, C. J. Höller, M. Mai, J. Heck, S. J. Sedlmaier,
S. Schmiechen, C. Feldmann, W. Schnick and K. Müller-
Buschbaum, J. Mater. Chem., 2012, 22, 10179.
J.-C. Rybak, L. V. Meyer, J. Wagenhöfer, G. Sextl and
K. Müller-Buschbaum, Inorg. Chem., 2012, 51, 13204.
Volumetric uptake and specific surface areas were determined
by adsorption–desorption isotherms of different gases
(
−196 °C for N , 99.999% and Ar, 99.999%, as well as −75° C
2
for CO , 99.9993% and −153° C for Kr, 99.996%) obtained on
2
1
0 (a) L. E. Kreno, K. Leong, O. K. Farha, M. Allendorf,
R. P. Van Duyne and J. T. Hupp, Chem. Rev., 2012, 112,
a Quantachrome Autosorb 1C apparatus according to the
BET theory. Prior to the adsorption measurements, the
−
3
1105; (b) J. Lei, R. Qian, P. Ling, L. Cui and H. Ju, Trends
samples were degassed at a vacuum of 10 mbar for 4.5 h at
25 °C. The results were compared to the non-activated frame-
Anal. Chem., 2014, 58, 71.
2
1
1 (a) D. Liu, K. Lu, C. Poon and W. Lin, Inorg. Chem., 2014,
work 1.
53, 1916; (b) Z. Hu, B. J. Deibert and J. Li, Chem. Soc. Rev.,
2
014, 43, 5815.
Vibrational spectroscopy
1
1
2 Z. Xie, L. Ma, K. E. deKrafft, A. Jin and W. Lin, J. Am. Chem.
Soc., 2009, 132, 922.
3 (a) A. Lan, K. Li, H. Wu, D. H. Olson, T. J. Emge, W. Ki,
M. Hong and J. Li, Angew. Chem., Int. Ed., 2009, 48, 2334;
FTIR spectra in the MIR area were recorded using a Thermo
Nicolet FTIR-380 spectrometer; samples were measured in
dried KBr pellets under inert conditions. Spectra in the FIR
area were recorded using a Bruker FTIR-IS66V-S spectrometer
and the samples were measured in dried PE-pellets.
(b) S. Pramanik, C. Zheng, X. Zhang, T. J. Emge and J. Li,
J. Am. Chem. Soc., 2011, 133, 4153.
14 X. Zhou, H. Li, H. Xiao, L. Li, Q. Zhao, T. Yang, J. Zuo and
W. Huang, Dalton Trans., 2013, 42, 5718.
1
5 Z. Liu, W. He and Z. Guo, Chem. Soc. Rev., 2013, 42,
Acknowledgements
1568.
Support by the Deutsche Forschungsgemeinschaft within the 16 Y. Takashima, V. M. Martinez, S. Furukawa, M. Kondo,
priority program SPP1362 “Metal–Organic Frameworks” is
S. Shimomura, H. Uehara, M. Nakahama, K. Sugimoto and
gratefully acknowledged.
S. Kitagawa, Nat. Commun., 2011, 2, 168.
This journal is © The Royal Society of Chemistry 2015
Dalton Trans.

View Article Online
Paper
Dalton Transactions
1
7 (a) B. Chen, L. Wang, Y. Xiao, F. R. Fronczek, M. Xue, 30 Y. Cui, Y. Yue, G. Qian and B. Chen, Chem. Rev., 2012, 112,
Y. Cui and G. Qian, Angew. Chem., Int. Ed., 2009, 48, 500; 1126.
b) Y.-Y. Jiang, S.-K. Ren, J.-P. Ma, Q.-K. Liu and Y.-B. Dong, 31 Y. Cui, B. Chen and G. Qian, Coord. Chem. Rev., 2014,
(
Chem. – Eur. J., 2009, 15, 10742.
273–274, 76.
1
8 K. Müller-Buschbaum, Z. Anorg. Allg. Chem., 2005, 631, 811. 32 A. Broadbent, Color Res. Appl., 2004, 29, 267.
1
9 K. Müller-Buschbaum, S. G. Torres, P. Larsen and 33 (a) W. J. Evans, D. G. Giarikos and J. W. Ziller, Organometal-
C. Wickleder, Chem. Mater., 2007, 19, 655.
lics, 2001, 20, 5751; (b) J.-S. Li, B. Neumüller and
K. Dehnicke, Z. Anorg. Allg. Chem., 2002, 628, 45;
(c) G. B. Deacon, N. M. Scott, B. W. Skelton and
A. H. White, Z. Anorg. Allg. Chem., 2006, 632, 1945;
(d) J.-S. Li, B. Neumüller and K. Dehnicke, Z. Anorg. Allg.
Chem., 2002, 628, 2785.
2
0 A. Zurawski, M. Mai, D. Baumann, C. Feldmann and
K. Müller-Buschbaum, Chem. Commun., 2011, 47, 496.
1 A. Zurawski, J.-C. Rybak, L. V. Meyer, P. R. Matthes,
V. Stepanenko, N. Dannenbauer, F. Würthner and
K. Müller-Buschbaum, Dalton Trans., 2012, 41, 4067.
2
2
2 J.-C. Rybak, M. Hailmann, P. R. Matthes, A. Zurawski, 34 C. J. Höller, P. R. Matthes, M. Adlung, C. Wickleder and
J. Nitsch, A. Steffen, J. G. Heck, C. Feldmann,
K. Müller-Buschbaum, Eur. J. Inorg. Chem., 2012,
S. Götzendörfer, J. Meinhardt, G. Sextl, H. Kohlmann,
5479.
S. J. Sedlmaier, W. Schnick and K. Müller-Buschbaum, 35 G. M. Sheldrick, SHELXS-97, Program for the solution of
J. Am. Chem. Soc., 2013, 135, 6896.
crystal structures, Göttingen, 1997.
2
2
3 K. Müller-Buschbaum, Z. Naturforsch., 2006, 61b, 792.
4 K. Müller-Buschbaum and Y. Mokaddem, Eur. J. Inorg.
Chem., 2006, 2000.
36 G. M. Sheldrick, SHELXL-97, Program for the refinement of
crystal structures, Göttingen, 1997.
37 (a) A. L. Spek, PLATON-2000, A Multipurpose Crystallographic
Tool, Utrecht, 2000; (b) XPREP v6.12, Program for symmetry
analysis and data reduction of diffraction experiments, Bruker
AXS Inc., Madison, USA, 2001.
2
5 X.-L. Zheng, Y. Liu, M. Pan, X.-Q. Lü, J.-Y. Zhang,
C.-Y. Zhao, Y.-X. Tong and C.-Y. Su, Angew. Chem., Int. Ed.,
2
007, 46, 7399.
2
2
2
2
6 R. S. Rowland and R. Taylor, J. Phys. Chem., 1996, 100, 7384.
7 P. Dorenbos, J. Lumin., 2000, 91, 91.
8 S. P. Nakamura, Soc. Photo-Opt. Ins., 1997, 3002, 26.
9 E. Bernarducci, P. K. Bharadwaj, K. Krogh-Jespersen,
J. A. Potenza and H. J. Schugar, J. Am. Chem. Soc., 1983,
38 J. V. de Mello, H. F. Wittmann and R. H. Friend, Adv.
Mater., 1997, 9, 230.
39 (a) A. L. I. Stevels and A. J. M. Schrama de Pauw, J. Electro-
chem. Soc., 1974, 123, 691; (b) S. Marks, J. Heck,
P. O. Burgos, C. Feldmann and P. W. Roesky, J. Am. Chem.
Soc., 2012, 134, 16983.
1
05, 3860.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2015