Research Update : Preserving the photoluminescence efficiency of near infrared emitting nanocrystals when embedded in a polymer matrix

Near infrared light emitting nanocrystals are known to lose efficiency when embedded in a polymer matrix. One of the factors leading to reduced efficiency is the labile nature of the ligands that may desorb off the nanocrystal surface when the nanocrystals are in the polymer solution. We show that adding trioctylphosphine to the nanocrystal-poly(methylmethacrylate) solution prior to film casting enhances the photoluminescence efficiency. The solid films’ photoluminescence quantum efficiency values are reduced by less than a factor of two in the solid form compared to the solution case. We demonstrate record efficiency values of 25% for lead sulfide nanocrystals solid films emitting at 1100 nm.

Nanoscale inorganic materials display unique size, shape, and composition dependent electronic and optical properties.][6][7][8][9][10] Despite the progress made in many aspects of nanocrystal's synthesis and devices made of them, there seems to remain an issue that holds back some applications.There are many reports that although the photoluminescence (PL) efficiency of near infrared (NIR) emitting NCs in solution could be 20% or above, the quantum efficiency (QE) measured for solid films is in the range of 1% and often below it. 7To maintain above 1% efficiency, a higher gap shell is introduced 11 and an almost 4%, 12 or 12% for low NC loading films, 13 is considered very high.
Our studies of the role of ligands in affecting the energy levels of nanocrystals 2,3,10 have emphasized the role of ligand exchange which is based on the fact that most ligands would dynamically detach and reattach to the nanocrystal's surface.This has led us to postulate that diluting the nanocrystals in a matrix also dilutes the ligands and thus favors the situation where part of the nanocrystal's surface becomes non-passivated thus dropping the PL efficiency.Ligand desorption in dilute solutions and the role of surface coverage in affecting the emission quantum yield have been reported in Refs.14-16.Here we report that adding excess ligands to the solution of, commercially available, lead sulfide PbS nanocrystal results is in PL efficiency, of NCs embedded in a poly(methylmethacrylate) (PMMA) matrix, of 25% and 12% for nanocrystals emitting at 1100 nm and 1500 nm, respectively.
For this study, we purchased two types of commercially available colloidal lead sulfide (PbS) nanocrystals (NCs) series C (SCR) from CAN GmbH (www.can-hamburg.com)as solutions in toluene.The NCs size distribution was slightly improved through size selection procedures carried out by the manufacturer.A polymer host, poly(methylmethacrylate) PMMA solution in toluene, and additives of trioctylphosphine (TOP) and/or aniline, both purchased from Aldrich, were used to prepare free standing films from polymer-NCs blends.The use of relatively thick films was to ensure sufficient absorption in the NIR to ensure accurate efficiency measurements (see method below).To characterize the as-received NCs, we used a 1 mm quartz cuvette to hold the NCs toluene solutions.Near-infrared (NIR)-visible absorption spectra were measured using UV-3101 PC spectrophotometer (Shimadzu Scientific Instruments, Inc.) in 700-1700 nm region.Photoluminescence (PL) and photoluminescence quantum efficiency (PL QE) measurements were performed using the integrated system, based on the FS920 fluorimeter (Edinburgh Instruments Ltd., UK), equipped with liquid nitrogen cooled germanium photo-detector with lock-in amplification and an integrating sphere (Labsphere, Inc.IS-040-SL with UV-VIS-NIR reflectance coating).The labsphere was fiber coupled to the FS920 and excited by the monochromatic xenon lamp (450 W) light at 886 nm.The entire system response was normalized by a calibrated detector (Newport 818 IR) and a multi-function optical meter (Newport 1835C) in 800-1700 nm region.The PL QE was performed following the procedure described in Ref. 17.
Figure 1 shows the absorption (dashed line) and emission (full line) spectra for the 1100 nm (Figure 1(a)) and the 1500 nm (Figure 1(b)) batches.The first excitonic peak absorption is found to be at ∼1000 nm and ∼1400 nm and the PL emission peak is found to be at about 100 nm longer wavelength.The PL QE measurements described above yielded QE in solution of 45% and 20% for the 1100 nm and 1500 nm batches, respectively.
Next, we moved to examine the PL properties of the NCs when embedded within a polymer matrix.As oxygen may affect the NC surface and lead to ligand desorption, 18,19 all film preparations were carried out inside an inert glove box (<1 ppm O 2 and H 2 O).Free standing films were produced from blends of PMMA solution in toluene (60 mg/ml) with ∼4 wt.% (∼1 vol.%) of NCs and where applies also ∼5 vol.% of additives.The blend was prepared in a small glass Petri dish and stirred for 1 h using a magnetic stirrer.The blend was kept inside the glove box during the solvent evaporation and once dry the film was easily removed from a glass and finally dried for 3 h in the vacuum oven at 60 • C. Thicknesses of prepared films were in the range 0.2-0.3mm.
The empty round symbols in Figure 2 show the absorption (dashed line) and emission (full line) spectrum of the 1100 nm (Figure 2(a)) and 1500 nm (Figure 2(b)) batches in a PMMA matrix.We note that, compared to Figure 1, the excitonic peak in the absorption is broadened and less pronounced and that the emission spectrum is red shifted.The PL QE was measured to be 10% and 6% for the 1100 nm and 1500 nm PMMA based films, respectively.The ∼4 fold reduction in PL QE together with the smearing of the excitonic peak and its red shift suggests that the surface of the nanocrystals was damaged and that despite the low volume fraction, aggregates might have formed.The labile nature of the ligands is often used to perform ligand exchange 2,20 and it is known that the process propagates as the original ligand desorbs and the excess amount of the new ligand FIG. 1. Absorption (dashed line) and emission (full line) spectrum of the ∼1100 nm and the ∼1500 nm NCs measured as solution in toluene.The corresponding quantum efficiency was 45% and 20% for the 1100 nm and 1500 nm batches, respectively.ensures that the adsorption would be of the new ligand.It is also known that the presence, or lack of, ligands affects the nanocrystal's optical properties and that at least some of the ligands covering the NC may be removed in dilute solutions. 14Considering the labile nature of the ligands and the excess amount of PMMA, it is most likely that at least some of the ligands were detached from the NCs surface and dispersed in the polymer solution and/or matrix.We thus attribute the decrease in the NCs properties to this ligand desorption mechanism.Assuming that our postulation of the ligand desorption is correct, one could follow the same logic used in the ligand-exchange procedures and provide excess amount of free ligands such that when a ligand is desorbed, there would be another ligand, as a reservoir, to take its place and re-passivate the surface.To this end, we repeated the film preparation procedure and this time added 5 vol.% of TOP to the NC-PMMA solution.We chose TOP as it was shown that synthesis resulting in TOP capping produced highly luminescent PbS nanocrystals. 21The full circles in Figure 2 show the absorption (dashed line) and emission (full line) spectrum of the 1100 nm (Figure 2(a)) and 1500 nm (Figure 2(b)) batches in a PMMA+TOP matrix.Note that both the absorption and the emission characteristics are largely recovered following the addition of TOP.The PL QE measurements yielded efficiencies of 25% and 12% for the 1100 nm and 1500 nm batches on PMMA+TOP matrix, respectively.Such efficiencies for NIR emitting NCs in a solid matrix are clearly record values.
While adding TOP resulted in improved optical properties, we noted that the film became less uniform and there was a clear vertical phase segregation.We tried using aniline instead of TOP and while the film properties were recovered, the optical properties were almost as if no ligand was introduced (PL QE was 14%).We next tested a 50:50 mix of TOP:aniline and found that both the film properties and the optical properties were improved although not as high as TOP only films.The PL QE of TOP:aniniline PMMA films was 20% for the 1100 nm batch.The PL spectrum taken as part of this TOP/aniline study is shown in Figure 3 clearly indicating the importance of TOP.
Finally, we need to ascertain that the role of TOP involves an attachment of TOP to the nanocrystals and at least partial exchange of the oleic acid ligands.To do so, we have followed the protocol described in Ref. 2 where it was shown that partial exchange of ligands leads to stark shift of the absorption peak.Knowing the wt.% of the nanocrystals in the solution, one can estimate the total weight of ligands that are attached to the nanocrystals (assuming ideal coverage).Then a weighted amount of TOP is added to create the relative percentage shown in Figure 4. Figure 4 shows the absorption spectrum around the peak of the first exciton peak for two PbS solutions.The first (solid line) is for the as-received solution with all the nanocrystals being covered by oleic acid.The second (dashed line) is for a solution where TOP was added to form 20% of the total weight of ligands.Figure 4 shows a clear red shift and the inset shows that this shift increased between 0%, 11%, and 20% in qualitative agreement with previously reported results for CdSe. 2 To conclude, we have demonstrated the record PL efficiency values for nanocrystal based solid films emitting above 1000 nm.The commonly reported low efficiency values are attributed to ligand desorption that may be mediated by extrinsic presence of oxygen 18,19 or by the labile nature the ligands where the NCs are diluted in a, polymer, solution. 14By processing the films in inert atmosphere, we achieved PL QE of 10% and 6% for the 1100 nm and 1500 nm batches in PMMA, which is similar to the values in Ref. 13. Adding the trioctylphosphine (TOP) to the NC-PMMA solution prior to film casting enhanced the PL QE to record values of 25% and 12% for PbS nanocrystals solid films emitting at ∼1100 nm and ∼1500 nm, respectively.We expect that studies of different ligands (as hexadecylamine, HDA) 14 may further boost the PL QE of solid films.As the method is general, it should also improve the NCs performance when the polymer is a semiconducting one and thus lead to higher efficiency LEDs and possibly also to better photo-cells. 22ga Solomeshch is grateful for the support by the Center for Absorption in Science of the Ministry of Immigrant Absorption and the Committee for Planning and Budgeting of the Council for Higher Education under the framework of the KAMEA Program.

040702- 3 O
FIG.2.Absorption (dashed line) and emission (full line) spectrum of the ∼1100 nm (a) and the ∼1500 nm (b) NCs measured as solid films.The empty symbols are for PbS NCs in PMMA matrix and the full symbols are for the case where TOP is added to the NC-PMMA solution prior to film casting.The quantum efficiency values of the TOP containing films were 25% and 12% for the 1100 nm and 1500 nm batches, respectively.

040702- 4 OFIG. 4 .
FIG. 3. Emission spectrum of the ∼1100 nm NCs batch in a PMMA matrix and for three different additives.The round symbols are for TOP additive, the diamonds are for aniline, and the squares are for 50:50 TOP: aniline.