Material reference

Gallium arsenideGaAs

The direct-gap workhorse of the 780–1100 nm range: VCSELs, pump diodes, and high-power bars. A strong second-order nonlinearity keeps it interesting for integrated nonlinear optics.

GaAs was the material of the first room-temperature semiconductor lasers, and the GaAs/AlGaAs system remains close to ideal for device engineering: AlAs and GaAs are lattice-matched to within about 0.1 % at every aluminum fraction, so heterostructures, quantum wells, and DBR mirror stacks can be grown with near-total compositional freedom. That accident of crystallography is why 850 nm VCSELs, with their dozens of alternating AlGaAs mirror pairs, are a GaAs technology.

Key properties

Property	Value	Conditions
Refractive index $n$	3.374	1550 nm, 300 K
Bandgap $E_g$	1.424 eV (direct)	300 K
Bandgap wavelength	871 nm	300 K
Emission coverage	~630–1100 nm	AlGaAs / InGaAs on GaAs
Second-order nonlinearity $d_{14}$	~120 pm/V	1533 nm
Electro-optic coefficient $r_{41}$	~1.5 pm/V
Thermo-optic coefficient $dn/dT$	$\approx +2.3 \times 10^{-4}$ K⁻¹	near-IR
Thermal conductivity	55 W·m⁻¹·K⁻¹	300 K

What the numbers mean in practice

The 1.424 eV direct gap puts native GaAs emission at 870 nm; adding aluminum pushes it shorter (up to the direct–indirect crossover near $x \approx 0.45$ ), while strained InGaAs wells pull it longer, covering the 9xx nm pump band that feeds every EDFA and fiber laser. If a photon at 808, 850, 940, or 980 nm did work today, a GaAs-family device almost certainly produced it.

Below the gap, GaAs is transparent well into the mid-IR and carries one of the largest second-order nonlinearities of any mainstream photonic material, roughly four times lithium niobate's $d_{33}$ . The catch is that its cubic symmetry rules out birefringent phase matching, so quasi-phase-matching requires orientation patterning (OP-GaAs) or clever waveguide-mode tricks rather than the ferroelectric poling that makes PPLN straightforward.

Compared with InP, GaAs wafers are larger and cheaper and the technology base (including RF electronics) is broader; the limitation is simply that lattice-matched gain past ~1.1 μm gets hard, which is why the telecom bands belong to InP.

References

T. Skauli et al., "Improved dispersion relations for GaAs and applications to nonlinear optics," J. Appl. Phys. 94, 6447 (2003).
I. Shoji, T. Kondo, A. Kitamoto, M. Shirane, R. Ito, "Absolute scale of second-order nonlinear-optical coefficients," J. Opt. Soc. Am. B 14, 2268 (1997).
I. Vurgaftman, J. R. Meyer, L. R. Ram-Mohan, "Band parameters for III-V compound semiconductors and their alloys," J. Appl. Phys. 89, 5815 (2001).
E. D. Palik (ed.), Handbook of Optical Constants of Solids, Academic Press (1985).