View Article Online View Journal
PCCP Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: M. J. Fernée, C. Sinito, P. Mulvaney, P. Tamarat and B. Lounis, Phys. Chem. Chem. Phys., 2014, DOI: 10.1039/C4CP02022G.
This is an Accepted Manuscript, which has been through the Royal Society of Chemistry peer review process and has been accepted for publication. Accepted Manuscripts are published online shortly after acceptance, before technical editing, formatting and proof reading. Using this free service, authors can make their results available to the community, in citable form, before we publish the edited article. We will replace this Accepted Manuscript with the edited and formatted Advance Article as soon as it is available. You can find more information about Accepted Manuscripts in the Information for Authors. Please note that technical editing may introduce minor changes to the text and/or graphics, which may alter content. The journal’s standard Terms & Conditions and the Ethical guidelines still apply. In no event shall the Royal Society of Chemistry be held responsible for any errors or omissions in this Accepted Manuscript or any consequences arising from the use of any information it contains.
www.rsc.org/pccp
Physical Chemistry Chemical Physics
View Article Online The optical phonon spectrum in CdSe colloidal quantum dots. DOI: 10.1039/C4CP02022G
Mark J. Fernée1,2, Chiara Sinito1,2, Paul Mulvaney3, Philippe Tamarat1,2, Brahim Lounis1,2 *.
1Univ Bordeaux, LP2N, F-‐33405 Talence, France. 2Institut d’Optique & CNRS, LP2N, F-‐33405 Talence, France.
Published on 03 July 2014. Downloaded by Kansas State University on 08/07/2014 17:19:04.
3School of Chemistry, The University of Melbourne, Parkville, Victoria, 3010,
Australia. Absract: The direct coupling of excited electronic states to optical phonons in single CdSe colloidal quantum dots is explored using both photoluminescence emission and excitation spectroscopies. We find a broad optical phonon spectrum associated with a single fine structure state. Multiple peaks in the optical phonon sideband are ascribed to different optical phonon types emanating from both the core and shell layers. A mixed emission process that involves the simultaneous generation of two different types of optical phonon is also observed. In general, rather than a single mode, each designated phonon type is associated with a dispersed family of modes. Narrow optical phonon sidebands, consistent with the dominant LO mode are observed in some nanocrystals. A linewidth analysis indicates that optical phonon lifetimes are in the 10 picosecond range. We demonstrate the ability to selectively excite a specific band-‐edge state by directly exciting it’s LO phonon sideband. Introduction: Optical phonons are responsible for the relaxation of highly excited carriers in bulk semiconductors. However their role in carrier relaxation in quantum confined semiconductors, or quantum dots, is less well known. A key factor that distinguishes quantum dots from their bulk counterpart is that discrete electronic levels in quantum dots replace the bulk band-‐structure, hence they are often referred to as artificial atoms. In this regime, it was proposed that if the separation between the energy levels didn’t match the optical phonon energy, energy relaxation would be inhibited due to a phonon bottleneck1, 2. The search for a phonon bottleneck has been widely conducted with colloidal nanocrystals (NCs) due to the large degree of quantum confinement and hence wide energy level separations that can be achieved3-‐5. It was found that there were sufficient additional relaxation pathways6, 7 in these materials to preclude the observation of a phonon bottleneck4. In fact, it required significant NC engineering to finally observe this effect8, which highlighted the many possible relaxation pathways in these materials. Therefore it is not clear what role optical phonons play in the relaxation pathway following excitation to energies far above the band-‐edge. Optical phonon coupling to NCs has been studied using a variety of different techniques both in the time domain and frequency domain. The time domain techniques encompass coherent photon echo9, 10 and transient grating11
1
Physical Chemistry Chemical Physics Accepted Manuscript
Page 1 of 9
Page 2 of 9
View Article Online 12 techniqies as well as state-‐resolved differential absorption pump-‐probe DOI: 10.1039/C4CP02022G techniques. In general these techniques resolve a single LO phonon mode from an ensemble of NCs. Frequency domain techniques include photoluminescence13-‐ 17 and Raman spectroscopies18-‐24. Of all the techniques used to study the optical phonon coupling, only the Raman spectra indicate that the optical phonon region of CdSe core/shell NCs contains other optical phonon modes, even though the spectra are significantly broadened. Recent studies of CdSe/CdS core shell NCs reveal a rich optical phonon spectrum with modes associated with the dominant longitudinal optical (LO) phonon of both the CdSe core and CdS shell as well as so-‐called surface optical (SO) and CdSe/CdS interface (IF) modes21. Interestingly, another recent study that included detailed atomistic modeling of the Raman response suggests that the so called SO modes do not actually involve the surface of the NC and are still predominantly associated with the CdSe core22. This same study also predicted a dispersed phonon mode spectrum at the single NC level that combined to give the ensemble Raman response. This mode dispersion is likely to be a fundamental consequence of the nano-‐scale material and deserves further investigation. In general, techniques that study NC ensembles mask the effects of sample inhomogeneity. Thus single NC techniques can provide complementary information that is not available to ensemble techniques. The optical phonon replica has been detected with single NC spectroscopy14-‐17. Individual spectra exhibit a range of apparent weights of the LO phonon replica, indicating an inhomogeneity in the exciton-‐phonon coupling strength between different NCs14. However, in general single NC photoluminescence (PL) spectroscopy of the LO phonon replica is complicated by low signal to noise ratios as well as spectral diffusion, both of which have precluded high resolution spectroscopy of the phonon replica. In this letter we study the optical phonon spectrum in single NCs. Low resolution measurements are conducted using PL spectroscopy. Increased resolution is obtained using resonant photoluminescence excitation (RPLE) spectroscopy of the optical phonon sideband. Selective excitation of the observed sideband features are used to identify the optical phonon spectrum associated with a single band-‐edge exciton state. Our experimental setup has been described elsewhere26, 27. Briefly, we use commercial NCs with a wurtzite crystal structure emitting at 655 nm embedded in a polyvinyl alcohol matrix or a CdSe/4CdS/1ZnS core/shell/shell NCs with a zincblende crystal structure embedded in a polymethylmethacrylate substrate and deposited on to clean glass coverslips. The sample is mounted in a liquid helium bath cryostat where the low-‐pressure helium exchange gas medium can maintain a temperature of 2 Kelvin. The experiment uses an epifluorescence microsope geometry with a high numerical aperture microscope objective contained inside the cryostat and operated in a scanning confocal mode so that individual NCs can be isolated and studied. The resonant PL excitation scans are conducted using circularly polarized excitation from a cw dye laser operated in multimode configuration with a ~10
2
Physical Chemistry Chemical Physics Accepted Manuscript
Published on 03 July 2014. Downloaded by Kansas State University on 08/07/2014 17:19:04.
Physical Chemistry Chemical Physics
Physical Chemistry Chemical Physics
View Article Online GHz mode envelope, which sets the resolution of the technique. The tuning f the DOI:o10.1039/C4CP02022G laser is conducted using a computer-‐controlled stepper motor driving an intra-‐ cavity lyot filter using a ~16 GHz step size. The output power of the dye laser is actively stabilized using a noise eater, resulting in a constant power at the sample over the entire scan range. The red-‐shifted emission using a low pass filter with OD 7 rejection of the laser scatter resulting in background-‐free detection for the low excitation power of 200 nW used in this experiment. RPLE of the band edge exciton is conducted by excitation of the NCs zero phonon lines with a tunable cw laser and the detection of red shifted luminescence emitted at the optical phonon replica. The phonon replica is at a fixed energy transition (~26meV) relative to the zero phonon lines. For the optical phonon absorption region, the RPLE signal is collected via the integrated ZPL PL, which is >20 meV red-‐shifted from the optical phonon sideband absprption. The low pass filter transition region occurs over ~10 meV and the filter cutoff wavelength is adjusted to each NC by changing the incidence angle of the filter and so is suitable for both RPLE variants described above. For this study, NCs with multiple epitaxial shells are necessary in order to obtain bright and spectrally stable emission at cryogenic temperatures. In particular we have found that an outer ZnS layer is necessary for both bright emission and to inhibit excessive dynamics involving photo-‐generated charges and surface states28, which can result in low photoluminescence efficiency and spectral stability. In Fig. 1 we show the photoluminescence (PL) spectrum obtained from a single NC at 2 K. In this case, there is a single zero phonon line (ZPL), characteristic of emission from the trion state28-‐30. Importantly, we find that both the optical phonon energy and coupling to the excited state exhibit no noticeable change between the neutral17 and trion states of the NC. Although we note that the trion created by photocharging has an external counter-‐charge and so the overall NC charge remains unchanged. The single ZPL of the trion state enables a detailed examination of the optical phonon side bands associated with a single spectral line using an integration time of 14 minutes, enabling the detection of both the first and second phonon replicas. We plot the spectrum intensity with both linear and logarithmic intensity axes. The logarithmic data clearly reveals the structure in the weak phonon bands from which we identify both the contributions from the CdSe core optical phonons as well as those from the CdS shell. In addition to the features generally attributed to the LO phonon is the appearance of a clear shoulder at energies above the CdSe LO peak, corresponding to emission of phonons with lower energy than that of the LO phonon. This shoulder is also clearly repeated in the second replica. The appearance of such a feature is consistent with the interaction with SO and interface phonons that have been observed in Raman spectroscopy of CdSe NCs21. We also note that the second replica is a compound spectrum of a (1+1) type, where one of the optical phonons is the CdSe LO phonon. Thus the second replica is approximately a factor of 0.1 smaller than the previous replica but is otherwise reproduced over the same energy scale.
3
Physical Chemistry Chemical Physics Accepted Manuscript
Published on 03 July 2014. Downloaded by Kansas State University on 08/07/2014 17:19:04.
Page 3 of 9
Page 4 of 9
View Article Online
Published on 03 July 2014. Downloaded by Kansas State University on 08/07/2014 17:19:04.
DOI: 10.1039/C4CP02022G
Figure 1. PL spectrum of a single charged NC at 2 K obtained with an excitation irradiance of 25 Wcm-‐2 and an integration time of 14 minutes. The lower spectrum is plotted with a logarithmic vertical axis to enhance the optical phonon replicas.
The use of PL spectroscopy to study the optical phonon interaction with single NCs is not straightforward, as it requires long integration times and NCs exhibiting exceptional spectral stability. For example the data in Fig. 1 were obtained with the lowest spectral resolution to enhance the signal to noise ratio. In order to simultaneously increase the signal to noise ratio as well as the spectral resolution we use RPLE, which enables the study of optical phonon sideband absorption31 that exists on the high energy side of the ZPL, shifted by one optical phonon energy. Optical phonon sideband absorption corresponds to the absorption of a photon, which simultaneously creates an exciton state as well as an optical phonon. In theory, the absorption sideband should be the mirror image of the red-‐shifted optical phonon sideband observed in emission in the adiabatic limit. However, it has been shown that optical phonon interaction is actually non-‐adiabatic31, which may affect the naïve symmetry expectation between the optical phonon sideband observed in emission and absorption. The optical phonon absorption band can be extremely important for quantum technologies, as specific phonon sidebands can enable off-‐resonant excitation of a specific exciton state32-‐34. In Fig. 2 we use two RPLE scans to reveal the spectral region close to the band-‐ edge. The low energy RPLE scan obtains its signal from the red-‐shifted optical phonon band (as shown in Fig. 1) in order to reveal the band-‐edge detail fine structure states. Here we have identified spectral lines corresponding to the optically allowed transitions (bright exciton states). The second part of the spectrum consists of an RPLE scan that obtains its signal from the ZPL emission in order to enhance the sensitivity to the optical phonon sideband. Here we see what appear to be two regions associated with optical phonons in the CdSe core35-‐37 and CdS shell regions. Arrows indicate the LO phonon energies relative to the 0U state, which has a large oscillator strength. Of particular interest is the
4
Physical Chemistry Chemical Physics Accepted Manuscript
Physical Chemistry Chemical Physics
Physical Chemistry Chemical Physics
Published on 03 July 2014. Downloaded by Kansas State University on 08/07/2014 17:19:04.
View Article Online appearance of a narrow peak at the LO phonon energy as well as other structure DOI: 10.1039/C4CP02022G at lower energies which we will discuss in more detail below.
Figure 2. RPLE spectra of a single NC where the band-‐edge fine structure signal is obtained by monitoring the LO phonon replica, while the signal from higher energy optical phonon sideband scan is derived from the band-‐edge PL. LO phonon energies for both the CdSe core and CdS shell are indicated.
The RPLE signal around the LO phonon energy associated with the CdSe core, is expanded in Fig. 3a,b. In Fig. 3a, we show the optical phonon sideband region obtained from four different NCs. Each of these NCs has a rod-‐like band-‐edge spectrum with a dominant 0U line observed in RPLE of the band edge. Thus the optical phonon sideband should be dominated by modes coupled to the 0U state. This then explains the emergence of a similar three-‐peaked structure observed across these NCs. We tentatively associate the three peaks to different optical phonon modes, nominally associated with the SO branch, and interface mode (IF) associated with the CdSe/CdS core shell interface and an LO phonon mode21. However we see in Fig. 3b that these modes actually have significant structure associated with each peak. The series of four RPLE scans of the same NC shown in Fig. 3b is used to show the repeatability of the structure between scans. Multiple scans of the optical phonon sideband are necessary as for such broad features, the RPLE scan is more susceptible to spectral diffusion, which is responsible for the differences between scans. Addition scans (not shown) show the peaks are very sensitive to spectral diffusion, which tends to blur or remove the structure, thus ensuring that they are not experimental artefacts. The observed richly structured mode families have been predicted using atomistic calculations for small CdSe NCs22. One might expect an enhancement of this effect in core/shell structures due to interfacial strain introducing variations in the crystal lattice. We also notice that in some NCs there appears a dominant sharp LO mode, although shifted to higher energy than the LO phonon associated with bulk CdSe. Such a blue shift is expected from NCs as the CdS shell thickness increases21.
5
Physical Chemistry Chemical Physics Accepted Manuscript
Page 5 of 9
Page 6 of 9
View Article Online
Published on 03 July 2014. Downloaded by Kansas State University on 08/07/2014 17:19:04.
DOI: 10.1039/C4CP02022G
Figure 3. RPLE scans of the CdSe optical phonon sideband. (a) RPLE scans from four different NCs indication the presence of thee dominant phonon modes. (b) Successive RPLE scans from a single NC revealing underlying structure in each of the modes.
We note that the different optical phonon modes are relatively sharp, of order 100 µeV full width at half maximum (fwhm), indicative of optical phonon lifetimes of approximately 10 picoseconds. This lifetime is comparable to the 5 ps decoherence lifetime found in pure CdSe NCs using a pump/probe technique12 and is also compatible with the slowest intraband relaxation rates reported for CdSe NCs38 and similar to lifetimes observed in other materials39. We test the association of the optical phonon spectrum to the different band-‐ edge states by directly exciting different energies along the optical phonon region and recording the band-‐edge exciton PL spectrum. In Fig 4a we show RPLE scans of the band-‐edge states and the optical phonon spectrum with different excitation points are indicated. This NC exhibits zero field splitting of the 1L state that has been associated with NC anisotropy that breaks the cylindrical symmetry42, 43. The resultant PL observed with this optical phonon sideband excitation are shown in Fig. 4b and compared to the standard PL spectrum obtained by exciting far above the band-‐edge. We first excite at precisely 1 LO phonon energy from the highest 0U state (P1 excitation) and see that compared with the non-‐resonant excitation PL case, emission from the 0 U state is strongly enhanced, indicating the LO phonon peak is associated with this state33, 34, 44-‐46. This spectrum remains unchanged as we lower the excitation energy to the position, P2, indicating that the entire optical phonon spectrum is associated with the 0U state, which has the largest oscillator strength of the band-‐edge states in this rod-‐like region. Finally at the lowest energy position, P3, we find that the lower energy states are predominantly excited, marking the edge of the optical phonon region associated with the 0U state. Whilst clearly demarking the optical phonon spectrum associated with a single state, our results also show that it is possible to use optical phonon sidebands to selectively excite different band-‐edge states.
6
Physical Chemistry Chemical Physics Accepted Manuscript
Physical Chemistry Chemical Physics
Physical Chemistry Chemical Physics
View Article Online
Published on 03 July 2014. Downloaded by Kansas State University on 08/07/2014 17:19:04.
DOI: 10.1039/C4CP02022G
Figure 4. Direct excitation of the optical phonon sideband at 2 K. (a) An RPLE scans of a single NC revealing the band-‐edge and optical phonon sideband structures. The LO phonon energy relative to the 0U state is indicated. The labeled points represent the different excitation energies employed. (b) The band-‐edge emission spectra obtained using standard non-‐resonant pumping (PL) and those obtained by exciting the various labeled optical phonon sideband modes (P1-‐P3).
In conclusion, we have probed the coupling of the excited state of a NC to optical phonons at the single NC level, revealing a complex optical phonon structure that have previously been assigned as LO, SO and IF modes. The apparent dispersion in the mode energies could have detrimental implications for energy dissipation models that assume a single LO phonon frequency. In fact the relatively long optical phonon lifetimes indicate that single NCs are relatively high-‐Q phonon resonators and that optical phonons are only weakly coupled to the local environment. Finally we have demonstrated for the first time the possibility to use optical phonon sideband excitation to directly excite specific band-‐edge fine structure states. Such new excitation schemes have the potential for selective quantum state preparation, which can be used to tailor the emission properties single NCs. Precise control of the band-‐edge state populations can have important implications for novel quantum applications33, 34, 44-‐46 and advanced light sources. References: 1. H. Benisty, C. M. Sotomayor-‐Torres and C. Weisbuch, Physical Review B, 1991, 44, 10945–10948. 2. U. Bockelmann and G. Bastard, Physical Review B, 1990, 42, 8947–8951. 3. S. Xu, A. A. Mikhailovsky, J. A. Hollingsworth and V. I. Klimov, Physical Review B, 2002, 65, 045319. 4. R. R. Cooney, S. L. Sewall, K. E. H. Anderson, E. A. Dias and P. Kambhampati, Physical Review Letters, 2007, 98, 177403.
7
Physical Chemistry Chemical Physics Accepted Manuscript
Page 7 of 9
5. 6. 7. 8. 9. 10. Published on 03 July 2014. Downloaded by Kansas State University on 08/07/2014 17:19:04.
11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.
Page 8 of 9
View Article Online B. L. Wehrenberg, C. Wang and P. Guyot-‐Sionnest, The Journal of Physical DOI: 10.1039/C4CP02022G Chemistry B, 2002, 106, 10634-‐10640. P. Kambhampati, Accounts of Chemical Research, 2010, 44, 1-‐13. P. Kambhampati, The Journal of Physical Chemistry C, 2011, 115, 22089-‐ 22109. A. Pandey and P. Guyot-‐Sionnest, Science, 2008, 322, 929-‐932. R. W. Schoenlein, D. M. Mittleman, J. J. Shiang, A. P. Alivisatos and C. V. Shank, Physical Review Letters, 1993, 70, 1014-‐1017. M. R. Salvador, M. W. Graham and G. D. Scholes, The Journal of Chemical Physics, 2006, 125, 184709. V. M. Huxter, A. Lee, S. S. Lo and G. D. Scholes, Nano Letters, 2009, 9, 405– 409. D. M. Sagar, R. R. Cooney, S. L. Sewall, E. A. Dias, M. M. Barsan, I. S. Butler and P. Kambhampati, Physical Review B, 2008, 77, 235321. M. Nirmal, C. B. Murray and M. G. Bawendi, Physical Review B, 1994, 50, 2293-‐2300. S. A. Empedocles, D. J. Norris and M. G. Bawendi, Physical Review Letters, 1996, 77, 3873-‐3876. N. Le Thomas, E. Herz, O. Schops, U. Woggon and M. V. Artemyev, Phys Rev Lett, 2005, 94, 016803. M. J. Fernee, B. N. Littleton, S. Cooper, H. Rubinsztein-‐Dunlop, D. E. Gomez and P. Mulvaney, Journal of Physical Chemistry C, 2008, 112, 1878-‐1884. L. Biadala, Y. Louyer, P. Tamarat and B. Lounis, Physical Review Letters, 2009, 103, 037404. A. V. Baranov, Y. P. Rakovich, J. F. Donegan, T. S. Prenova, R. A. Moore, D. V. Talapin, A. L. Rogach, Y. Masumoto and I. Nabiev, Physical Review B, 2003, 68, 165306. V. M. Dzhagan, M. Y. Valakh, A. E. Raevskaya, A. L. Stroyuk, S. Y. Kuchmiy and D. R. T. Zahn, Nanotechnology, 2007, 18, 285701. L. Liu, X. Xiao-‐Liang, L. Wen-‐Tao and L. Hai-‐Fei, Journal of Physics: Condensed Matter, 2007, 19, 406221. N. Tschirner, H. Lange, A. Schliwa, A. Biermann, C. Thomsen, K. Lambert, R. Gomes and Z. Hens, Chemistry of Materials, 2011, 24, 311-‐318. C. Lin, D. F. Kelley, M. Rico and A. M. Kelley, ACS nano, 2014, 8, 3928-‐3938. H. Lange, M. Artemyev, U. Woggon, T. Niermann and C. Thomsen, Physical Review B, 2008, 77. S. V. Kershaw, A. S. Susha and A. L. Rogach, Chemical Society Reviews, 2013, 42, 3033-‐3087. D. M. Sagar, R. R. Cooney, S. L. Sewall and P. Kambhampati, The Journal of Physical Chemistry C, 2008, 112, 9124-‐9127. M. J. Fernée, P. Tamarat and B. Lounis, Journal of Physical Chemistry Letters, 2013, 4, 609-‐618. Y. Louyer, L. Biadala, J. B. Trebbia, M. J. Fernee, P. Tamarat and B. Lounis, Nano Letters, 2011, 11, 4370-‐4375. M. J. Fernée, C. Sinito, Y. Louyer, C. Potzner, T.-‐L. Nguyen, P. Mulvaney, P. Tamarat and B. Lounis, Nature Communications, 2012, 3, 1287. M. J. Fernée, B. N. Littleton and H. Rubinsztein-‐Dunlop, ACS Nano, 2009, 3, 3762-‐3768.
8
Physical Chemistry Chemical Physics Accepted Manuscript
Physical Chemistry Chemical Physics
Physical Chemistry Chemical Physics
30. 31. 32. 33.
Published on 03 July 2014. Downloaded by Kansas State University on 08/07/2014 17:19:04.
34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46.
View Article Online Y. Louyer, L. Biadala, P. Tamarat and B. Lounis, Applied Physics Letters, DOI: 10.1039/C4CP02022G 2010, 96, 203111. V. M. Fomin, V. N. Gladilin, J. T. Devreese, E. P. Pokatilov, S. N. Balaban and S. N. Klimin, Physical Review B, 1998, 57, 2415-‐2425. B. D. Gerardot, D. Brunner, P. A. Dalgarno, P. Ohberg, S. Seidl, M. Kroner, K. Karrai, N. G. Stoltz, P. M. Petroff and R. J. Warburton, Nature, 2008, 451, 441-‐444. I. A. Akimov, D. H. Feng and F. Henneberger, Physical Review Letters, 2006, 97, 056602. M. A. Pooley, D. J. P. Ellis, R. B. Patel, A. J. Bennett, K. H. A. Chan, I. Farrer, D. A. Ritchie and A. J. Shields, Applied Physics Letters, 2012, 100, 211103. H. Htoon, P. J. Cox and V. I. Klimov, Physical Review Letters, 2004, 93. A. Hundt, T. Flissikowski, M. Lowisch, M. Rabe and F. Henneberger, physica status solidi (b), 2001, 224, 159–163. F. Gindele, K. Hild, W. Langbein and U. Woggon, Physical Review B, 1999, 60, R2157-‐R2160. P. Guyot-‐Sionnest, B. Wehrenberg and D. Yu, The Journal of Chemical Physics, 2005, 123, 074709. X.-‐Q. Li and Y. Arakawa, Physical Review B, 1998, 57, 12285-‐12290. V. I. Klimov, The Journal of Physical Chemistry B, 2000, 104, 6112-‐6123. S. L. Sewall, R. R. Cooney, K. E. H. Anderson, E. A. Dias and P. Kambhampati, Physical Review B, 2006, 74, 235328. S. V. Goupalov, Physical Review B, 2006, 74, 113305. S. V. Goupalov, Physical Review B, 2009, 79, 233301. K. Kowalik, O. Krebs, A. Lemaitre, J. A. Gaj and P. Voisin, Physical Review B, 2008, 77, 161305. H. Kumano, H. Kobayashi, S. Ekuni, Y. Hayashi, M. Jo, H. Sasakura, S. Adachi, S. Muto and I. Suemune, Physical Review B, 2008, 78, 081306. T. Flissikowski, A. Hundt, M. Lowisch, M. Rabe and F. Henneberger, Physical Review Letters, 2001, 86, 3172-‐3175.
9
Physical Chemistry Chemical Physics Accepted Manuscript
Page 9 of 9