PerspectiveApplied Physics

Seeking the Ultimate Nanolaser

See allHide authors and affiliations

Science  13 Oct 2006:
Vol. 314, Issue 5797, pp. 260-261
DOI: 10.1126/science.1131322

Semiconductor lasers generally emit a large amount of undesired spontaneous emission before starting lasing oscillation, which degrades their efficiency and performance substantially. Therefore, lasers that emit almost no spontaneous emission have long been sought. Such devices are called “thresholdless lasers,” where light output versus excitation power has no obvious threshold characteristic and lasing occurs at extremely low excitation powers. These lasers should have the maximum allowable performance and thus be very useful for optical applications. One promising approach has been to construct lasers with a nanocavity in a photonic crystal, in which the optical properties are structurally designed rather than intrinsic to the material. The photonic crystals and nanocavities can then be tailored to control spontaneous emission to achieve thresholdless operation. Recent progress in the engineering of photonic crystal nanocavities and their combination with quantum dots has accelerated this effort (14).

Several key issues (5) must be addressed before thresholdless lasers can be realized. The threshold behavior of semiconductor lasers arises from spontaneous emission coupled to the many optical modes inherent to the laser cavity. Only one of these can be the lasing mode. The undesired spontaneous emission into other modes dissipates the excited carriers in the semiconductor and, consequently, the laser efficiency is degraded. Before the ultimate laser can be realized, the following issues must be addressed: (i) Optical modes that induce undesired spontaneous emission should be suppressed where possible; (ii) a single-cavity mode with a sufficiently high Q factor (the so-called quality factor of the cavity) and a small modal volume is essential; and (iii) excited carriers should be concentrated to emit light coupled to the single-cavity mode.

Connecting the dots.

A high-Q nanocavity (red) is formed in a 2D photonic-crystal slab that inhibits spontaneous emission. Quantum dots embedded in the nanocavity confine electrons (and holes) three-dimensionally and have sharp gain functions. Such a configuration may allow thresholdless laser operation.

Regarding the issue (i), the suppression of spontaneous emission has recently been demonstrated (2, 3). Progress in the development of two-dimensional (2D) photonic-crystal slabs (see the figure) has been integral to this achievement. The 2D slab structure facilitates a quasi-three-dimensional (3D) confinement of photons as a result of the large refractive-index contrast perpendicular to the slab. Although spontaneous emission from light emitters inside the 2D slab structure can be coupled to confined or leaky optical modes, it is possible to couple ∼94% of the spontaneous emission to the confined mode. Therefore, the use of a 2D photonic-bandgap structure can inhibit ∼94% of the spontaneous emission (3). Recent experiments have suppressed the spontaneous-emission rate of this system by roughly the theoretical limit (∼15 times) (6).

As for the fabrication of appropriate cavity modes, a single mode can be introduced by forming an artificial defect in a 2D photonic-crystal slab (see the figure) (1). The modal volume (V) of this defect (nanocavity) can be on the order of a cubic wavelength. When the Q factor of the cavity is sufficiently large, the emission coupled to the single-cavity mode can be substantially enhanced by a factor of Q/V, which is called the Purcell effect. When the emission rate in the laser mode is much greater than that coupled to the residual leaky optical modes in the 2D slab, thresholdless operation (7) would be possible. Thus, the Q factor of the nanocavity is an important measure of the optimization of the desired emission process. Nanocavity Q factors have improved from hundreds in 1999–2000 (8) to ∼50,000 in 2003 (1), ∼600,000 in 2005 (9), and >1 million currently (10, 11). Although Q factors of >1 million are not essential for thresholdless lasers, the realization of such ultrahigh-Q nanocavities is important to control the interaction between photonic and electronic systems.

Ongoing studies also aim to demonstrate that the carriers stored by the suppression of spontaneous emission can be used to induce emission coupled to the single-cavity mode instead of the residual leaky modes in a 2D slab (6). Quantum dots (QDs), which can confine carriers three-dimensionally, are the most promising light emitters to be introduced into nanocavities. This 3D carrier confinement allows nonradiative processes to be suppressed. In addition, the gain curve becomes sharp due to the delta-function-like density of states of QDs (12). Furthermore, QDs can reach absorption saturation easily due to their strong nonlinearity, and a high Q factor of the nanocavity can be maintained even during the initial stages of the excitation. However, the bottleneck blocking the demonstration of thresholdless operation arises because high-quality QDs can be obtained only by self-assembly methods; hence, the wavelengths and positions of the QDs in the nanocavity are random (see the figure). If the cavity mode is resonant with respect to the wavelength and position of the QD, the Purcell effect occurs (2, 13), and the carriers stored by inhibiting spontaneous emission can be used mostly for emission coupled to the single-cavity mode, allowing thresholdless operation to occur. However, if the wavelengths and positions of QDs are not resonant with the cavity mode, the Purcell effect is suppressed and the emission rate of the cavity mode cannot be improved (2), leading to the consumption of excited carriers by emission coupled to the residual leaky modes. Thus, a clear demonstration of thresholdless operation remains to be achieved. Nevertheless, a recent paper (4) reported that the self-tuned QD gain effect is feasible, in which nearly thresholdless behavior might be obtained even in the off-resonant case. Although a detailed investigation of this effect is necessary, the report could be useful in addressing the matter of carrier concentration.

Major progress toward the realization of thresholdless nanolasers has clearly been achieved with 2D photonic crystal-based nanocavities and their fusion with QDs. However, to accomplish thresholdless laser operation, more needs to be done, including detailed investigations to clarify the interactions between the nanocavity and QDs in an off-resonant condition, and between individual QDs inside the nanocavity. There also needs to be progress in the development of 3D photonic crystals (14, 15). The suppression of undesired spontaneous emission would be at least 10 to 100 times as great as that in a 2D photonic-crystal slab. Even if the QDs and the nanocavity mode were offresonant, thresholdless operation would be expected because spontaneous emission coupled to residual leaky modes would be reduced to <0.06 to 0.6%. And finally, we must find an appropriate method of current injection (16) by which the Q factor of the nanocavity is not degraded. The future looks bright on all these fronts.

References and Notes

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
View Abstract

Stay Connected to Science

Navigate This Article