Infrared dipole antenna enhanced bysurface phonon polaritons

Hyun Chul Kim1,2 and Xing Cheng1,*1Department of Electrical and Computer Engineering, Texas A&M University, Mail Stop 3128 TAMUS,

College Station, Texas 77843, USA2Currently at DRAM Process Architecture Team, Memory Division, Samsung Electronics Company, Ltd.,

San #16 Banwol-Dong, Hwasung-City, Gyeonggi-Do, 445-701, Korea*Corresponding author: chengx@ece.tamu.edu

Received April 29, 2010; revised August 31, 2010; accepted October 11, 2010;posted October 14, 2010 (Doc. ID 127640); published November 3, 2010

In this Letter, we propose a gold dipole antenna formed on a SiC substrate to achieve a strong concentration of mid-IR radiation based on a synergistic integration of the IR dipole antenna and the resonance excitation of a surfacephonon polariton. Numerical simulation based on the finite-difference time-domain technique shows that the in-tensity enhancement can be greater than 107 times at the mid-IR spectral region. The influence of the geometricparameters (i.e., antenna length, gap dimension, antenna thickness, and antenna width) on the antenna fieldenhancement is also studied. The strong intensity enhancement can find important applications in highly sen-sitive mid-IR photodetectors and in molecular detection and identification by surface-enhanced IR absorptionspectroscopy techniques. © 2010 Optical Society of AmericaOCIS codes: 240.0240, 240.5420, 300.6340, 260.3060.

The IR detector is a photodetector that responds to IRradiation and is widely used in industrial processesand scientific equipment. There are two types of IR de-tectors: thermal and photonic [1]. The thermal-type de-tects the IR energy based on conductivity change dueto IR-induced heating. It does not require cooling, butit has slow response time and low detectivity. In contrast,the photonic-type IR detector based on optical genera-tion of free carriers has a high detection sensitivityand fast response time. High dark current due to ther-mally generated carriers significantly lowers the perfor-mance of photonic IR detectors. Cryogenic cooling isnecessary to suppress thermal generation of the carriers,which makes the detector system bulky and unsuitablefor portable devices. It is thus highly desirable to developsensitive and compact uncooled photonic-type IR detec-tors by focusing the incident IR radiation using varioustechniques.Surface plasmon polaritons (SPPs) are transverse sur-

face charge waves accompanying electromagnetic fieldslocalized at an interface between a metal and a dielectricat the visible and near-IR region [2]. In recent years, me-tal nanostructures have received considerable attentionfor their ability to guide and manipulate SPPs at the na-noscale [3]. They have offered many potential applica-tions, such as nanolithography, near-field scanningoptical microscopy, optical spectroscopy, data storage,and biosensing. However, at lower frequencies (i.e.,mid-IR range), the plasmonic response of metal is eitherfar too weak or nonexistent. Recently, it has been an-nounced that SPPs have a counterpart in the mid-IR spec-tral region, the so-called surface phonon polaritons(SPhPs) [4]. SPhPs arise due to the coupling of the elec-tromagnetic field to the lattice vibrations of polar dielec-trics at mid-IR frequencies. The physics of theseexcitations are conceptually analogous to those of bothpropagating and localized SPPs.Another technique to concentrate electromagnetic ra-

diation in a confined geometry is using the antenna struc-ture. Resonant optical antennas can confine a strongly

optical near field in a subwavelength volume, whichhas been recently demonstrated for dipole antennasand bow-tie antennas at the visible region [5,6]. Opticalantennas operating in the IR wavelengths are also shownto be able to strongly confine the IR intensity [7–9]. In thisLetter, we propose a synergistic integration of SPhPmodes and optical antenna at the IR wavelengths tosignificantly increase the IR field enhancements com-pared to those achievable with IR antennas or SPhPmodes alone.

Silicon carbide (SiC) is an attractive wide bandgapsemiconductor for high-temperature and high-powermicroelectronics [10]. Another interesting property ofthis material is its negative dielectric permittivity inthe mid-IR spectral range between 10.3 and 12:6 μm,as shown in Fig. 1(a) [11]. Here, the real part of the di-electric constant, ReðεÞ, is negative in the so-called Rest-strahlen band between the transverse and longitudinaloptical phonon frequencies, which gives rise to the exci-tation of SPhPs. Note that their damping characteristicsgiven by ImðεÞ can be much smaller than those for SPPs.

The dispersion relation for the wave vector of theSPhPs is given by [12]

kSPhPðωÞ ¼ωc

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiε1ðωÞε2

ε1ðωÞ þ ε2

s; ð1Þ

where kSPhP is the wave vector of the SPhPs and ε1ðωÞand ε2 are dielectric constants for SiC and air, respec-tively. As shown in Fig. 1(b), we calculated the dispersionrelation of the SPhPs at the air/SiC interface. SPhPs canonly propagate close to the interface between two dif-ferent media, and the amplitude of the fields decays ex-ponentially with the distance from the interface. We seethat the resonant condition between the two media isRe½ε1ðωÞ� ¼ −ε2; thus the maximum Re½kSPhP=k0� is at10:55 μm, the resonant wavelength. The propagationlength of SPhPs given by ð2Im½kSPhP�Þ−1 decreaseswith the increasing of Re½kSPhP=k0�. The real part of

3748 OPTICS LETTERS / Vol. 35, No. 22 / November 15, 2010

0146-9592/10/223748-03$15.00/0 © 2010 Optical Society of America

the dielectric permittivity of SiC at the wavelength of10:6 μm used in all simulations in this study is ReðεÞ ¼−1:15 [13].The schematic diagram of the IR antenna is shown in

Fig. 2(a). Optical energy can be coupled into SPhP modesat discontinuities in the materials [14,15]. Similarly in thisstudy, the edge of the gold dipole on the SiC substratecan excite SPhPs, which is the so-called edge couplingmethod [15,16].The intensity enhancement in the gap region of the

gold dipole antenna on various substrates and incidentwavelengths, including the SiC substrate at 10:2 μm(no SPhP excitation), is simulated and analyzed by the3D finite-difference time-domain method. The dielectricpermittivity of gold and SiC at the 10:6 μm wavelength isεAu ¼ −3128:7þ 1558:5i and εSiC ¼ −1:15þ 0:13i, re-spectively [17], and the IR radiation to excite the SPhPsis a circularly polarized wave normally incident to the topof the dipole antenna. Figure 2(b) shows the intensity dis-

tribution in the x–y plane for a gold dipole antenna on theSiC substrate. It is clearly seen that the field intensity isstrongly localized at the gap and edges of the antenna.

Figure 3 shows the intensity enhancement in the gapregion of the gold dipole antenna on the SiC substrateas a function of antenna length. The corresponding di-pole antennas have 100 nm in gap dimension, 100 nmin antenna thickness, and 400 nm in antenna width.When the antenna length increases, the dipole antennaexhibits an increase of the field intensity until the anten-na length reaches the first resonant length. After that, thefield intensity decreases gradually. The physical origin ofthis effect is connected to the coupling strength of thetwo antenna arms. A strong field enhancement of about2 × 107 times is achieved at a resonant antenna length of16:4 μm. To verify the impact of the resonant excitationof the SPhPs, simulations are also performed for severalcases: dipole antenna on Si substrate, freestanding IRantenna, and dipole antenna on SiC substrate at 10:2 μmwavelength (no SPhP excitation). The simulation resultsare also plotted in Fig. 3. All parameters are identical tothe aforementioned gold dipole antenna on the SiC sub-strate. The maximum field enhancement of the dipole an-tenna on the SiC substrate at 10:2 μm, at which point theSiC substrate has positive dielectric permittivity and doesnot support SPhPs, is about 2 × 104 times at the resonantantenna length. Likewise, the maximum field enhance-ments of the dipole antenna on the Si substrate and free-standing IR antenna are achieved by approximately 80times and 1200 times, respectively, at the resonantlengths. It is clearly seen that due to the synergisticaction of the dipole antenna and the resonant excitationof the SPhPs, field enhancement in the gap region canreach more than 4 orders of magnitude higher than thatof the freestanding dipole antenna. It is also noted thatthe Si substrate antenna has a much smaller first re-sonant length than the freestanding antenna due to thesubstrate effect of the higher refractive index (i.e., nSi ¼3:42 at 10:6 μm) [18]. Recent studies have shown that the

Fig. 1. (Color online) (a) Real and imaginary parts of the di-electric permittivity of SiC. (b) Real and imaginary parts of thedispersion relation for SPhPs at air/SiC interface are calculatedaccording to Eq. (1).

Fig. 2. (Color online) (a) Schematic diagram of an IR dipoleantenna enhanced by SPhPs. (b) Intensity distribution in thex–y plane (top view).

Fig. 3. Intensity enhancements in the freestanding IR antennaand IR antennas on SiC and Si substrates at 10:6 μm. Theintensity enhancement in IR antenna on the SiC substrate at10:2 μm is much lower, because no SPhPs can be excited at thiswavelength.

November 15, 2010 / Vol. 35, No. 22 / OPTICS LETTERS 3749

resonant length of the optical dipole antenna is about 20%shorter than the value predicted by antenna theory. Theantenna length of the optical antennas scales with the ef-fective wavelength of the incident light [19]. Both the Sisubstrate antenna and freestanding antenna have reso-nant lengths close to the predicted values. The dipole an-tenna on the SiC substrate, however, has a resonantlength larger than the predicted value. In our view, it ap-pears that the field enhancement originates not only fromdipole antenna coupling but also from the resonant cou-pling of the SPhPs between the incident IR radiation andthe electromagnetic surface modes due to negativedielectric permittivity provided by lattice vibrations inpolar crystals.Simulations are also performed to evaluate the effect

of geometric parameters on field enhancement, as shownin Fig. 4. The corresponding dipole antenna has an anten-na length of 16:4 μm, and all other parameters are iden-tical to those at which maximum field enhancement wasachieved. Among them, the field enhancement of the di-pole antenna obviously depends on its gap dimension: re-ducing the gap dimension rapidly increases the intensityenhancement, as shown in Fig. 4(a) [20]. Recently, No-votny demonstrated that the antenna thickness affectsthe effective wavelength that provides the resonantlength of an optical dipole antenna [19]. The effectivewavelength is linearly proportional to the thickness ofthe antenna. Figure 4(b) shows the antenna thicknessversus field enhancement. The field enhancement de-creases steadily with the antenna thickness, becausethe effective wavelength is gradually out of resonance

with the antenna with a fixed arm length. The effect ofantenna width is shown in Fig. 4(c). The field enhance-ment is not significantly varied with the changes in anten-na width. The maximum field enhancement is achieved atan antenna width of 400 nm.

In summary, we have numerically investigated the IRfield enhancement in a dipole antenna with SPhPs. Be-cause of the synergistic action of the dipole antennaand the resonant excitation of SPhPs, the strong field en-hancement in the gap region of the gold dipole antennaon SiC substrate reaches a value that is more than 4 or-ders of magnitude higher than that of the freestanding IRdipole antenna. The field enhancement varies strongly onthe gap dimension and antenna thickness. This strongfield confinement is expected to find promising applica-tions in molecule sensing by IR absorption and highlysensitive IR detectors.

References

1. A. Rogalski, Prog. Quant. Electon. 27, 59 (2003).2. H. Raether, Surface Plasmons (Springer, 1988).3. W. L. Barnes, A. Dereux, and T. W. Ebbesen, Nature 424,

824 (2003).4. S. A. Maier, Plasmonics—Fundamentals and Applications

(Springer, 2007).5. P. Muhlschlegel, H. J. Eisler, O. J. F. Martin, B. Hecht, and

D. W. Pohl, Science 308, 1607 (2005).6. A. Sundaramurthy, P. J. Schuck, N. R. Conley, D. P. Fromm,

G. S. Kino, and W. E. Moerner, Nano Lett. 6, 355 (2006).7. I. Codreanu and G. D. Boreman, Infrared Phys. Tech. 43,

335 (2002).8. I. Codreanu, F. J. Gonzalez, and G. D. Boreman, Infrared

Phys. Tech. 44, 155 (2003).9. C. Fumeaux, M. A. Gritz, I. Codreanu, W. L. Schaich, F. J.

González, and G. D. Boreman, Infrared Phys. Tech. 41,271 (2000).

10. C. Rockstuhl, M. G. Salt, and H. P. Herzig, J. Opt. Soc. Am. B22, 481 (2005).

11. N. Ocelic and R. Hillenbrand, Nat. Mater. 3, 606 (2004).12. P. B. Catrysse and S. Fan, Phys. Rev. B 75, 075422 (2007).13. J. Renger, S. Grafstrom, L. M. Eng, and R. Hillenbrand,

Phys. Rev. B 71, 075410 (2005).14. A. V. Zayats, I. I. Smolyaninov, and A. A. Maradudin, Phys.

Rep. 408, 131 (2005).15. A. J. Huber, B. Deutsch, L. Novotny, and R. Hillenbrand,

Appl. Phys. Lett. 92, 203104 (2008).16. A. Huber, N. Ocelic, D. Kazantsev, and R. Hillenbrand, Appl.

Phys. Lett. 87, 081103 (2005).17. E. D. Palik, Handbook of Optical Constants of Solids

(Academic, 1985).18. K. B. Crozier, A. Sundaramurthy, G. S. Kino, and C. F.

Quate, J. Appl. Phys. 94, 4632 (2003).19. L. Novotny, Phys. Rev. Lett. 98, 266802 (2007).20. H. Fischer and O. J. F. Martin, Opt. Express 16, 9144 (2008).

Fig. 4. Field enhancement as a function of (a) gap dimension,(b) antenna thickness, and (c) antenna width.

3750 OPTICS LETTERS / Vol. 35, No. 22 / November 15, 2010