Optical Phonon Behaviors of Photocharged Nanocrystals: Effects of

Subscriber access provided by Queen Mary, University of London

Spectroscopy and Photochemistry; General Theory

Optical Phonon Behaviors of Photocharged Nanocrystals: Effects of Free Charge Carriers Butian Zhang, Ruiheng Chang, Kexin Wang, Jing-Tao Lu, and Shun Wang J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.8b01831 • Publication Date (Web): 15 Aug 2018 Downloaded from http://pubs.acs.org on August 23, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

Optical Phonon Behaviors of Photocharged Nanocrystals: Effects of Free Charge Carriers †

†

†

§

Butian Zhang , Ruiheng Chang , Kexin Wang , Jing-Tao Lü , and Shun Wang †

*,†,§

MOE Key Laboratory of Fundamental Physical Quantities Measurement & Hubei

Key Laboratory of Gravitation and Quantum Physics, School of Physics, Huazhong University Science and Technology, Wuhan 430074, China §

School of Physics & Wuhan National High Magnetic Field Center, Huazhong

University Science and Technology, Wuhan 430074, China

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT

Page 2 of 27

For semiconductor nanocrystals (NCs), the precise knowledge of

phonons in presence of free carriers is important for understanding their electronic and photonic properties in device applications. Using Raman spectroscopy, this study investigates the effects of free charge carriers on optical phonon behaviors of NCs. The adoption of the photocharging method allows us to introduce free charge carriers into NCs without inducing other side effects. In the photocharged ZnO NCs, lower longitudinal optical (LO) phonon frequencies and weaker LO overtones relative to the fundamentals were found, which was explained by the screening and band-filling effects caused by the induced free carriers. The free carrier effects on optical phonon behaviors of NCs, usually neglected in previous studies, should be taken into consideration when discussing the electronic and photonic properties of NC-based devices. TOC GRAPHICS

KEYWORDS longitudinal optical phonons, photochemical doping, electron-phonon coupling, Raman spectroscopy, zinc oxide nanocrystals


Page 3 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


For nanocrystal (NC)-based devices, phonons can greatly affect the device performances such as thermal conductivities, charge carrier mobilities, and emission line widths.1-3 Therefore, the precise knowledge of NCs’ phonon behaviors under given conditions is important for understanding their electrical and optical properties in device applications. Depending on the specific requirements of the application, herein the “given conditions” refer to different material properties such as the size and composition, as well as varied external circumstances such as the electric and optical fields. Under such conditions, the phonon behaviors predicted for intrinsic bulk semiconductors without external fields may no longer be applicable. 4-7 In the fabrication of NCs and operation of NC-based devices, introduction of free carriers is common and sometimes unavoidable. Equilibrium free carriers may come from non-stoichiometry or aliovalent doping, while non-equilibrium free carriers can be generated under electric field or optical excitations.8-11 This highlights the importance of investigating the free carrier effects on phonon behaviors. For bulk semiconductors, the free carrier effects on phonon behaviors have been widely demonstrated in Si, Ge, GaAs from using Raman spectroscopy.12-14 Phonon frequencies

and

spectral

shapes

were

found

to

exhibit

material-

and

carrier-density-dependent changes when free carriers were introduced. Qualitative analysis revealed phonon-involved mechanisms such as the self-energy and plasmon-phonon coupling effect.15-17 Further quantitative studies established a relationship between free carrier densities and phonon frequencies, allowing Raman spectroscopy to be a non-destructive and facile technique to determine the free carrier densities in semiconductors.18 Similarly, we expect that the investigation of free carrier effects on phonon behaviors of NCs is beneficial for understanding their



Page 4 of 27

phonon-involved physical mechanisms, and for developing a non-destructive approach to characterize the free carrier concentration in NCs. Here, we select ZnO NCs as a model system and resonant Raman scattering (RRS) technique as a tool to study the free carrier effects on phonon behavior of NCs. Upon the introduction of free carriers, changes in longitudinal optical (LO) phonon frequencies and LO overtones to fundamental intensity ratios were observed, which is correlated with the screening and band-filling effects induced by the generated free carriers. A key to the successful observation of the free carrier effects is the adoption of a particular photocharging method for the accumulation of free carriers in NCs. The conventional methods of introducing free carriers, such as the aliovalent doping and photoexcitation, are usually accompanied with other complicated “side effects” in crystalline size, lattice strains, or local temperature.19-22 These effects can also affect the phonon behaviors, making it difficult to extract the influences from free carriers alone. Therefore, a reliable study of free carrier effects on phonon behaviors requires a strategy to accumulate free carriers without causing other effects on phonon behaviors. Photocharging method (Scheme S1) exactly fulfills such requirements. The irradiation of UV light on ZnO NCs dispersed in deaerated solvents excites electron-hole pairs, after which process the excited electrons stay in the conduction band, unable to recombine with the holes as the holes are rapidly scavenged by hole quenchers (e.g., ethanol molecules).23-25 The photocharged ZnO NCs will retain their charged state until they are discharged by electron quenchers such as oxygen, which can be simply achieved by exposure to air. The photocharging and discharging is a reversible process, and the main difference in photocharged and pristine ZnO NCs is the free carrier density while other effects should be minimal or negligible. Carrier densities ranging from 0.7×1020- 6×1020 cm-3 could be achieved in photocharged 4 ACS Paragon Plus Environment

Page 5 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ZnO NCs,23-26 corresponding to a few to a few tens of charges per nanocrystal for smaller ZnO QDs (size < 4 nm) and several tens to several hundred of charges per nanocrystal for larger ZnO NCs (size ~ 10 nm). Such carrier densities are comparable to or higher than those of the aliovalent doped and electrically doped ZnO crystals.27-28 A preliminary study by Yamamoto et al. demonstrated that the photocharging process would affect the Raman spectra of ZnO QDs.29 In this study, the photocharging method was improved to achieve a more homogeneous and repeatable photocharging of ZnO NCs. Furthermore, the selection of size-varying ZnO NCs, including QDs and larger NCs, allowed us to systematically investigate the effects of free charge carriers on the intensities of overtones relative to fundamentals. The free carrier effects on phonon behaviors were firstly examined in ZnO QDs (Figure 1). The four QD samples, characterized by Transmission Electron Microscopy (TEM) and absorption spectra, were found to be crystalline and well dispersed in ethanol with an average size of 2.8-3.9 nm (Figure 1a and S1). After verifying the successful photocharging of the sample by absorption and photoluminescence (PL) spectroscopy (Figure S2), we investigated the free carrier effects on phonon behaviors by comparing the RRS spectra of the pristine and photocharged ZnO QDs (Figure 1b). In both of the RRS spectra, a rising background coming from UV PL of ZnO QDs is visible, which is steeper for photocharged ZnO QDs due to an enhanced UV PL intensity upon photocharging. For pristine ZnO QDs, the Raman spectrum exhibit two peaks located at ~575 and ~1150 cm-1, which correspond to the first and second longitudinal optical (1LO and 2LO) phonon modes respectively. After photocharging, the 1LO peak intensity decreased while no 2LO peak was observed. Upon exposure to air, the changes in 1LO and 2LO peaks were reversed (Figure S3), confirming that these changes originated from the accumulated free carriers. 5 ACS Paragon Plus Environment


Page 6 of 27

Figure 1. Effects of free carriers on phonon behaviors of ZnO QDs dispersed in ethanol. (a) TEM and HRTEM (inset) image of ZnO QDs. (b) RRS spectra of pristine and photocharged ZnO QDs excited by a 325 nm laser. Linear background (inset, grey lines) was used for peak fitting in the 1LO region. (c) Two-peak Voigt fitting of Raman spectra in the 1LO region. The grey scatters, black curves, and olive/blue curves are the data, total fit and individual peak fits, respectively. Focusing on the spectral shape in the 1LO region, we observed that a shoulder peak appeared at a lower frequency side of the main 1LO peak (Figure 1b inset), and a two-peak Voigt fitting locates this shoulder peaks at 530-540 cm-1 (Figure 1c). This peak did not originate from the solvent or unwashed reactants (Figure S4) and was assigned as the surface optical (SO) phonon mode (Supporting Information).6, 30-31 After photocharging, the spectral shape change in the 1LO region mainly arises from two sources: a decreased intensity ratio of the 1LO and SO peaks (I1LO/ISO, Figure 2a), which will be addressed in a later discussion, and a red shift of 2.5-8.4 cm-1 in the 1LO peak (Figure 2b). Several mechanisms were proposed for the phonon frequency shift in presence of free carriers. The plasmon-phonon coupling effect may result in the shift of the LO frequency in doped semiconductors,32-33 however this is not applicable in our case because the coupling between plasmon and phonon are highly unlikely due to a large mismatch of their frequencies (~575 cm-1 and ~3000-5000 6 ACS Paragon Plus Environment

Page 7 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


cm-1).24 An intuitive explanation for the red shift is that the interatomic force constants are reduced due to the screening of the ion-ion interaction by free carriers.14 Apart from this, a bound phonon model was applied by Yu’s group specifically in the photoexcited QD system.34-35 That is, when the energy separation between the discrete electronic levels exceeds the LO phonon energy, the dielectric constant as a function of free carrier density will locally modify the LO frequency. A spectral broadening of 1LO peak were observed in these photoexcited QDs and in photocharged ZnO QDs film.29 This could be a result of inhomogeneous charging among QDs, which were caused by different incident power for individual QDs in photoexcited samples,34-35 and a varied number of hole quencher molecules for individual QDs in photocharged film samples.29 In our study, a narrowing, rather than a broadening of the 1LO peak, indicates more homogeneous charging among photocharged ZnO QDs dispersed in ethanol (Figure 2c). This spectral narrowing effect could be correlated with a longer phonon lifetime36 or a narrower distribution in frequencies of phonons that giving the 1LO line. The narrower distribution in frequencies of phonons is reasonable considering that the intensity change of 1LO peak after photocharging observed in our experiments is usually accompanied with the change of relative contributions from phonons with different wave vector directions (A1-LO and E1-LO mode) and values ( over the Brillouin zone), leading to different LO linewidth.37-38



Page 8 of 27

Figure 2. Characteristics of RRS peaks for pristine and photocharged ZnO QDs. (a) Intensity ratios of 1LO and SO peaks for pristine and photocharged ZnO QDs. (b) Peak positions of 1LO mode for pristine and photocharged ZnO QDs. (c) Spectral widths of 1LO mode for pristine and photocharged ZnO QDs. Now, let us revisit the changes in RRS peak intensity: the decreased I1LO/ISO and the disappeared 2LO peak. The RRS peak intensity for a phonon mode is usually associated with its coupling strengths to electrons in available excitonic states resonant to the laser frequency.5 Taking into consideration that multiple excitonic transitions may contribute to the RRS spectra, these observations could result from reduced overall EPC strength or a removal of the excitonic transitions with larger EPC strength by band filling. After photocharging, a filling of the lower-lying states was indicated by the bleaching of the excitonic peaks in the absorption spectra (Figure S2). This band filling effect may reduce the I1LO/ISO and 2LO intensity if the 1LO intensity is preferentially enhanced by the lower-lying excitonic transition.39-40 On the other hand, it is expected that the EPC strength will be reduced in presence of the free carriers due to the screening of the interactions between excitons and lattice vibrations. We attempted to find evidence from I2LO/I1LO, a value commonly used for evaluating the EPC strength. However, it is difficult to determine the I2LO/I1LO for photocharged 8 ACS Paragon Plus Environment

Page 9 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ZnO QDs, since the PL background swamps the 2LO peak (Figure 1b). To verify the effect of free charge carriers on EPC strength, we used larger ZnO NCs. The PL spectra of larger ZnO NCs shift to longer wavelength and are situated further away from the Raman signals, thus should induce a weakened background signal. The synthesis of larger ZnO NCs in ethanol resulted in particle aggregation. To obtain well-dispersed NCs, the ZnO QDs were transferred into dodecylamine (DDA) for continued particle growth. Four samples of ZnO NCs with average size of 8.5, 9.7, 9.9, and 11.5 nm and band gaps of 3.320-3.349 eV were obtained by varying the aging time (Figure 3a and S6). As expected, the PL background of the RRS spectra is weakened and the LO overtones (2LO and 3LO) can be clearly observed (Figure 3b), with their intensity decreased after photocharging (Figure 3c). For four QD samples, the 1LO peak exhibited a 2.1-4.2 cm-1 red shift and an average 2.3- and 3.3-fold red shifts were observed for 2LO and 3LO overtones (Figure 4a). Importantly, these photocharging induced changes were reversible by discharging (Figure S7), highlighting the effects of free carriers. Compared with ZnO QDs photocharged in neat ethanol, the red shift is smaller and linewidth is almost unchanged (Figure 4b), which are correlated with a lower free carrier density for these ZnO NCs photocharged in a mixture of toluene and ethanol.23-25 Another difference is that the bound phonon model applied in QDs is not applicable for these larger NCs in a weak confinement regime. For this case, we suggest the LO red shift to be a consequence of the screened ion-ion interaction.



Page 10 of 27

Figure 3. Effects of free carriers on phonon behaviors of ZnO NCs dispersed in toluene.

(a) TEM and HRTEM (inset) image of ZnO NCs. (b) RRS spectra of

pristine and photocharged ZnO NCs excited with a 325 nm laser. Reference spectrum of toluene was given. (c) 1LO, 2LO and 3LO peaks of the pristine and photocharged ZnO NCs. The background was removed and the intensity was normalized by the peak intensity of toluene.

Figure 4. Characteristics of LO peaks for pristine (black) and photocharged (red) ZnO NCs. (a) Peak positions of 1LO, 2LO, and 3LO peaks. (b) Spectral linewidths of 1LO, 2LO, and 3LO peaks. (c) Relative intensities of LO peaks, I2LO/I1LO and I3LO/I2LO. As noted above, we are interested in EPC strength before and after photocharging, which may be associated with the relative intensities of LO and LO overtones. For 10 ACS Paragon Plus Environment

Page 11 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


this purpose, I2LO/I1LO and I3LO/I2LO were calculated for pristine and photocharged ZnO NCs (Figure 4c). However, subsequent attempts at evaluating the EPC strength encountered a problem: could these relative intensities be directly correlated with the magnitude of EPC? In RRS studies, there are two mechanisms for understanding the origin of LO overtones: the virtual-state-mediated ‘true’ RRS process and the real-state-mediated cascade process (Figure 5a). In the ‘true’ RRS process, the nLO

phonons (n ≥ 1) are emitted in a one-step manner and the relative intensities of

various LO lines are associated with the EPC strength of the resonant electronic states. In the cascade process (i.e., hot luminescence process), the nLO phonons (n ≥ 2) are emitted in a multi-step manner during the relaxation of an electron to the conduction band bottom. The absolute intensity of 2LO and higher overtones are determined by τ ( )/τ ( ) , where τ ( ) and τ ( ) is respectively the lifetime of the radiative recombination and that of the phonon emission via a real transition.41 The τ ( ) and the absolute intensity are affected by the EPC strength, but the relative intensities of various LO lines are independent of EPC strength and only determined by electron and hole dispersion curves. However, I2LO/I1LO is an exception. Since the emission of 1LO phonons is forbidden by selection rules in cascade process,41-42 the I2LO/I1LO reflects the probability of cascade process to that of a physically different process giving the 1LO line (e.g., impurity-induced Raman scattering19). The distinction between ‘true’ Raman and cascade processes is a long-standing challenge43-44 and the RRS spectra were ambiguously interpreted in existing studies of ZnO nanostructures (Table 1). Some studies suggested the ‘true’ RRS process and correlated the I2LO/I1LO with the EPC strength,45-49 while others applied the cascade model in which the I2LO/I1LO has a completely different interpretation.19, 42, 50-55 For QDs, because the separation between their discrete energy 11 ACS Paragon Plus Environment


Page 12 of 27

levels exceeds the LO phonon energy, the cascade process is forbidden and thus the LO overtones can only originate from the ‘true’ RRS process. However, it is still unclear for ZnO NCs with larger size and less pronounced quantum confinement effect. For our ZnO NCs ranging from 8.5-11.5 nm, the energy separation is estimated to be 58-118 meV by two confinement models (Equation S1a-S1b), comparable with the LO phonon energy of 71 meV in ZnO (Figure 5b). Such a large energy separation will make the occurrence of cascade process rather doubtful, despite that we cannot completely exclude this possibility. What’s more, after photocharging, it is expected that the τ ( ) is significantly lengthened due to the screening effect at a carrier density of 1020 cm-3,56-57 which should increase the absolute intensities of 2LO and 3LO overtones originated from the cascade process. On the contrary, weaker overtones were found in the photocharged species, providing us an additional evidence that the cascade process is absent or negligible in our samples.


Page 13 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1. Applied physical models for interpreting the multiphonon RRS spectra of ZnO nanostructures in previous literatures and this work.

Sample

Crystalline size

I2LO/I1LO

Applied physical model

Ref.

ZnO

3-5 nm

-

‘True’ RRS

49

ZnO

3.5-12 nm

0.4-0.5

‘True’ RRS

45

Mn-doped ZnO

12.5 nm

0.4

Cascade

55

ZnO

20 nm

0.8

Cascade

42

ZnO

9-20 nm

0.38-1.07 ‘True’ RRS

46

ZnO

24 nm

2.94

Cascade

19

O-deficient ZnO

Optical Phonon Behaviors of Photocharged Nanocrystals: Effects of

Recommend Documents