ZnS Quantum Dots as Noncontact Probes

Jan 13, 2015 - Collin R. Howder , Bryan A. Long , Dieter Gerlich , Rex N. Alley , and Scott L. Anderson. The Journal of Physical Chemistry A 2015 119 ...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/JPCC

Thermally Brightened CdSe/ZnS Quantum Dots as Noncontact Probes for Surface Chemistry Studies of Dark Nanoparticles Trapped in the Gas Phase Collin R. Howder, Bryan A. Long, David M. Bell, and Scott L. Anderson* Department of Chemistry, University of Utah, Salt Lake City, Utah 84112, United States S Supporting Information *

ABSTRACT: Charged CdSe/ZnS quantum dots (QDs) trapped in the gas phase are transformed by CO2 laser heating, resulting in brightening of their photoluminescence intensities by more than 2 orders of magnitude. The transformation is shown to be thermally driven, and self-limiting, i.e., once the QDs have been fully brightened, they are unaffected by further CO2 laser irradiation at intensities up to 1 kW/cm2. The transformation clearly involves loss of the 10.6 μm chromophore, which appears to be the ligand layer. The thermally brightened QDs are tested for use as noncontact probe particles, allowing dark nanoparticles to be detected by cotrapping them with a brightened QD. We show that cotrapping has negligible effect on the secular frequency of the cotrapped particles, and that it is possible to simultaneously monitor changes in the mass and charge of up to three cotrapped nanoparticles for hours, allowing the rate of surface reactions to be measured. Application to studying the surface chemistry of free nanoparticles is discussed. easily detected by light scattering,21 and laser-induced fluorescence/luminescence works for NPs such as dye-doped polymer spheres35,39 or semiconductor nanocrystal quantum dots (QDs).39,40 Unfortunately, many NPs of interest from a surface chemistry perspective are too small for light scattering detection, and are nonluminescent. In principle, such NPs could be monitored by labeling them with one or more fluorescent probe molecules or NPs, however, the probes would tend to perturb the small dark particles, and might not remain bright under reaction conditions. Here we demonstrate a noncontact approach to optically probing dark NPs by cotrapping with a bright probe NP, and demonstrate that thermally brightened CdSe/ZnS quantum dots are near-ideal probe NPs. In the NPMS method used here, a charged NP is trapped in a 3D quadrupole (Paul) ion trap by an applied radio frequency (rf) potential with voltage V0 and angular frequency Ω, under conditions21 where the particle motion is well described as slow, large amplitude secular motion, with superimposed small amplitude driven motion at frequency Ω. For an ideal quadrupole trap, the secular motion is harmonic with angular frequencies ωz and ωr for motion in the axial and radial directions, respectively. From the secular frequency, the mass-

I. INTRODUCTION Many nanoscience problems involve reactions of nanoparticles that are small enough (100 nm) are © XXXX American Chemical Society

Special Issue: Steven J. Sibener Festschrift Received: October 30, 2014 Revised: January 11, 2015

A

DOI: 10.1021/jp5109027 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

II. EXPERIMENTAL METHODS Our instrument, described in detail elsewhere,48 uses electrospray ionization (ESI − Micromass Z spray) to put charged nanoparticles in the gas phase. Mercaptoundecanoic acidcapped aqueous CdSe/ZnS core−shell QDs (5.5 nm CdSe core, 1 nm ZnS shell, solution emission λmax = 650 nm) were obtained from NN laboratories in concentrations of 1 mg/mL. The QD stock solution is a pH 8 Na+-containing buffer solution, and is diluted 100:1 in methanol and electrosprayed, generating particles including QD monomers and small aggregates of 2−5 QDs with charges ranging from ∼+3e to +20e per QD. Under these conditions, the acid groups should be largely ionized, and positive charge should be due to excess Na+. After ESI, gas phase QD ions are thermalized in a hexapole ion guide (P = 8−12 mTorr argon), then enter a quadrupole ion guide used as a mass prefilter to reject ions or particles with M/Q outside the range of interest. The nanoparticles are then injected into a split-ring-electrode quadrupole (Paul) ion trap, based on a design by Gerlich et al.49 A pneumatically operated valve, built into an ion lens, separates the hexapole and quadrupole and is used to isolate the trap chamber vacuum from the high flux of gas entering from the ESI source. The valve is opened only for ∼10 s at a time for trap filling. To slow QDs for trapping, the pressure in the trap region is set to ∼5−15 mTorr. For the experiments discussed below, the hexapole, quadrupole, and trap were operated at a frequency of 145 kHz, with amplitude of 500 V. Trapped QDs are detected optically, via 532 nm laserinduced PL. Because the PL quantum yield is below unity, and the emission is Stokes-shifted, a significant fraction of the 532 nm photon energy is retained in the QDs as heat. To prevent QD overheating and sublimation, the trap chamber is flooded with argon buffer gas to a pressure of 1 to 20 mTorr whenever the 532 nm laser is on. A continuous wave (cw) 532 nm laser with 500 mW maximum power (Ultralasers) was loosely focused through the trap, with power measured at the trap exit. The focused beam diameter at the trap center was measured by trapping a single polystyrene particle (d ∼ 100 nm), and then scanning the laser focus through the center of the trap while recording the intensity of light scattered by the particle. Taking the radius of thermal motion of the particle (∼30 μm) into account,50 the beam waist was estimated to be 500 μm in diameter, and the corresponding laser intensities were typically kept in the 50−150 W/cm2 range. PL signal is collimated by an aspheric lens, passed through a 532 nm notch filter to remove scattered light, and then focused (with 2× magnification) onto an avalanche photodiode (APD) detector with 100 μm × 100 μm active area (Laser components Count 50). The detection volume is determined by the overlap between the APD detection area and the laser focus and is estimated to be ∼50 μm × 50 μm × 500 μm. A filter wheel in the beam path can be used to insert long pass color filters (Thorlabs FGL) to make spectral measurements. Spectra are determined by subtracting signals measured through the sequence of filters, with background measured using analogous experiments made on an empty trap. All spectra are corrected for the avalanche photodiode (APD) detection efficiency, which is ∼30% at 400 nm, 85% at 650 nm, dropping to ∼13% at 1000 nm, and to ∼6% at 1100 nm. Light (10.6 μm) from a duty-factor-modulated, quasi-cw CO2 laser (Synrad 10 W) is used for particle heating. The CO2 laser enters and exits the instrument through ZnS Cleartran

to-charge ratio, M/Q, can be determined, as shown for the case of axial motion:

2 V0 M = Q ωz Ωz 02 where z0 is a trap geometric parameter (2.96 mm). Q, and hence M, can be determined by observing a series of steps in M/Q caused by charge-changing collisions between the NP and ions or electrons in the trap background (Q/ΔQ = ωz/Δωz). Secular motion is detected by measuring variations in light scattering or photoluminescence signal as the NP oscillates in the trap, thus modulating its overlap with the detection volume. The envisioned noncontact probe approach involves trapping a dark NP of interest together with an optically bright probe NP with the same charge polarity, so that the two NPs repel each other. Coulomb repulsion also couples the motion of the two NPs, so that in principle, the motion and properties of the dark NP can be inferred from its effects on the motion of the bright NP. As demonstrated below, for the correct coupling strength, the method works and is quite straightforward to interpret. The probe NP must obviously be optically bright under all the conditions of interest, but it also must have mass that is less than or comparable to that of the dark NPs of interest, so that the dark NP can scatter it out of the detection volume. The size range of greatest interest for NP surface chemistry is below ∼10 nm, thus probe NPs with masses in the 105 Da range are needed. For NP surface chemistry experiments, it will be necessary to heat the dark NP by CO2 laser irradiation (10.6 μm) for cleaning/annealing, and the NPs will also be exposed to various gaseous reactants. Therefore, the probe NP must survive intense CO2 laser irradiation and reactant exposures, remaining luminescent and not undergoing large M or Q changes, which would complicate analysis. CdSe/ZnS core/ shell QDs are available with masses in the ∼1−7 × 105 Da range and, as shown below, have properties that appear almost ideal for use as probe NPs. Semiconductor QDs have been studied in detail in the condensed phase,41−43 where they are often used as luminescent probes, however, few studies of QDs in the gas phase have been reported, with the notable exception of Xiong et al., who measured photoemission from gas-phase QDs.44 We reported NPMS studies of gas-phase CdSe/ZnS core/shell QDs,39,40 demonstrating the possibility of monitoring a single QD for several days while continuously measuring M, Q, and the intensity and spectrum of photoluminescence (PL) from the QD. As-trapped QDs have low PL intensities, however, upon heating, the PL brightens dramatically and red shifts. In condensed phase measurements, thermal damage also leads to red-shifted emission attributed to creation of surface trap states,45−47 however, trap state creation in the condensed phase is associated with decreased PL intensity. This report has several goals. The thermal brightening process is explored in more detail, and then brightened QDs are examined with respect to their suitability for use as probe particles. The effects on M, Q, and PL intensity of both prolonged CO2 laser irradiation and exposure to oxidizing environments at elevated temperatures are examined. Finally, the noncontact probe method is demonstrated, by measuring M/Q for a dark particle via its influence on the motion of a cotrapped probe QD, and simultaneous measurement of M and Q for three cotrapped QDs. B

DOI: 10.1021/jp5109027 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

second relative to the 532 nm baseline (i.e., ∼12 counts per 100 ms time bin). QDs were trapped at ∼10 mTorr, and the pressure was held constant for the duration of the experiment. After a few seconds, the emission spectrum for the as-trapped QDs was collected, as shown in Figure 1B. The vertical lines overlaying the spectra show the filter cutoff wavelengths, and the points centered between the cutoffs show the integrated intensity in each spectral region. As we have observed previously, the emission spectrum of the as-trapped QDs closely matches the spectrum for the same QDs in solution, but is much weaker (