Nanocavity-Enhanced Giant Stimulated Raman ... - ACS Publications

Subscriber access provided by Iowa State University | Library

Communication

Nanocavity Enhanced Giant Stimulated Raman scattering in Si nanowires in the visible light region Daksh Agarwal, Mingliang Ren, Jacob S Berger, Jinkyoung Yoo, Anlian Pan, and Ritesh Agarwal Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.8b04666 • Publication Date (Web): 25 Jan 2019 Downloaded from http://pubs.acs.org on January 26, 2019

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 22 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

Nano Letters

Nanocavity Enhanced Giant Stimulated Raman scattering in Si nanowires in the visible light region Daksh Agarwal1, MingLiang Ren1, Jacob S. Berger1, Jinkyoung Yoo2, Anlian Pan3, and Ritesh Agarwal1* 1 Department

of Materials Science and Engineering, University of Pennsylvania, Philadelphia, PA

19104, USA 2

Center for Integrated Nanotechnologies, Los Alamos National Laboratory, Los Alamos, NM

87545, USA 3

Key Laboratory for Micro-Nano Physics and Technology of Hunan Province, College of

Materials Science and Engineering, Hunan University 410082, China * E-mail: [email protected]

Silicon photonics has been a very active area of research especially in the past two decades in order to meet the ever-increasing demand for more computational power and faster device speeds and their natural compatibility with CMOS. In order to develop Si as a useful photonics material, essential photonic components such as light sources, waveguides, wavelength convertors, modulators and detectors need to be developed and integrated. However, due to the indirect electronic bandgap of Si, conventional light emission devices such as LEDs and lasers cannot be built. Therefore, there has been considerable interest in developing Si-based Raman lasers, which are nonlinear devices, and require large stimulated

1 ACS Paragon Plus Environment

Nano 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

Page 2 of 22

Raman scattering (SRS) in an optical cavity. However, due to the low quantum yield of SRS in Si, Raman lasers have very large device footprints and high lasing threshold, making them unsuitable for faster, smaller and energy efficient devices. Here, we report strong SRS and extremely high Raman gain in Si nanowire optical cavities in the visible region, with measured SRS threshold as low as 30 kW/cm2. At cavity mode resonance, light is confined into a low mode volume and high intensity electromagnetic mode inside the Si nanowire due to its high refractive index, which leads to strong SRS at low pump intensities. Electromagnetic calculations reveal greater than six orders of magnitude increase in Raman gain coefficient at 532 nm pump wavelength, compared to the gain value at 1.55 µm wavelength reported in literature, despite the 108 higher losses at 532 nm. Because of the high gain in such small structures, we believe that this is a significant first step in realizing a monolithically integrable nanoscale low powered Si Raman laser.

Keywords: Raman spectroscopy, silicon, nanowire, nanocavity, stimulated Raman scattering, Silicon photonics, Raman lasing

Silicon photonics offers a cost effective solution for building the next generation of ultrafast, small and energy efficient optoelectronic devices1, because of existing and highly developed processes in the microelectronic industry along with compatibility with CMOS fabrication.2 These optoelectronic devices also require a laser for transmitting the modulated electrical data using light2, and ideally an optical source fabricated from Si for seamless integration with other Si-based components. But due to the indirect electronic bandgap of Si3, building 2 ACS Paragon Plus Environment

Page 3 of 22 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

Nano Letters

conventional emission based lasers in which an electron-hole pair recombination leads to a photon emission, is extremely difficult and as a result Raman scattering has been explored as a source for stimulated emission from Si3-5. But relatively high losses in Si6 and an inherently low Raman gain coefficient value of ~10 cm/GW (at 1.55 µm)7 have made it difficult to achieve net optical gain in Si cavities. To overcome this challenge, a large cavity of size 4.8 cm, which wasn’t monolithically integrable, was used to achieve net gain using pulsed optical pumping8 as well as continuous wave (CW) optical pumping9 but with a power requirement of ~470 mW (power intensity ~ 30MW/cm2) to attain a net gain of 58%. Si Raman lasers have been fabricated using similar cavities with lasing thresholds of 9 W (pulsed pumping)10 and 0.4 mW (CW pumping)11, which are in general prohibitive for practical applications. Monolithically integrable ring-resonator cavities optimized for low losses decreased lasing threshold to ~50 mW CW power but still required a reverse bias p-n junction for efficient operation to minimize free carrier absorption losses12. These centimeterscale lasers with large lasing thresholds are incompatible with the goal of making small, low powered energy efficient devices which require nanoscale, low lasing threshold and monolithically integrable Si laser. These nanoscale devices are particularly important to achieve seamless integration with electronic devices such as transistors in sub-100 nm length scales. Although a micrometer scale Si Raman laser has been demonstrated using photonic crystals13, it still has an overall large device footprint and is also not monolithically integrable with other devices. Si nanostructures provide an interesting platform to investigate optical properties which are otherwise challenging to realize in bulk. For example, we recently reported broadband white light emission from Si nanowires coupled to plasmonic cavities14,15 which is not observable in bulk Si because of its indirect bandgap. Si rich particles in a SiO2 matrix have shown large Raman gain16,17, and Raman lasing was observed in micrometer scale Si photonic crystals13. But compared 3 ACS Paragon Plus Environment

Nano 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

Page 4 of 22

to other nanostructures such as nanoparticles and photonic crystals, nanowires are easier to integrate in devices because of their simple one-dimensional cylindrical geometry. They can also function as device elements and nanoscale interconnects, and hold potential for integration with future nanophotonic and electronic devices. In this work we report a large net Raman gain in Si nanowires from stimulated Raman scattering (SRS) in the visible spectral region with an extremely low threshold despite large absorption losses in this region. The overlap of the cavity mode of Si nanowire with the pump wavelength leads to high electric field intensity modes inside the nanowire (~30 times more intense than in bulk Si). This combined with the low mode volume of these electromagnetic modes leads to SRS at (external) pump intensities as low as 30 kW/cm2. A Raman gain coefficient value of 4×107 cm/GW at 532 nm is calculated, which is more than 106 greater than the reported value for bulk Si at 1.55 µm excitation18. Such high Raman gain can pave way for a monolithically integrable nanoscale Si laser, which may also enable seamless integration of a nanoscale optical source with other nanscale electronic and optical components, leading to the development of next generation of low powered nanodevices and nanosystems. To study SRS, Si nanowires (typical lengths >10 µm) grown via low-pressure chemical vapor deposition (see supporting information for details), were dispersed onto a glass substrate. Raman spectroscopy was performed on these nanowires using CW laser excitation at two different wavelengths (532 and 659.4 nm, spot size ~1µm) and varying pump intensity. The pump light was polarized along the nanowire long axis to increase in-coupling of optical power (schematic in Figure 1a, see supporting information for details). Excitation power dependent Raman measurements performed on a Si nanowire of diameter 100 nm showed that with increase in pump intensity, the Raman peaks red shift and broaden (Figure 1b). The integrated Stokes intensitycalculated from area under each Raman peak (Figure 1c, see supporting information for details), 4 ACS Paragon Plus Environment

Page 5 of 22 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

Nano Letters

also increased with an increase in pump intensity. At lower pump intensities, the Stokes scattering was linearly dependent on pump intensity indicating spontaneous Raman scattering but at higher pump intensities this dependence became nonlinear, suggesting the onset of SRS. At a pump intensity of 1 MW/cm2, the measured Stokes intensity was twice the value calculated from linear extrapolation of Stokes intensity at the lowest pump intensity (Figure 1c, broken line), corresponding to a 100% enhancement in Stokes scattering. The superlinear dependence of Stokes scattering on pump intensity is more clearly visible in log-log scale (inset- Figure 1c). The slope of this curve which is close to 1 at low pump intensity increased to ~2 at a pump intensity of 1 MW/cm2, clearly indicating nonlinearity and SRS at higher pump intensities19. The threshold for SRS, i.e., when the Raman scattering process becomes nonlinear is only ~30 kW/cm2 (pump intensity at which the slope of the log-log plot becomes greater than 1). It is indeed interesting to observe SRS in single Si nanostructures on a 100 nm lengthscale and that too in the visible spectrum where losses are ~8 orders of magnitude higher than at 1.55 µm (see supporting information for details). To understand the mechanism of strong SRS in the otherwise lossy spectral region, numerical calculations were performed to calculate the spatial distribution of electric field intensity inside the nanowire via the finite difference time domain (FDTD) method (see supporting information). Calculations for the Si nanowire of diameter 100 nm showed that high electric field intensity-whispering gallery (TM11) mode20 is excited in the nanowire near 532 nm (pump wavelength, Figure 2a) and 547 nm (Stokes wavelength, Figure 2b) with a fourfold enhancement in electric field intensity compared to incident electric field intensity. There is also a good spatial overlap of the electromagnetic modes at the pump and at Stokes wavelength inside the nanowire which was shown to be a critical factor in achieving SRS at low powers in nanostructured Si13. 5 ACS Paragon Plus Environment

Nano 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

Page 6 of 22

The average electric field intensity (Figure 2c) inside the nanowire is also enhanced by a factor of 1.5 (compared to incident electric field intensity) at both the wavelengths, which is much stronger than the average electric field intensity in bulk Si (only 5% of incident electric field intensity). Moreover, the mode volume of these electromagnetic modes is estimated to be ~3.5×10-4 µm3 in the Si nanowire compared to ~2.1×10-3 µm3 in bulk Si (see supporting information for details). This increase in electric field intensity inside the nanowire at both the pump and Stokes wavelength along with the spatial overlap of electromagnetic modes combined with their small mode volume leads to strong SRS in Si nanowires at low intensities. The strong electric field intensity inside the Si nanowire and above bandgap excitation wavelength (bandgap of Si at 302 K = 1.09 µm)3 also leads to strong absorption of incident light. However, because of the indirect bandgap of Si3 most of the absorbed light is converted into heat21, which coupled with the heat generated from Raman scattering causes an increase in nanowire temperature leading to a red shift22 and broadening23 of the Raman peak at higher pump intensities (Figure 1b). Nanowire temperature calculated from the red shift in Raman peak24 (see supporting information for details) increased with increase in pump intensity. A maximum temperature of 865K was attained at a pump intensity of 1 MW/cm2 (Figure 3a, black curve-left axis), an increase of ~570K from 295K. The broadening of Raman peak happens in part due to increased lattice anharmonicity25 at higher temperature and has also been previously reported during SRS measurements15. A comparison of the experimentally measured increase in full width at half maximum (FWHM) value of the Raman peak (Figure 3a, orange solid curve-right axis) with numerical calculations for expected increase in FWHM value at higher temperature24 (Figure 3a, orange broken curve-right axis) revealed more broadening (more FWHM) in the experimentally measured Raman peak than predicted from calculations. This extra increase in FWHM is caused 6 ACS Paragon Plus Environment

Page 7 of 22 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

Nano Letters

by non-uniform laser heating of the Si nanowire26 and asymmetric line broadening associated with Fano interactions between optically excited charge carriers and the lattice27, each of which has been shown to cause Raman peak broadening in Si nanowires at high pump intensities. While SRS has been reported earlier in bulk Si10,11, Si nanoparticles14, carbon nanotubes28 and GaP nanowires29, gain associated peak narrowing has not been observed, which is consistent with our observations. To further investigate the effect of temperature on SRS, FDTD calculations were performed to calculate the average electric field intensity inside the nanowire as a function of wavelength and temperature using temperature dependent permittivity of Si30. These calculations show that an increase in temperature leads to a decrease in the average electric field intensity confined inside the nanowire at pump and at Stokes wavelength (Figure 3b). At 865K, the product of electric field intensities inside the cavity at pump and at Stokes wavelength is decreased by a factor of 3 compared to the value at room temperature (Figure 3b, inset). Since SRS depends on the product of electric field intensities inside the cavity at pump and at Stokes wavelength, a threefold decrease in SRS intensity at 865 K (pump intensity- 1 MW/cm2) is expected. In other words, under otherwise identical conditions, the SRS intensity at pump intensity of 1 MW/cm2 would be three times the value experimentally recorded if the temperature of the system were to be maintained at 295 K. Other experimental19,31 and theoretical32 studies performed to investigate the temperature dependence of Raman scattering in Si also concluded that the intensity of Stokes scattering decreases with an increase in temperature. To analyze SRS without the deleterious effect of temperature in Si nanowires, temperature corrected Stokes intensity (for Figure 1c) was calculated using the empirical formula for temperature dependence of Stokes scattering provided by Compaan et.al33 (also see supplementary figure S1) which showed an even larger increase in 7 ACS Paragon Plus Environment

Nano 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

Page 8 of 22

Raman scattering intensity (Figure 3c). The slope of the temperature corrected curve was also steeper indicating stronger Raman nonlinearity if the deleterious effect of increased nanowire temperature were to be removed. At a pump intensity of 1 MW/cm2 the increase in temperature corrected Stokes scattering intensity was ~400% compared to a 100% increase in experimentally measured Stokes intensity highlighting the decrease in SRS intensity due to increase in temperature. It should be noted that the temperature correction factor for Stokes intensity for bulk Si at 865K is 2.5 which is similar to the value of 3 calculated by numerical calculations for nanowire. One of the reasons for higher value from numerical calculations could be the fact that at higher temperature, the cavity mode of the nanowire shifts away from the pump wavelength, leading to poor confinement of light at pump and Stokes wavelength and a further decrease in Raman scattering (and correspondingly a bigger temperature correction factor), an effect not seen in bulk. To investigate the effect of confinement of electric field intensity inside the nanowire at pump and Stokes wavelengths on SRS and also the wavelength tunability of SRS, FDTD calculations were performed to calculate the average electric field intensity inside the nanowire as a function of their diameter and excitation wavelength. These calculations reveal wavelength and nanowire diameter dependent electromagnetic modes where the electric field intensity is strongly confined inside the nanowires (Figure 4a). Based on these calculations it can be inferred that experiments conducted on a nanowire selected with high average electric field intensity inside the nanowire at the pump and Stokes wavelength should show SRS and hence Stokes intensity should be superlinearly dependent on pump intensity. Alternately, experiments on a nanowire with low electric field intensity at pump and Stokes wavelength should show spontaneous Raman scattering and Stokes intensity should vary linearly with pump intensity. To test this hypothesis, Raman 8 ACS Paragon Plus Environment

Page 9 of 22 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

Nano Letters

scattering experiments were done on several nanowires with varying diameter at two pump wavelengths 532 and 659.4 nm (green and red broken lines in Figure 4a). Results for four representative wires plotted in Figure 4b (with high and low average electric field intensity at each of the two pump wavelengths) show a good correlation between electric field intensity inside the nanowire and observance of SRS. Nanowires A (diameter, 220 nm) and B (diameter, 100 nm) have high average electric field intensity at 659.4 nm and 532 nm respectively. Correspondingly they show SRS and a superlinear dependence of Stokes scattering intensity on pump intensity. On the other hand, nanowires C (diameter,125 nm) and D (diameter, 100 nm) have low average electric field intensity at 532 and 659.4 nm respectively and hence do not show SRS, therefore confirming the dependence of SRS on electric field intensity inside the nanowire as well as the wavelength tunability of SRS. In order to quantify the gain observed in Si nanowires, calculations were performed using the model for pump and Stokes power evolution as proposed by Jones et.al34. Based on this model, net optical gain from SRS in a cavity is proportional to Stokes intensity from spontaneous Raman scattering and difference of Raman gain and losses in Si. By estimating spontaneous Raman scattering intensity from linear extrapolation of Stokes scattering intensity at lowest pump intensity, and losses from linear absorption coefficient of Si at 547 nm, a gain coefficient value of 4×107 cm/GW for Si nanowire of diameter 100 nm at 532 nm pump wavelength is calculated (see supporting information for details). This represents enhancement by a factor >106 in Raman gain coefficient compared to the value of 10 cm/GW reported for bulk Si at 1.55 µm18,19. Note that while our calculations assume 100% in-coupling and negligible losses of pump power inside the nanowire, accounting for lower in-coupling and pump power losses inside the nanowire, would give a larger value of gain coefficient than what is calculated here. Despite such high gain 9 ACS Paragon Plus Environment

Nano 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

Page 10 of 22

coefficient in these nanowires, lasing was not achieved primarily because of high losses leading to increase in cavity temperature which is detrimental to SRS. Similar to observations in bulk Si where a decrease in cavity losses by a factor of three decreased lasing threshold by an order of magnitude12; decreasing losses in Si nanowire should also help achieve lasing. In order to achieve that, experiments can be conducted at 1064 nm where losses are only ~ 0.1% of the losses at 532 nm. Moreover, the corresponding Stokes emission at 1190 nm would be below the bandgap of Si which would further decrease the loss of Stokes scattered light. Additionally, coating Si nanowires with a thick metal layer would provide an efficient heat sink to keep the nanowire at lower temperature and also lead to stronger confinement of light inside the nanowire20. Numerical calculations show (supplementary figure S2) a seventy-fold enhancement of electric field intensity inside the nanowire of diameter 475 nm coated with 250 nm Ag at 1060 nm excitation due to the formation of metal-dielectric cavity modes with stronger confinement. If Raman experiments were to be conducted on resonant Si nanowire cavities at this wavelength, it may lead to lasing in these Si nanowires due to lower losses and stronger confinement of light. Thus, by using the high Raman gain coefficient nanowires it should be possible to make a low powered Si nanolaser in the infrared wavelength region. In conclusion, we have demonstrated that by aligning the nanowire cavity mode with the pump wavelength, strong SRS can be observed in Si nanowires of various diameters at different wavelengths. Raman gain coefficient of Si nanowire can be enhanced by a factor greater than 106 at 532 nm excitation despite 8 orders of magnitude higher losses compared to at 1.55 µm. This enhancement is demonstrated to be arisen from higher electric field intensity and lower mode volume of the electromagnetic modes inside the Si nanowire compared to bulk Si at pump and Stokes wavelength, as well as the spatial overlap of these modes inside the cavity. This is a 10 ACS Paragon Plus Environment

Page 11 of 22 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

Nano Letters

significant step towards the development of a homogeneous Si nanolaser with low lasing threshold, and could lead to the development of first monolithically integrable Si Raman nanolaser in the infrared as well as telecom wavelength range


Nano 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

Page 12 of 22

Supporting Information Methods for nanowire synthesis and diameter measurement, optical measurements, calculations for temperature, Stokes intensity, temperature dependence of Stokes intensity, wavelength dependent losses in Si, Raman gain coefficient and numerical calculations. Author contributions DA and RA conceived the concept and designed the optical experiments. JY synthesized silicon nanowires. DA performed optical experiments and numerical calculations. DA, MLR, JB, AP and RA analyzed the results and wrote the manuscript.

Acknowledgements We thank funding from the University Research Foundation at the University of Pennsylvania for their support. This work was performed in part at CINT, a U.S. Department of Energy, Office of Basic Energy Sciences User Facility at Los Alamos National Laboratory (Contract DE-AC52-06NA25396) and Sandia National Laboratories (Contract DE-AC0494AL85000). A.P. thanks the support from the National Natural Science Foundation of China (Nos. 51525202, 61574054).


Page 13 of 22 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

Nano Letters

References 1. Alduino, A.; Paniccia, M. Nat. Photonics 1, 153-155 (2007). 2. Paniccia, M. Nat.Photonics 4, 498-499 (2010). 3. Liang, D.; Bowers, J. E. Nat. Photonics 4, 511-517 (2010). 4. Liu, A.; Liao, L.; Rubin, D.; Basak, J.; Chetrit, Y.; Nguyen, H.; Cohen, R.; Izhaky, N.; Paniccia, M. Semicond. Sci. Technol. 23, 064001 (2008). 5. Leuthold, J.; Koos, C.; Freude, W. Nat. Photonics 4, 535-544 (2010). 6. Claps, R.; Raghunathan, V.; Dimitropoulos, D.; Jalali, B. Opt. Express 12, 2774-2780 (2004). 7. Jalali, B. J. of Lightwave Tech. 24, 4600-4615 (2006). 8. Liu, A.; Rong, H.; Paniccia, M. Opt. Express 18, 4261-4268 (2004). 9. Jones, R.; Rong, H.; Liu, A.; Fang, A. W.; Paniccia, M.J. Opt. Express 13, 519-525 (2005). 10. Boyraz, O.; Jalali, B. Opt. Express 12, 5269-5273 (2004). 11. Rong, H.; Jones, R.; Liu, A.; Cohen, O.; Hak, D.; Fang, A.; Paniccia, M. Nature 433, 292294 (2005). 12. Rong, H.; Xu, S.; Kuo, Y.; Sih, V.; Cohen, O.; Raday, O.; Paniccia, M. Nat. Photonics 1, 232-237 (2007). 13. Takahashi, Y.; Inui, Y.; Chihara, M.; Asano, T.; Terawaki, R.; Nosa, S. Nature 498, 470-474 (2013). 13 ACS Paragon Plus Environment

Nano 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

Page 14 of 22

14. Cho, C.; Aspetti, C. O.; Park, J.; Agarwal, R. Nat. Photonics 7, 285 – 289 (2013) 15. Aspetti, C. O.; Cho, C.; Agarwal, R.; Agarwal, R. Nano Lett. 14, 5413 – 5422 (2014) 16. Sirleto, L.; Ferrara, M. A.; Niktin, T.; Novikov, S.; Khriachtchev, L. Nat. Comm. 3, 1220 (2012). 17. Niktin, T.; Novikov, S.; Khriachtchev, L. Appl. Phys. Lett. 103, 151110 (2013). 18. Hon, N. K.; Soref, R.; Jalali, B. J. Appl. Phys. 110, 011301 (2011). 19. Boyd, R. W. Nonlinear Optics, 3rd Edition, Academic Press: Cambridge, Massachusetts (2008). 20. Ren, M.; Liu, W.; Aspetti, C. O.; Sun, L.; Agarwal, R. Nat. Comm. 5, 5432 (2014). 21. Agarwal, D.; Aspetti, C. O.; Cargnello, M.; Ren, M.; Yoo, J.; Murray, C. B.; Agarwal, R. Nano Lett. 17, 1839-1845 (2017). 22. Niu, J.; Sha, J.; Yang, D. Scr. Mater. 55, 183-186 (2006). 23. Su, Z.; Sha, J.; Pan, G.; Liu, J.; Yang, D.; Dickinson, C.; Zhou, W. J. Phys. Chem. B 110, 1229-1234 (2006). 24. Balkanski, M.; Wallis, R.; Haro, E. Phys. Rev. B 28, 1928-1934 (1983). 25. Ipatova, I. P.; Maradudin, A. A. Phys. Rev. 155, 882-895 (1967). 26. Piscanec, S.; Cantoro, M.; Ferrari, A. C.; Zapien, J. A.; Lifshitz, Y.; Lee, S. T.; Hofmann, S.; Robertson, J. Phys. Rev. B 68, 241312 (2003). 27. Gupta, R.; Xiong, Q.; Adu, C. K.; Kim, U. J. ; Eklund, P. C. Nano Lett. 3, 627-631 (2003). 14 ACS Paragon Plus Environment

Page 15 of 22 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

Nano Letters

28. Zhang, B. P.; Shimazaki, K.; Shiokawa, T.; Suzuki, M.; Ishibashi, K.; Saito, R. Appl. Phys. Lett. 88, 241101 (2006). 29. Wu, J.; Gupta, A. K.; Gutierrez, H. R.; Eklund, P. C. Nano Lett. 9, 3252-3257 (2009). 30. Jellison Jr, G. E.; Modine, F. A. J. Appl. Phys. 76, 3758-3761 (1994). 31. Compaan, A.; Trodah, H. J. Phys. Rev. B 29, 793-801 (1984). 32. Venkateshwarlu, K. Nature 159, 96-97 (1947). 33. Compaan, A.; Lee, M. C.; Trott, G. J. Phys. Rev. B 32, 6731-6741 (1985). 34. Jones, R.; Rong, H.; Liu, A.; Fang, A. W.; Paniccia, M. J.; Hak, D.; Cohen, O. Opt. Express 13, 519–525 (2005).


Nano 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

Page 16 of 22

Figure Captions Figure 1: Stimulated Raman Scattering in Si nanowires: a) Schematic of Raman experiments on a Si nanowire in backscattering geometry. Green arrows indicate the direction of the propagation of the incident wave and the back scattered Raman Stokes wave. Black broken arrow shows the direction of polarization of incident electric field, inset shows the scanning electron microscopy image of a representative Si nanowire (scale bar is 400 nm); b) Pump intensity dependent Stokes Raman spectra of a Si nanowire of diameter 100 nm with a 532 nm pump, legend shows the pump intensity; c) Integrated Stokes intensity of the same wire as a function of pump intensity (linear scale). Broken line is the extrapolation of Stokes intensity using the Raman scattering data at the lowest pump intensity and represents the expected Stokes intensity from spontaneous Raman scattering. Inset shows the same data in log-log scale.

Figure 2: High electric field intensity inside the nanowire at cavity mode resonance: a, b) Electric field intensity distribution (normalized to incident electric field intensity) inside the Si nanowire of diameter 100 nm at pump wavelength (532 nm) and at Stokes wavelength (547 nm). Region inside the inner white circle represents the nanowire. Region in between the two white circles represents the native oxide layer; c) Average electric field intensity inside the nanowire (normalized to incident electric field intensity) inside Si nanowire and bulk Si as a function of excitation wavelength.

Figure 3: Effect of Temperature on Stimulated Raman Scattering: a) left axis: Temperature of the Si nanowire as a function of pump intensity, right axis: experimentally measured increase 16 ACS Paragon Plus Environment

Page 17 of 22 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

Nano Letters

in full width at half maximum (FWHM) values and calculated24 temperature dependent increase in FWHM values (line connecting the points is a guide to the eye). Black and orange horizontal arrows indicate the y-axis corresponding to the temperature and FWHM data points; b) FDTD calculations for average electric field intensity inside the nanowire (normalized to incident electric field intensity) as a function of temperature. Inset shows the product of pump electric field intensity (Ipump) and Stokes electric field intensity (ISt), which is directly proportional intensity of SRS, inside the nanowire as function of temperature; c) Experimentally measured and temperature corrected integrated Stokes intensity (line connecting the points is a guide to the eye), and linear extrapolation of Stokes intensity from the lowest pump intensity as a function of pump intensity of Si nanowire of diameter 100 nm

Figure 4: Effect of nanowire cavity mode alignment with pump wavelength on Stimulated Raman Scattering: a) FDTD calculations of average electric field intensity inside the nanowire (normalized to incident electric field intensity) as a function of nanowire diameter and excitation wavelength. Vertical broken lines represent the two pump wavelengths- 532 nm (green line) and 659.4 nm (red line); b) Experimentally measured integrated Stokes intensity as a function of laser intensity in Si nanowires of diameter 220 nm (A) and 100 nm (D) at 659.4 nm and of diameter 100 nm (B) and 125 nm (C) at 532 nm. Inset shows the schematic of Raman experiments.


Nano Letters

Figure 1

a)

b)

Counts (a.u.)

1.0

2

1.03 MW/cm 2 0.361 MW/cm 2 0.082 MW/cm 2 0.012 MW/cm

λpump = 532 nm

0.5

0.0 460

480

500

520

-1

540

560

Raman shift (cm )

c)

2

Log (Stokes Intensity)

3

Stokes Intensity (a.u.)

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

Page 18 of 22

λpump = 532 nm

1

0.1

0.01 0.01

1

0 0.0

0.1 Log(Pump Intensity)

1

Si nanowire Spontaneous Raman

0.4

0.8

2

1.2

Pump Intensity (MW/cm )


Page 19 of 22

Figure 2

a)

|E|2

532 nm

Si

b)

|E|2 547 nm

Si

c) Average field intensity (a.u.)

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

Nano Letters

2

λpump λRaman

Si NW Si bulk

1

0 500

600

Wavelength (nm)

700


Nano Letters

Figure 3

a) 16

Temperature

700 8

500

300 0.0

Average field intensity (a.u.)

b)

Experimental 24 Calculated

0.4 0.8 2 Pump Intensity (MW/cm )

2

[I(St)I(Pump)]  SRS

Temperature (K)

900

1

FWHM increase

0

1.2

3

2

1

0 300

600

Temperature (K)

900

300 K 500 K 750 K 850 K

0 500

600

700

Wavelength (nm)

c) 5.0


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

Page 20 of 22

Temperature corrected Experimental Spontaneous Raman

2.5

0.0 0.0

0.4

0.8

2

1.2



Page 21 of 22

Figure 4

a)

|E|2

b) 3

λ =659.4 nm λ =532 nm

A

C B

D


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

Nano Letters

Spontaneous Raman λ=659.4nm (A,D) λ=532nm (B,C)

2 B, d=100 nm A, d=220 nm

C, d=125 nm

1

D, d=100 nm

0 0.0

0.3

0.6

0.9

1.2 2



Nano 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

Page 22 of 22

TOC Graphic


Nanocavity-Enhanced Giant Stimulated Raman ... - ACS Publications

Recommend Documents