Charge hopping revealed by jitter correlations in the photoluminescence spectra of single CdSe nanocrystals

Mark J. Fernée, Brad Littleton, Taras Plakhotnik, Halina Rubinsztein-Dunlop, Daniel E. Gómez, and Paul Mulvaney

Phys. Rev. B 81, 155307 – Published 8 April 2010

Abstract

Spectral fluctuations observed in single CdSe/ZnS nanocrystals at 5 K are found to be entirely the result of discrete charge hops in the local environment of the nanocrystal, which occur at a rate comparable to the acquisition time of a single spectrum. We show that intervals between discrete spectral hops introduce a correlation between the successive measurements of the emission wavelength of single CdSe nanocrystals. This correlation can be recovered even in the presence of noise, but is shown to be sensitive to the experimental acquisition time, in good agreement with theory. However, we only find correlations for the smaller of the two nanocrystal sizes studied and discuss this in terms of size-dependent time scales correlated with the amount of excess energy dissipated in the nanocrystal due to hot-carrier relaxation.

Received 2 February 2010

DOI:https://doi.org/10.1103/PhysRevB.81.155307

Authors & Affiliations

Mark J. Fernée^*, Brad Littleton^†, Taras Plakhotnik, and Halina Rubinsztein-Dunlop

Centre for Quantum Computer Technology, School of Physical Sciences, The University of Queensland, Queensland 4072, Australia

Daniel E. Gómez^‡ and Paul Mulvaney

School of Chemistry, The University of Melbourne, Parkville, Victoria 3010, Australia

^*FAX: +61 7 3365 1242; fernee@physics.uq.edu.au
^†Present address: Dept. of Physics, King’s College London, the Strand, London WC2R 2LS, United Kingdom.
^‡Present address: Materials Science and Engineering, CSIRO, Private Bag 33, Clayton South MDC, Victoria 3169, Australia.

Article Text (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand

Issue

Vol. 81, Iss. 15 — 15 April 2010

Reuse & Permissions

Access Options

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
(Color online) (a) A series of 512 successive spectra obtained from a single NC of $R \sim 2.5 nm$ at 5 K using a pump irradiance of $27 W / {cm}^{- 2}$ and a 7 s integration time. (b) An expanded region of greater stability illustrating the effect of SD on individual spectra. (c) A histogram of peak locations for the data in (a). (d) Average spectrum obtained by aligning each spectrum to its peak position (peak FWHM= $295 μ eV$ ).Reuse & Permissions
Figure 2
(Color online) (a) Histogram of spectral jump magnitudes (blue circles—bin width $30 μ eV$ ) obtained from a series of 3163 individual 1 s spectra from a single $R \sim 2.5 nm$ core diameter NC with a pump irradiance of $66 W / {cm}^{2}$ . (b) Histogram of spectral jump magnitudes (red squares—bin width $30 μ eV$ ) obtained from a series of 3552 individual 1 s spectra from a single $R \sim 4 nm$ core diameter NC with a pump irradiance of $52 W / {cm}^{2}$ . The dashed lines represent Lorentzian fits to the data. Panels (c)–(f) show conditional probability $P (| ν_{n} | \leq Δ ∣ | ν_{n - l} | \leq Δ)$ as a function of the jump separation, $l$ , with $Δ = 60 μ eV$ . (c) Single $R \sim 2.5 nm$ core diameter NC shown in (a) (blue open circles). (d) Single $R \sim 4 nm$ core diameter NC shown in (b) (red open squares). (e) Ensemble of 20 single $R \sim 2.5 nm$ NCs comprising 2945 jumps. (f) Ensemble of 21 single $R \sim 4 nm$ NCs comprising 2339 jumps. $60 μ eV$ bin widths are used throughout.Reuse & Permissions
Figure 3
(Color online) Spectral jump time series are generated from calculating the difference between two successive spectral peak measurements, $ν_{n} = p_{n + 1} - p_{n}$ . Correlations between successive spectral jumps in spectral jump time series can arise from the presence of a common spectrum (asterisk) in the case when calculating the joint probability for observing two intervals with no jumps (i.e., $| ν | = 0$ ). Only three spectra are required when calculating the joint probability for successive jumps to have zero magnitude (a), while four spectra are required in all other cases (b).Reuse & Permissions
Figure 4
(Color online) Variation in the correlation, $C$ , as a function of the average number of spectral jumps, $N$ .Reuse & Permissions
Figure 5
(Color online) Variation in the correlation, $C$ as a function of the bin width, $Δ$ for the data of Fig. 2c (blue solid line). The same analysis applied to a $10^{5}$ point simulated SD data set generated using Lorentzian distribution of random jumps with a FWHM of $200 μ eV$ with an average jump rate $N = 1$ and a Gaussian peak measurement error with $σ_{M} = 20 μ eV$ (red dashed).Reuse & Permissions
Figure 6
(a) Jump distributions for the data in Fig. 2a with $15 μ eV$ bin widths, using 1 s (open circles) and 2 s (open diamonds) integration times. (b) Conditional probabilities $P (| ν_{n} | \leq Δ ∣ | ν_{n - l} | \leq Δ)$ , $Δ = 60 μ eV$ .Reuse & Permissions

Physical Review B

covering condensed matter and materials physics