Diode Laser

Semiconductor Diode Lasers: Complete Theory from First Principles

Semiconductor diode lasers are the most efficient and compact coherent light sources available. Their operation relies on the fundamental physics of direct band-gap semiconductors, non-equilibrium carrier statistics, and optical feedback in a p-n junction.

1. Semiconductor Band Structure

A laser diode is constructed from a direct band-gap semiconductor. For emission at $\lambda \approx 808$ nm (commonly used to pump Nd:YAG lasers), the material is typically Gallium Aluminum Arsenide (GaAlAs).

1.1 Energy Bands

Valence Band (VB): The highest energy band that is completely filled with electrons at absolute zero.
Conduction Band (CB): The next lowest energy band, which is empty at equilibrium at absolute zero.
Band Gap ($E_g$): The energy difference between the bottom of the conduction band ($E_c$) and the top of the valence band ($E_v$).

\[ E_g = E_c - E_v \]

For GaAlAs at 808 nm emission:

\[ E_g \approx 1.53 \text{ eV} \quad \Rightarrow \quad \lambda = \frac{hc}{E_g} \approx 808 \text{ nm} \]

1.2 Direct vs. Indirect Band Gap

Optical transitions must conserve both energy and crystal momentum ($k$).

Direct Band Gap: The minimum of the CB and the maximum of the VB occur at the same wavenumber ($k$). Since the momentum of a photon is negligible, transitions are vertical in $k$-space ($\Delta k \approx 0$). This allows for highly efficient photon emission.
Indirect Band Gap (e.g., Si, Ge): The band extrema are at different $k$ values. A transition requires a phonon to conserve momentum, making radiative recombination inefficient and unsuitable for lasers.

2. The p-n Junction under Forward Bias

A diode laser is essentially a forward-biased p-n junction containing an active region (either bulk or quantum well).

Equilibrium: A built-in potential barrier prevents current flow.
Forward Bias: An external voltage $V$ reduces the barrier, injecting electrons from the n-side and holes from the p-side into the active region.

This injection creates a non-equilibrium state where carrier densities far exceed their thermal equilibrium values.

3. Quasi-Fermi Levels (Non-Equilibrium Statistics)

In thermal equilibrium, a single Fermi level $E_F$ describes the entire system. Under strong injection, the electrons in the CB and holes in the VB relax to thermal equilibrium within their respective bands much faster than they recombine. However, the two bands are not in equilibrium with each other.

We define two Quasi-Fermi Levels:

$E_{F_c}$: Quasi-Fermi level for electrons in the conduction band.
$E_{F_v}$: Quasi-Fermi level for holes in the valence band.

The probability of occupation is given by Fermi-Dirac statistics:

Conduction Band: \[ f_c(E) = \frac{1}{1 + \exp\left(\frac{E - E_{F_c}}{k_B T}\right)} \]
Valence Band: \[ f_v(E) = \frac{1}{1 + \exp\left(\frac{E - E_{F_v}}{k_B T}\right)} \]

4. Microscopic Origin of Optical Gain

Consider the interaction of a photon of energy $h\nu = E_2 - E_1$ with the semiconductor.

4.1 Stimulated Emission Rate

An electron drops from state $E_2$ (CB) to $E_1$ (VB). The rate depends on the probability that $E_2$ is occupied AND $E_1$ is empty.

\[ R_{stim} \propto f_c(E_2) [1 - f_v(E_1)] \rho(\nu) \]

4.2 Stimulated Absorption Rate

An electron is promoted from $E_1$ to $E_2$. The rate depends on $E_1$ being occupied AND $E_2$ being empty.

\[ R_{abs} \propto f_v(E_1) [1 - f_c(E_2)] \rho(\nu) \]

4.3 Net Stimulated Rate

The net rate of photon generation is the difference:

\[ \Delta R \propto R_{stim} - R_{abs} \propto [f_c(E_2) - f_v(E_1)] \rho(\nu) \]

5. Condition for Optical Gain (Bernard-Duraffourg Condition)

For amplification (gain), stimulated emission must exceed absorption ($\Delta R > 0$):

\[ f_c(E_c) > f_v(E_v) \]

Substituting the Fermi-Dirac distributions and simplifying leads to the fundamental condition for transparency and gain:

\[ E_{F_c} - E_{F_v} > h\nu > E_g \]

Physical Interpretation:
The separation of the quasi-Fermi levels must exceed the photon energy. This is the semiconductor equivalent of population inversion ($N_2 > N_1$). It implies that the semiconductor must be degenerately doped or heavily injected so that the bottom of the CB is filled with electrons and the top of the VB is filled with holes.

6. Optical Cavity and Feedback

To sustain oscillation, the gain medium is placed in a resonator.

Fabry-Pérot Cavity: The cleaved facets of the semiconductor crystal act as mirrors due to the high refractive index difference between semiconductor ($n \approx 3.5$) and air ($n=1$).
Reflectivity: Using Fresnel equations, $R \approx (\frac{n-1}{n+1})^2 \approx 0.3$.

This 30% reflectivity is sufficient for lasing due to the extremely high gain of semiconductors.

7. Threshold Condition

Let $g$ be the material gain per unit length. The threshold is reached when the gain compensates for all losses.

\[ g_{th} = \alpha_{int} + \alpha_{mirror} = \alpha_{int} + \frac{1}{2L} \ln\left(\frac{1}{R_1 R_2}\right) \]

The injection current required to reach this gain level is the Threshold Current ($I_{th}$).

8. Carrier Clamping and Output Power

Below Threshold ($I < I_{th}$): Carrier density $N$ increases with current. Spontaneous emission dominates (LED mode).

Above Threshold ($I > I_{th}$): The carrier density clamps (stays constant) at the threshold value $N_{th}$. Any additional injected carriers recombine immediately via stimulated emission, contributing directly to the laser field.

8.1 Power Output Formula

The rate of injected electrons above threshold is $(I - I_{th})/e$. Assuming each electron generates one photon of energy $h\nu$, the ideal power is $P_{ideal} = \frac{h\nu}{e}(I - I_{th})$.

However, not all photons escape the cavity. We define the Differential Quantum Efficiency ($\eta_d$):

\[ \eta_d = \frac{\alpha_{mirror}}{\alpha_{int} + \alpha_{mirror}} \]

The final output power relation is:

\[ P_{out} = \eta_d \frac{h\nu}{e} (I - I_{th}) \]

8.2 Slope Efficiency

The slope of the P-I curve above threshold is constant:

\[ \frac{dP}{dI} = \eta_d \frac{h\nu}{e} \quad [W/A] \]

This linear relationship makes modulation easy, allowing diode lasers to carry data in fiber optic communications.

9. Temperature Dependence of Wavelength

The emission wavelength is determined by the band gap $E_g$. The band gap of semiconductors decreases as temperature increases ($dE_g/dT < 0$).

From $\lambda = hc/E_g$, we find that wavelength increases with temperature:

\[ \frac{d\lambda}{dT} \approx 0.3 \text{ nm}/^\circ C \]

Application Note for Pumping Nd:YAG:
Since the absorption lines of Nd:YAG are narrow (~1-2 nm), the diode laser wavelength must be precisely tuned to the absorption peak (808 nm). This requires precise temperature control of the diode using thermoelectric coolers (TECs).

10. Summary: Why Diode Lasers?

Efficiency: High electrical-to-optical conversion (>50%).
Compactness: Micrometer-scale active regions.
Tunability: Wavelength can be temperature-tuned to match absorption lines.
Direct Modulation: Can be modulated at GHz speeds for communications.

These properties make diode lasers the ideal pump source for solid-state lasers (DPSS systems), replacing inefficient flashlamps.

18nucleons

Search This Blog