Nonradiative De-excitation Mechanisms in Long-Lived Erbium(III

May 7, 2009 - Nonradiative De-excitation Mechanisms in Long-Lived Erbium(III) ... ReceiVed: December 11, 2008; ReVised Manuscript ReceiVed: March 26, ...
3 downloads 0 Views 2MB Size
Download PDF
7474

J. Phys. Chem. B 2009, 113, 7474–7481

Nonradiative De-excitation Mechanisms in Long-Lived Erbium(III) Organic Compounds ErxY1-x[(p-CF3-C6F4)2PO2]3 Ignacio Hernández,*,† R. H. C. Tan,†,‡ J. M. Pearson,‡ P. B. Wyatt,‡ and W. P. Gillin† Department of Physics, and School of Biological and Chemical Sciences, Queen Mary UniVersity of London, Mile End Road, London E1 4NS, U.K. ReceiVed: December 11, 2008; ReVised Manuscript ReceiVed: March 26, 2009

We have performed a spectroscopic study of ErxY1-x[(p-CF3-C6F4)2PO2]3 aimed at understanding nonradiative de-excitation mechanisms. These fluorinated compounds have a long lifetime for the erbium 4I13/2 f 4I15/2 emission at λ ∼ 1540 nm, but the lifetime increases with decreasing x. We have studied the lifetime as a function of morphology, temperature, and high hydrostatic pressure. We have demonstrated the occurrence of energy migration and calculated the corresponding activation energy. Moreover, using high pressure techniques, we provide evidence that cross-relaxation involving energy transfer from an excited erbium in the 4I13/2 promoting a neighbor in the same state to 4I9/2 is the dominant mechanism at ambient conditions for short erbium-erbium distances. The model explains the observed dynamics of excited states in the series and is tested against the Yb[(p-CF3-C6F4)2PO2]3 compound. Introduction Since their discovery,1 erbium-doped fiber amplifiers have become widely used for telecommunications given that erbium(III) ion luminescence (4I13/2 f 4I15/2) at λ ∼ 1.5 µm is in the low loss window for silica. Thus, Er3+ has become extensively used in a number of hosts.2–5 However, amplification over small distances is limited because of the relatively low erbium absorption cross-section and low concentrations (∼1020 ions/cm3) that must be used in all these systems for avoiding quenching of luminescence by clustering or erbium-erbium interactions.6,7 Organic hosts have been proposed and studied as a means to overcome this problem: the organic ligands acting as sensitizers7–11 would result in much higher absorption crosssections than those found for the free erbium ion, and the chelated ions would be kept separated, allowing higher concentrations.7,8 However, ligands and coordinating solvent molecules usually contain C-H and O-H bonds that can cause vibrational quenching of electronically excited rare-earths.10–15 This limits any application of such complexes in infrared emitting devices. Exclusion of coordinated water, deuteration, and fluorination of the ligands can increase the lifetime of infrared luminescence from these lanthanide complexes.7,8,12,14,15 Our group has recently quantified the role of C-H oscillators in the quenching of erbium by deuterating the CsEr(HFA)4, H-HFA ) 1,1,1,5,5,5-hexafluoro-2,4-pentanedione system, which has only one C-H bond per ligand. Any hydrogen within 20 Å acts as an efficient quenching center.16 Mancino et al. have shown that an evaporated thin film of Er(F-tpip)3, where F-tpip ) [(C6F5)2PO]2N-, presents a luminescence lifetime of 224 µs, which is much longer than its nonfluorinated analogue (τ ∼ 5 µs).17 We have recently investigated complexes involving (C6F5)2PO2- (perfluorodiphenylphosphinate) and (p-CF3-C6F4)2PO2- (perfluoroditolylphosphinate) ligands.18,19 According to powder X-ray diffraction, the perfluorodiphenylphosphinates are thought to be isostructural * Corresponding author. E-mail: [email protected]. † Department of Physics. ‡ School of Biological and Chemical Sciences.

with the previously studied20 nonfluorinated versions in the case of trivalent yttrium and rare-earths, and appear to be very stable but highly insoluble coordination polymers.18,21,22 Direct erbium excitation or UV excitation into the ligand states produces infrared luminescence showing relatively long lifetimes between 300 and 500 µs.18,19 However, these perfluorinated compounds show significantly shorter lifetimes than expected from isolated ions, suggesting that other quenching mechanisms are operating. Research has been carried out in diluted erbium materials ErxY1-x[(p-CF3-C6F4)2PO2]3. Erbium concentrations in the polymer were varied from 100 to 0.1 mol % by substituting it with yttrium. Y3+ has approximately the same ionic radius as Er3+ but is optically inert. Dilution leads to an increase of the lifetime to τ ∼ 730 µs for x < 0.4 due to a decrease of the erbium-erbium interaction.19 A comparable decrease has been observed along the ErxY1-x[(C6F5)2PO2]3 series23 and in Er0.5Yb0.5[(C6F5)2PO2]3.21 This work reports new results concerning these materials, aiming at exploring nonradiative quenching mechanisms due to erbium-erbium interactions. The choice of these materials is due to the fact that they are prototypes for active medium in amplifier optical waveguides (although implementable materials could require functionalization for improved solubility). Ideally, visible excitation into the ligands would be more advantageous for the application,7,11 but the lack of visible sensitization is preferred for the present fundamental study, as the ligands are less likely to play a role in the erbium(III) de-excitation dynamics.7 We have focused on compounds with the perfluoroditolylphosphinate ligand as they show a slightly longer photoluminescence lifetime than the perfluorophenylphosphinate, favoring sensitivity to changes. We have studied the photoluminescence quenching as a function of morphology of the material, temperature, and pressure as well as erbium content. Morphology and particle size can affect the optical properties of a material because of the different surface to volume ratios and the existence of preferential directions in the crystals. The influence of the temperature on dynamical processes in photoluminescence properties of materials is well-known. Phenomena

10.1021/jp810932s CCC: $40.75  2009 American Chemical Society Published on Web 05/07/2009

Nonradiative De-excitation Mechanisms

J. Phys. Chem. B, Vol. 113, No. 21, 2009 7475

Figure 1. (left) Energy diagram for the Er3+ ion in ErxY1-x[(p-R-C6F4)2PO2]3 R ) CF3, F showing the approximate energy difference (in corresponding wavenumber units) between excited and ground electronic states. The shaded region represents an estimate of ligand-related low energy excited states, L*.18 Excitation at wavelengths in the UV or in direct excitation in erbium states 2H11/2 to 4I13/2 results in luminescence from the first excited state (1538 nm). (right) Emission spectrum for Er[(p-CF3-C6F4)2PO2]3 at ambient conditions, high pressure and low temperature. Note that changes in the crystal field hardly affect the emission wavelength.

such as upconversion can be enhanced at low temperature, while higher temperatures usually favor energy migration to quenching centers or trap depopulation leading to phosphoresecence. In addition, the temperature dependence of nonradiative de-excitation mechanisms is of interest as it provides the activation energies, helping distinguish which processes are the most relevant at the desired working temperature. High-pressure has been proved of great help in studying dynamical processes involving the excited states of inorganic and organic solids.24–27 Optical spectroscopy under high pressure has been widely used for rare-earth compounds.28 However, research in organolanthanides has been mostly devoted to study the pressure-induced tuning of sensitization of the rare-earth emission (mostly in the visible range) and the interplay between the ligand and the metal, neglecting the interion interaction.25,26 We propose the use of high-pressure techniques for understanding the interaction between rare earths within organic molecular solids. High pressure is a tool to produce denser materials without altering the chemical composition and to contrast results concerning chemical series, since it avoids creation of disorder, or aggregates. Pressure-induced structural changes in soft matter include intermolecular distance reduction, intramolecular distance changes, anisotropic strain, and phase transitions toward closer packing crystalline or amorphous phases.29,30 Pressureinduced energy changes in the electronic states of the organic material can be significant for the ligand states, but the effect of crystal field changes in rare-earths is relatively small.31 Pressure can enhance sensitization of the material, as it has been observed for organic complexes including Eu, Tb, Sm, and Nd,26,31,32 and, in general, we expect ligand-metal and metal-metal interactions to become more intense. However, anisotropic changes and the role of structural phase transitions make it difficult to predict a priori the luminescent behavior of organic or inorganic samples at high pressure. Interaction mechanisms might be affected, leading to photoluminescence enhancement or decrease both in pure and diluted materials.33–35 In addition, high pressure techniques allow discrimination as to whether the interion interaction is mainly due to exchange (short-range) or electric/magnetic multipole (long-range)36,37

based on the dependence of the lifetime on the erbium-erbium distance given the adequate structural characterization. Experimental Section Ligands were prepared as their free acids forms.18,19,23,38 The complexes were obtained by mixing solutions of the ligand acids and metal chlorides in appropriate proportions. Slow mixing procedures in which the two components reacted at the interface of immiscible solvents such as water and diethyl ether were explored as described elsewhere.39 Samples thus produced have been characterized by electron microscopy and spectroscopically. Selected samples were used for temperature and pressure experiments. Electron microscopy was performed at the Nanovision Center in Queen Mary University of London, using FEI Quanta 3D ESEM and FEI Inspect-F electronic microscopes. The Quanta microscope allows measurements in a wet atmosphere, precluding the charging of the insulating material. Carbon or gold coatings were used in the Inspect-F microscope to overcome problems derived from the electrical charging of the material. No appreciable differences in the characteristic properties of the materials were observed by the use of these coatings. Photoluminescence was excited using ∼7 ns pulses from a Continuum Panther optical parameter oscillator (OPO) laser pumped with a Surelite I laser. This laser is tunable in the 415 to 2500 nm range. The setup was arranged in a 90° configuration for illumination detection, together with glass filters for minimizing direct light detection. Excitation at 520 nm was employed for conventional luminescence experiments, given the high oscillator strength of the resonant 4I15/2 f 2H11/2 transition of Er3+. 2H11/2 excited-state decays nonradiatively to the first excited state, 4I13/2, which yields the infrared λ ∼ 1540 nm emission when decaying to the ground state (Figure 1). Luminescence was dispersed in a Triax 550 spectrometer and detected using a Hamamatsu R5509-72 nitrogen-cooled photomultiplier tube. Spectral resolution for spectra was approximately 0.5 nm. Lifetimes were recorded at the peak of the photoluminescence spectrum using a LeCroy LT372 oscilloscope. A resistor box allowed changing the total impedance of the system for signal magnification at the expense of time

7476

J. Phys. Chem. B, Vol. 113, No. 21, 2009

Hernandez et al.

Figure 2. (left) Schematic formula and electron microscopy picture for ErxY1-x[(p-R-C6F4)2PO2]3 R ) CF3, x ) 1. R ) F and/or x < 1 samples show an analogous morphology. (right) Averaged lifetime and stretching factor for the I(t) decay (see text) correlate with the rod diameter in the case of R ) F (squares), CF3 (circles), x ) 1. Lifetime shows no correlation in the case of x * 1 nor with the fiber length for any case.

resolution. The maximum impedance employed was 1 MΩ for the high pressure experiments, allowing an event discrimination time in the range of ∼10 µs (approximately 50 times smaller than the photoluminescence lifetimes). Integration of the corresponding I(t) curves at the selected time-intervals provides the corresponding time-resolved spectra. An Optistat Oxford Instruments continuous flow cryostat allowed measurements in the 77-300 K range and a Linkam LTS 350 hot-stage in the 300-600 K range. The sample chamber was purged with flowing nitrogen during the whole experiment to avoid chemical degradation by the presence of oxygen or water vapor at high temperatures. High-pressure experiments required the adaptation of the conventional spectroscopic setup to work on microsamples within a diamond anvil cell. Small amounts of powdered samples were placed inside the hydrostatic cavity (diameter ∼ 250 µm) that was drilled with a precision mechanical drill in a stainless steel gasket which was previously indented to 80 µm40 (original thickness: 500 µm). The pressure-transmitting medium was 300 000 cst silicone oil (Dow Corning). Some small ruby chips were also included in the cavity for determining the pressure from the redshift of the R1 and R2 photoluminescence lines.41 Stated pressure values and errors represent the average of the measurements obtained at different places and an estimate of the corresponding standard deviation, respectively. Additional errors in the measurements coming from line broadening are taken into account. The hydrostatic cavity was pressurized by a custom-made piston-cylinder miniature diamond anvil cell (DAC)42 equipped with 0.6 mm diameter tip type IIA diamonds. The miniDAC is made out of BeCu, and it is suitable for the range 0-20 GPa. The force was applied externally and continuously through a hydraulic ram, allowing a fine-tuning of the pressure inside the cavity. Experiments at high temperature and high-pressure required a modified cell with 1 mm tip sapphire anvils and stainless steel body, designed to fit the same force-driving mechanism. A 0.6 mm diameter hydrostatic cavity was used, accordingly. High temperatures were achieved through a constant flow of hot air. Temperature was measured by a thermocouple probing the inside of the cylinder of the cell and through the ruby lines.43 Light was focused in the sample chamber within the DAC by means of a high magnification lens (f ) 8 cm, Ø ) 5 cm) allowing selective excitation of the sample and the ruby. A 20× long working distance Mitutuyo microscope objective collected the light into the previously mentioned dispersion and detection

system. Because of the high pulsed laser power and the small size of the light spot, the energy density on the sample and surroundings becomes extremely high. Therefore, we included neutral density filters before the cell, attenuating the light to approximately 5%. The low signal due to the small size of the sample plus the lack of excitation power requires sacrificing time resolution and employing high impedance at the oscilloscope input. Nevertheless, discrimination higher than one fiftieth to one twentieth of the photoluminescence lifetime was kept for high pressure experiments. Raman measurements have been carried out with a Renishaw Ramanscope and a Spectraphysics 127 HeNe laser; the setup was adapted for work at high pressure. Results It has been reported that ErxY1-x[(p-R-C6F4)2PO2]3, R ) F, CF3 show nonexponential I(t) dependence for x > 0.7,18, 19 while the lower erbium concentrations provide single exponential decay curves. This fact has been attributed to the emission from a distribution of perturbed emitting centers, associated with the existence of energy migration. Therefore, the analysis reported in ref 19 was done with an I(t) curve following:

I(t) ) I0 exp[-(t/τ)β]

(1)

(William-Watts or stretched exponential distribution44), where I0 is the initial intensity and 0 < β e 1 is the stretching function. Using this stretched exponential function, it is possible to define a lifetime distribution function and calculate an average relaxation time,〈τ〉

〈τ〉 ) Γ(1/β)τ/β

(2)

where Γ is the Euler Gamma function. Lifetime as a Function of Morphology. Electron microscopy inspection of the ErxY1-x[(p-R-C6F4)2PO2]3 R ) F, CF3, x ) 0.3, 0.5, 1 (Figure 2, left)shows the fibrous nature of these compounds.21 The rod morphology (i.e., diameter and length) varies depending on the synthesis method. Diameters and length of the micro/nanoscopic rods were obtained by direct measurement on the electron-microscopy pictures. These diameters, 〈F〉, represent the average value from pictures taken at different spots. Dispersion was relatively small ( 0.3. Pressure produces a reduction in lifetime over the whole pressure range. Apart from multiphonon de-excitation enhancement through hardening of the bonds, which has been observed to be small (Roman), two different phenomena could, in principle, contribute to it: (1) pressure takes the erbium ions closer (at least in the favored anisotropic strain directions), enhancing metal-metal interaction, similar to a concentration increment;45 (2) high pressure tends to decrease the energy of the ligand states and enhance ligand-erbium interaction.31 The lowest ligand-related states are far from resonance from the emitting Er3+ state, 4I13/2 (Figure 1)18 so no transfer from erbium to ligand can be expected for P < 10 GPa on the basis of the decrease of the activation energy for the 4I13/2 f L* transfer (where L* represents any ligand excited state). Ea(4I13/2fL*) is in the order of magnitude of 2 [E(L*) - E(4I13/2)] (ref 31) which

is estimated to be higher than 1.6 eV (∼13000 cm-1). Even if E(L*) decreases around 62 meV/GPa (50 cm-1/kbar),26 this energy barrier would still be on the order of 1.3 eV for P ∼ 5 GPa, preventing this phenomenon from happening at ambient temperature. However, it could be important in the case of crossrelaxation involving higher states (see below) or for higher pressures than explored. Ligand states are close to higher states involved in the excitation. Therefore, the ligand-erbium sensitization can increase with pressure, accordingly.26,31 However, we cannot confirm this, given the associated experimental errors in the observation of intensity variations; uncertainties arising from changes in geometry in the spectroscopic setup when varying the pressure impede any quantitative analysis. The Er0.3Y0.7[(p-CF3-C6F4)2PO2]3# sample shows a photoluminescence lifetime of 680 µs at ambient conditions. The pressurized sample at 0.8 GPa shows a lifetime τ ∼ 530 µs, smaller than that of Er0.5Y0.5[(p-CF3-C6F4)2PO2]3# at room temperature (τx)0.5# ) 570 µs). Thermal dependence at high pressure elucidates whether the pressure increase to 0.8 GPa reduces the photoluminescence lifetime because of a decrease in Ea for migration, or by triggering nonradiative mechanisms by the shortening of distances (at this moderate pressure, activation of transfer to the ligand and multiphonon de-excitation probability can still be disregarded, as explained above). According to the thermal behavior at ambient pressure of the ErxY1-x[(p-CF3-C6F4)2PO2]3 series, the activation energy for excitation migration can be significantly reduced for x g 0.5 (changing from Ea ∼ 205 meV for x ) 0.3 to Ea ∼ 55 meV for x ) 0.5). If the effect of pressure was only due to changes in the Arrhenius term for nonradiative de-excitation probability to a value comparable to that of Er0.5Y0.5[(p-CF3-C6F4)2PO2]3#, it would lead to changes in the lifetime of more than 80 µs on passing from T ∼ 300 to 450 K. However, no appreciable changes can be obtained in this temperature range. It follows that changes in the Arrhenius term alone do not account for the enhancement of nonradiative processes and the activation energy for transfer from the Er3+ first excited-state to the ligand excited states remains negligible. Therefore, it can be concluded that high pressure induces the appearance of another contribution to τnr-1, depending on the erbium-erbium distance. Moreover, the temperature results suggest that this additional contribution also depends on temperature, reaching a minimum between 300 and 400 K for all samples. We have not found traces of cooperative upconversion from any of the possible states. Moreover, the photoluminescence lifetimes for excitation at 520, 800, and 978.1 nm are the same. In light of this, we disregard energy transfer mechanisms involving excited states other than 4I13/2. If they were significant, different photoluminescence lifetimes would be obtained at the different excitation wavelengths, as cross-relaxation channels would be accessible from certain excitation energies and not from the others. We propose that the luminescence in the erbium chelate is quenched by a cross-relaxation mechanism. The proposed mechanism is described in Figure 5. An erbium ion in the 4I13/2 excitedstate promotes a surrounding excited ion to a higher excited state.46 The promoted ion then relaxes nonradiatively to the first excitedstate or to the ground-state via excitation migration or transfer to the ligand, or two excited ions could sensitize the ligand cooperatively. The relatively long life of the first erbium excited-state and the proximity of energy E(4I9/2) ∼ 2 E(4I13/2) or E(L*) - 2 E(4I13/2) ∼ 0 favors these processes.

7480

J. Phys. Chem. B, Vol. 113, No. 21, 2009

Hernandez et al. Regarding the abrupt changes in photoluminescence lifetime at 2 GPa, the correlation with changes in the Raman spectrum relates it to a structural phase transition at this pressure. A phase transition (specially first-order) involves abrupt structural changes, affecting the lifetime.48,49 This kind of phenomenon can also explain the abrupt changes observed at T ∼ 125 K and in the case of Yb[(p-CF3-C6F4)2PO2]3 at T ∼ 135 K even though the temperature-induced transition does not necessarily lead to the same crystalline phase as the pressure-induced one. Further structural characterization work is planned in order to analyze this phenomenon, as well as to relate the pressure and chemical data to effective intermolecular or erbium-erbium geometry and work out which transfer mechanisms, and in what directions, are the most relevant in these systems. Conclusions

Figure 5. Cross-relaxation de-excitation for erbium pairs in concentrated ErxY1-x[(p-CF3-C6F4)2PO2]3 materials. Solid arrows represent two possible excitations, curly arrows nonradiative de-excitation, the dotted gray arrow radiative decay, and broken arrows energy transfer. The described mechanism represents energy transfer from an erbium in its first excited state, I13/2, to a neighbor excited ion, promoting it to the 4 I9/2state, with an associated energy E(I9/2) ∼ 2 E(I13/2). Subsequent nonradiative de-excitation from 4I9/2 implies the loss of photons. This cross-relaxation process is temperature-dependent through the density of states at the different energies and accessible to excitations between 520 and 971.8 nm. It is not possible in Yb complexes, which have only one accessible excited state, resulting in a different experimental behavior.

Changes in the ion-ion interactions, which are more probable at higher erbium concentrations and/or high pressure, account for the corresponding dependences. The increase of lifetime when heating in the 80-400 K range is explained by the temperature-induced reduction of this mechanism through a decrease of the otherwise overlapping density of states47 at the resonant energy. The proposed model is confirmed by the thermal behavior of Yb[(p-CF3-C6F4)2PO2]3 showing a monotonic increase of lifetime from 420 µs at 300 K to 526 µs at 130 K. The different behavior with respect to the erbium samples is explained in terms of the existence of a single excited state, 2F5/2, preventing the cross-relaxation described in Figure 5. Moreover, we can now understand why the Er0.5Yb0.5[(C6F5)2PO2]3 sample studied in ref 21 shows a comparable lifetime to that of the ErxY1-x[(pCF3-C6F4)2PO2]3, x e 0.3. Thus, τ-1 for the ErxY1-x[(p-CF3-C6F4)2PO2]3 systems is given by eq 3 with τrad-1 ∼ 0.00133 µs-1 and nr -1 nr -1 τ-1 nr ) τmigration (x,T) + τcross-rel (x,T)

τnrmigration-1

(6)

where represents the probability of excitation migration along the crystal and depends explicitly on temperature as given by eq 4. τnrcross-rel-1 stands for the nonradiative deexcitation probability due to cross-relaxation between erbium ions involving transfer from 4I13/2 (Figure 5). τnrcross-rel-1 depends on temperature and strongly on the concentration and, equivalently, the erbium-erbium distance or pressure. Changes from samples belonging to one chemical series to the ones in other series can be accounted for by an additional τnrimpurities-1 which depends on the presence of impurities due to the synthesis procedure. Additional multiphonon de-excitation channels may be active as well, although they are, apparently, not very significant for explaining the thermal and series evolution of the lifetime. Their role could be compared to that of the impurities, but equal for all the series.

We have studied the nature of photoluminescence quenching mechanisms in the ErxY1-x[(p-CF3-C6F4)2PO2]3 series. We have considered the influence of the morphology showing that larger diameter rods of the pure-erbium compounds of the series present a shorter photoluminescence lifetime. Diluted samples present no correlation between the sample morphology and photoluminescence lifetime. We have related these facts to the presence of excitation transfer and centers that are able to quench the photoluminescence. Low temperature behavior of the lifetime, τ(T), along the ErxY1-x[(p-CF3-C6F4)2PO2]3 series varies with the erbium concentration: x ) 0.3 shows no temperature-induced variations between 80 and 400 K, x ) 1 shows a decrease upon cooling (including an abrupt decrease at T ∼ 125 K), and x ) 0.5 shows a slight decrease followed by saturation starting at T ∼ 220 K. Heating above 400 K results in a monotonic decrease for all samples. The low temperature results are contrary to observations in Yb[(p-CF3-C6F4)2PO2]3, which show a continuous decrease in the lifetime from 440 µs at 300 K to 530 µs at T ) 135 K. High temperature behavior is explained through an Arrhenius type law for nonradiative processes related to excitation migration, the activation energy varying with the erbium content, x. However, the low temperature dependence for x > 0.3 requires another term accounting for cross-relaxation. An excited erbium ion in its first excited-state exchanges its energy with another excited erbium, promoting it to higher energy, which then deexcites nonradiatively, or equivalently both cooperatively transfer to low-lying ligand states. This is supported by the absence of upconversion, the fact that the intensity decay, I(t), does not depend on the excitation wavelength and the Yb[(pCF3-C6F4)2PO2]3 lifetime evolution at low temperatures, which is contrary to the one exhibited by the erbium compound. We have showed that compression of Er0.3Y0.7[(p-CF3C6F4)2PO2]3 leads to an important decrease in the lifetime, similar to that seen with the increase in erbium concentration. Er[(p-CF3-C6F4)2PO2]3’s lifetime also experiences a decrease when pressure is increased. This is due to enhancement of Er-Er interactions unlike what has been observed in the case of visible luminescence for Eu, Nd, and Pr chelates in which it was associated with metal-to-ligand energy transfer.31 High pressure and high temperature experiments for Er0.3Y0.7[(p-CF3C6F4)2PO2]3 show that the lifetime for this compound at P ∼ 0.8 GPa is hardly dependent on temperature. This proves that the cross-relaxation is favored not only by concentration but by pressure, highlighting the importance of the intermolecular distance. We have shown the potential of high-pressure techniques in explaining the optical behavior of soft materials. In particular

Nonradiative De-excitation Mechanisms we have employed them to study nonradiative mechanisms depending on intermolecular interactions. Further work is planned in order to study the pressuredependence of erbium-erbium geometry and accurately characterizing phase transitions. This may reveal the evolution of lifetime as a function the crystal structure, providing further information on the nature of the interion interaction and preferential pressure-induced strains and their influence on the optical properties of these materials. Acknowledgment. We thank Prof. D. J. Dunstan and G. Gannaway for the high-pressure equipment and Dr. M. K. P’ng for help with the electron microscopes at Nanovision. We also thank EPSRC National Mass Spectrometry Service Centre, Swansea, for mass spectroscopy measurements. This work has been financially supported by the Royal Academy of Engineering and the EPSRC. References and Notes (1) Mears, R. J.; Reekie, L.; Jauncey, I. M.; Payne, D. N. Electron. Lett. 1986, 22, 159. (2) Mears, R. J.; Reekie, L.; Jauncey, I. M.; Payne, D. N. Electron. Lett. 1987, 23, 1026. (3) Miniscalco, W.J. J. LightwaVe Technol. 1991, 9, 234. (4) Kaminow, I. P.; Li, T. Optical Fiber Telecommunications IV: Components; Academic Press: New York, 2002. (5) Kik, P. G.; Polman, A. MRS Bull. 1998, 23, 48. (6) van den Hoven, G. N.; Snoeks, E.; Polman, A.; van Dam, C.; van Uffelen, J. W. M.; Smit, M. K. J. Appl. Phys. 1996, 79, 1258. (7) Sloof, L. H.; van Blaaderen, A.; Polman, A.; Hebbink, G. A.; Klink, S. I.; van Veggel, F. C. J. M.; Reinhoudt, D. N.; Hofstraat, J. W. J. Appl. Phys. 2002, 91, 3955. (8) Kuriki, K.; Koike, Y.; Okamoto, Y. Chem. ReV. 2002, 102, 2347. (9) Lis, S.; Elbanowski, M.; Makowska, B.; Hnatejko, Z. J. Photochem. Photobiol., A 2002, 150, 233. (10) Curry, R. J.; Gillin, W. P. Curr. Opin. Solid State Mater. Sci. 2002, 5, 481. (11) Comby, S; Bu¨nzli, J. C. Handbook on the Physics and Chemistry of Rare Earths; Gschneidner, K. A.; Bu¨nzli J. C. G.; Pecharsky V. K. Eds., Elsevier:Amsterdam, 2007; Vol. 37. (12) Heller, A. J. Am. Chem. Soc. 1966, 88, 2058. (13) Stein, G.; Wu¨rzberg, E. J. Chem. Phys. 1975, 62, 208. (14) Winkless, L.; Tan, R. H. C.; Zheng, Y.; Motevalli, M.; Wyatt, P. B.; Gillin, W. P. Appl. Phys. Lett. 2006, 89, 111115. (15) Hasegawa, Y.; Wada, Y.; Yanagida, S. J. Photochem. Photobiol., C 2004, 5, 183. (16) Tan, R. H. C.; Motevalli, M.; Abrahams, I.; Wyatt, P. B.; Gillin, W. P. J. Phys. Chem. B 2006, 110, 24476. (17) Mancino, G; Ferguson, A. J.; Beeby, A.; Long, N. J.; Jones, T. S. J. Am. Chem. Soc. 2005, 127, 524. (18) Zheng, Y.; Pearson, J.; Tan, R. H. C.; Gillin, W. P.; Wyatt, P. B. J. Mater. Sci.: Mater. Electron. 2009, 20, S430.

J. Phys. Chem. B, Vol. 113, No. 21, 2009 7481 (19) Tan, R. H. C.; Pearson, J. M.; Zheng, Y.; Wyatt, P. B.; Gillin, W. P. Appl. Phys. Lett. 2008, 92, 103303. (20) Dunstan, P. O.; Vicentini, G. An. Acad. Bras. Cienc. 1975, 47, 97. (21) Song, L.; Hu, J.; Wang, J.; Liu, X.; Zhen, Z. Photochem. Photobiol. Sci. 2008, 7, 689. (22) Stucchi, E. B.; Scarpari, S. L.; Couto Dos Santos, M. A.; Leite, S. R. A. J. Alloys Compd. 1998, 89, 275. (23) Tan, R. H. C. Synthesis and characterization of perfluorinated lanthanide complexes for NIR applications, Ph.D. Thesis, Queen Mary University of London, 2008. (24) Drickamer, H. G.; Mitchell, D. J.; Schuster, G. B. in Radiationless Transitions; Lined, S. H. Ed., Academic Press: New York, 1980; pp 289316. (25) Yersin, H.; Bray K. L. Eds. Transition Metal and Rare Earth Compounds: Excited States, Transitions, Interactions; Springer: Berlin, 2001. (26) Drickamer, H. G. Annu. ReV. Phys. Chem. 1982, 33, 25. (27) Edwards, C. M.; Butler, I. S. Coord. Chem. ReV. 2000, 199, 1. (28) Tro¨ster, T. In Handbook on the Physics and Chemistry of Rare Earths; Gschneidner, K. A.; Bu¨nzli J. C. G.; Pecharsky V. K. Eds., Elsevier: Amsterdam, 2003; Vol. 33, pp 515-589. (29) Katrusiak, A. Cryst. Res. Technol. 1991, 26, 523. (30) Boldyreva, E. V. Acta Crystallogr., Sect. A: Found. Crystallogr. 2008, 64, 218. (31) Hayes, A. V.; Drickamer, H. G. J. Chem. Phys. 1982, 76, 114. (32) Zeng, Q, G.; Zhang, Z. M.; Ding, Z. J. Solid State Commun. 2006, 138, 86. (33) Herna´ndez, I.; Rodrı´guez, F. Phys. ReV. B 2003, 67, 012101. (34) Rodriguez, F.; Hernandez, I.; Moreno, M.; Alcala, R. J. Chem. Phys. 2003, 119, 8686. (35) Hernandez, I.; Rodrı´guez, F.; Hocheimer, H. D. Phys. ReV. Lett. 2007, 99, 027403. (36) Dexter, D. L. J. Chem. Phys. 1953, 21, 863. (37) Forster, T. Ann. Phys. (Leipzig, Ger.) 1948, 2, 55. (38) Oliver, K. W.; Rettig, S. J.; Thomson, R. C.; Trotter, J.; Xia, X. J. Fluorine Chem. 1997, 83, 47. (39) Pearson, J. M. Synthesis and Characterization of Luminescent Lanthanide Complexes containing Fluorinated Organic Ligands, M.Sc. Thesis, Queen Mary University of London, August 2007. (40) Dunstan, D. J. ReV. Sci. Instrum. 1989, 60, 3789. (41) Forman, R. A.; Piermarini, G. J.; Barnett, J. D.; Block, S. Science 1972, 176, 284. (42) Dunstan, D. J.; Scherrer, W. ReV. Sci. Instrum. 1988, 59, 627. (43) Vos, W. L.; Schouten, J. A. J. Appl. Phys. 1991, 69, 6744. (44) Lindsey, C. P.; Patterson, G. D. J. Chem. Phys. 1980, 73, 3348. (45) Inokuti, M.; Hirayama, F. J. Chem. Phys. 1965, 43, 1978. (46) Shi, W. Q.; Kurtz, R.; Machan, J.; Bass, M.; Birnbaum, M.; Kokta, M. Appl. Phys. Lett. 1987, 51, 1218. (47) van der Ziel, J. P.; Ostermayer, F. W.; van Uitert, L. G. Phys. ReV. B 1970, 2, 4432. (48) Rivera-Lopez, F.; Martin, I. R.; Da Silva, I.; Gonzalez-Silgo, C.; Rodriguez-Mendoza, U. R.; Lavin, V.; Lahoz, F.; Diaz-Gonzalez, S. M.; Martinez-Sarrion, M. L.; Mestres, L.; Fernandez-Urban, J. High Press. Res. 2006, 26, 355. (49) Ramirez, M. O.; Bausa´, L. E.; Speghini, A.; Bettinelli, M.; Ivleva, L.; Garcı´a Sole´, J. Phys. ReV. B 2006, 73, 035119.

JP810932S