Laser Fundamentals

The word LASER is an acronym for Light Amplification by Stimulated Emission of Radiation. A laser produces light that differs fundamentally from all other sources in four ways: it is highly monochromatic (narrow spectral linewidth), spatially coherent (well-defined wavefront), temporally coherent (long coherence length), and highly directional (low divergence) [1, 2]. These four properties arise from a single physical mechanism — stimulated emission — operating inside an optical feedback structure.

Every laser, regardless of type, contains three essential elements: (1) a gain medium that amplifies light through stimulated emission, (2) an energy source (the pump) that creates population inversion in the gain medium, and (3) an optical resonator (cavity) that provides feedback and determines the spatial and spectral properties of the output beam [1, 3]. Understanding each of these elements and how they interact is the foundation of laser physics.

Laser action depends on a quantum-mechanical process — stimulated emission — first described by Einstein in 1917. To understand stimulated emission, we must consider all three radiative processes that occur when light interacts with matter at the atomic level [1, 2].

2.1Three Radiative Processes

When an electromagnetic field interacts with a two-level atomic system having ground state energy $E_1$ and excited state energy $E_2$, three distinct processes can occur.

Absorption. An atom in the lower state absorbs an incoming photon of energy $h\nu = E_2 - E_1$ and transitions to the upper state. The rate of absorption transitions per unit volume is proportional to the population density in the lower state and the spectral energy density of the radiation field:

where $B_{12}$ is the Einstein B coefficient for absorption, $\rho(\nu)$ is the spectral energy density at frequency $\nu$, and $N_1$ is the population density of the lower state [1, 2].

Spontaneous emission. An atom in the excited state decays to the lower state and emits a photon of energy $h\nu$. The emitted photon has a random direction, phase, and polarization. The rate depends only on the upper-state population and is independent of the radiation field:

where $A_{21}$ is the Einstein A coefficient (the spontaneous emission rate, with units of s⁻¹). The reciprocal $\tau_{\text{sp}} = 1/A_{21}$ is the spontaneous emission lifetime [1, 2].

Stimulated emission. An incoming photon of energy $h\nu = E_2 - E_1$ interacts with an atom in the excited state and causes it to emit a second photon that is identical to the first — same frequency, direction, phase, and polarization. This is the process that produces coherent amplification:

where $B_{21}$ is the Einstein B coefficient for stimulated emission. The stimulated photon is a clone of the incident photon, and this cloning process — repeated across billions of atoms — is the origin of the laser beam's coherence and directionality [1, 2].

2.2Einstein Relations

Einstein showed that the A and B coefficients are not independent. By requiring that the rate equations yield the Planck blackbody spectrum at thermal equilibrium, two fundamental relations emerge [1, 2]. The ratio of spontaneous to stimulated emission coefficients is:

This relation reveals a critical frequency dependence: the ratio scales as $\nu^3$, meaning spontaneous emission increasingly dominates over stimulated emission at higher frequencies. This is the fundamental reason why X-ray and gamma-ray lasers are extraordinarily difficult to build [1, 2].

For non-degenerate levels ($g_1 = g_2$), the two B coefficients are equal. When the levels have different degeneracies:

where $g_1$ and $g_2$ are the degeneracies (statistical weights) of the lower and upper states, respectively [1, 2].

2.3Population Inversion

At thermal equilibrium, the population ratio between two energy levels follows the Boltzmann distribution:

At any finite temperature, $N_2 < N_1$ — there are always more atoms in the lower state than the upper state. Under these conditions, absorption dominates over stimulated emission, and the medium attenuates the beam. For net amplification to occur, we need $N_2 > N_1$ (for equal degeneracies), a condition called population inversion. Population inversion is a non-equilibrium state that cannot be achieved by heating; it requires a pump mechanism that selectively excites atoms into the upper laser level faster than they decay [1, 2, 3].

2.4Three-Level and Four-Level Systems

Population inversion cannot be achieved in a simple two-level system because, at best, the absorption and stimulated emission rates equalize at $N_1 = N_2$, producing transparency but no gain. Practical lasers use at least three energy levels [1, 3].

In a three-level system (e.g., ruby laser), atoms are pumped from the ground state (level 1) to a short-lived pump band (level 3), which rapidly decays non-radiatively to the metastable upper laser level (level 2). The laser transition occurs from level 2 back to level 1 (the ground state). Because the lower laser level is the ground state — which is heavily populated at thermal equilibrium — more than half the atoms must be pumped to achieve inversion. This high threshold makes three-level lasers inherently less efficient [1, 3].

In a four-level system (e.g., Nd:YAG), the laser transition terminates on an excited level (level 1) that sits above the ground state (level 0). Level 1 is nearly empty at thermal equilibrium and depopulates rapidly to the ground state via non-radiative relaxation. Population inversion between levels 2 and 1 is therefore achieved with minimal pump power, and four-level lasers typically have much lower thresholds than three-level lasers [1, 3].

The gain medium is the material in which population inversion is established and where amplification of light by stimulated emission occurs. The choice of gain medium determines the laser wavelength, power capability, bandwidth, and much of the system's practical character [1, 3].

3.1Gain Medium Classification

Gain media are classified by their physical state and the nature of the laser transition. The five principal types are:

1. Gas lasers. The gain medium is a gas or gas mixture. Laser transitions occur between discrete atomic or molecular energy levels. Examples include the helium-neon (He-Ne) laser at 632.8 nm, argon-ion laser at 488/514.5 nm, and carbon dioxide (CO₂) laser at 10.6 μm. Gas lasers generally produce excellent beam quality and narrow linewidths, but offer limited power scaling [1, 3].

2. Solid-state lasers. The gain medium is a crystalline or glass host doped with rare-earth or transition-metal ions. Examples include Nd:YAG (1064 nm), Nd:glass, Ti:sapphire (tunable 650–1100 nm), and Er:glass (1550 nm). Solid-state lasers can deliver high power, support short-pulse generation, and are mechanically robust [1, 3].

3. Semiconductor lasers (laser diodes). Population inversion occurs across the band gap of a semiconductor junction. The laser transition is between the conduction and valence bands, with wavelength determined by the band gap energy. Examples include GaAs (808–980 nm), InGaAs (900–1100 nm), and InGaAsP (1300–1550 nm). Semiconductor lasers are compact, electrically efficient, and the most widely produced lasers [1, 4].

4. Fiber lasers. The gain medium is an optical fiber whose core is doped with rare-earth ions (Yb, Er, Tm). The fiber waveguide provides a long interaction length and excellent thermal management due to the high surface-to-volume ratio. Fiber lasers achieve outstanding wall-plug efficiency (>30%) and superb beam quality, and they dominate industrial cutting and welding applications [1, 5].

5. Dye and liquid lasers. Organic dye molecules dissolved in a liquid solvent serve as the gain medium. Dye lasers are continuously tunable over a broad spectral range (typically 30–50 nm per dye) and were historically important for spectroscopy. They have been largely supplanted by tunable solid-state and semiconductor systems [1, 3].

3.2Gain Cross-Section and Small-Signal Gain

The gain cross-section $\sigma(\nu)$ quantifies the probability that a single inverted atom will contribute to stimulated emission at frequency $\nu$. It has dimensions of area (cm²) and is a property of the specific laser transition [1, 2]:

where $g(\nu)$ is the normalized line shape function, $n$ is the refractive index of the gain medium, and $\lambda$ is the wavelength. The small-signal gain coefficient describes the amplification per unit length in the absence of saturation:

where $\Delta N = N_2 - (g_2/g_1)N_1$ is the population inversion density. A beam propagating through the gain medium experiences exponential growth:

This exponential growth is the origin of the "amplification" in the laser acronym. The single-pass gain $G_0 = \exp(\gamma_0\,L)$, where L is the gain medium length, is a key parameter in laser design [1, 2].

3.3Line Broadening Mechanisms

The gain line shape $g(\nu)$ is not infinitely narrow; several physical mechanisms broaden the laser transition [1, 2]:

Homogeneous broadening affects all atoms identically. The primary mechanism is lifetime broadening (also called natural broadening), where the finite excited-state lifetime $\tau$ gives a Lorentzian line shape with full width at half maximum $\Delta\nu_{\text{nat}} = 1/(2\pi\tau)$. In solids and high-pressure gases, phonon interactions and collisions broaden the line further. In a homogeneously broadened medium, all atoms compete for the same photons, and the gain saturates uniformly across the line.

Inhomogeneous broadening arises when different atoms experience different transition frequencies. The dominant mechanism in gas lasers is Doppler broadening: atoms moving at different velocities see different photon frequencies in their rest frame, producing a Gaussian line shape with $\Delta\nu_D \propto \nu_0\sqrt{T/M}$, where T is the temperature and M is the atomic mass. In inhomogeneously broadened media, gain saturation creates spectral holes — narrow dips at specific frequencies — because only atoms with the matching velocity class are depleted [1, 2].

The broadening mechanism has direct implications for laser operation. Homogeneously broadened lasers tend to oscillate on a single longitudinal mode (the mode with highest gain clamps the inversion for all others), while inhomogeneously broadened lasers support multi-mode oscillation because different modes saturate different groups of atoms [1, 2].

The optical resonator (cavity) provides the positive feedback that converts a single-pass amplifier into a laser oscillator. Without a resonator, stimulated emission would amplify light in a single pass but produce neither the spectral purity nor the spatial coherence that define a laser. The resonator selects specific frequencies and spatial patterns, building up the intracavity field to the point where the gain per round trip exactly balances the total losses [1, 2, 3].

4.1The Fabry-Perot Cavity

The simplest laser cavity consists of two mirrors separated by a distance L, forming a Fabry-Perot resonator. One mirror (the high reflector, HR) has reflectivity close to 100%, while the other (the output coupler, OC) is partially transmitting to allow the output beam to escape. The gain medium fills part or all of the space between the mirrors [1, 2].

Light bounces back and forth between the mirrors, passing through the gain medium twice per round trip. On each pass, it experiences gain (stimulated emission) and losses (mirror transmission, scattering, absorption, diffraction). If the round-trip gain exceeds the round-trip loss, the intracavity power builds up exponentially until gain saturation reduces the gain to exactly balance the loss — the steady-state operating point [1, 2].

4.2Longitudinal Modes and Free Spectral Range

Standing waves form in the cavity only when an integer number of half-wavelengths fit between the mirrors. The resonance condition requires:

where q is a large integer (the longitudinal mode number), n is the refractive index inside the cavity, and L is the mirror separation. The frequency spacing between adjacent longitudinal modes is the free spectral range (FSR):

The number of longitudinal modes that oscillate simultaneously depends on the ratio of the gain bandwidth to the FSR. For a gain medium with bandwidth $\Delta\nu_g$, the approximate number of oscillating modes is:

For example, a He-Ne laser with a 1.5 GHz Doppler-broadened gain bandwidth and a 30 cm cavity (FSR = 500 MHz) can support approximately 3 longitudinal modes [1, 2].

4.3Transverse Modes

In addition to longitudinal modes, the resonator supports transverse electromagnetic (TEM) modes that describe the intensity pattern perpendicular to the beam propagation axis. These are designated TEM_mn, where m and n are the mode indices in the two transverse directions [1, 2].

The fundamental mode, TEM₀₀, has a Gaussian intensity profile and is the most desirable for most applications because it produces the smallest focused spot for a given beam diameter. Higher-order modes (TEM₁₀, TEM₂₁, etc.) have larger spatial extent and more complex intensity patterns with nulls. Most laser designs include an intracavity aperture to suppress higher-order transverse modes and force single-transverse-mode (TEM₀₀) operation [1, 2].

4.4Resonator Stability

A resonator is stable if a ray launched at a small angle to the optical axis remains confined after many round trips. The stability condition is expressed in terms of the g-parameters:

where $R_1$ and $R_2$ are the radii of curvature of the two mirrors (positive for concave mirrors). The resonator is stable when:

Resonator configurations that lie at the boundary of stability ($g_1 g_2 = 0$ or $g_1 g_2 = 1$) are marginally stable and highly sensitive to alignment and thermal effects. Named configurations include:

The spatial profile of a laser beam is one of its most important practical characteristics. The fundamental TEM₀₀ mode of a stable resonator produces a Gaussian beam — a beam whose transverse intensity profile is a Gaussian function at every plane along the propagation axis. Gaussian beam optics provides the mathematical framework for predicting how the beam propagates, focuses, and diverges [1, 2, 6].

5.1The Gaussian Beam

A Gaussian beam is fully characterized by its wavelength and a single parameter: the beam waist radius $w_0$, defined as the radius at which the field amplitude drops to $1/e$ of its on-axis value (equivalently, the intensity drops to $1/e^2$). The beam radius at a distance z from the waist is:

where $z_R$ is the Rayleigh range — the distance from the waist at which the beam area has doubled (the beam radius has increased by a factor of $\sqrt{2}$):

In the far field ($z \gg z_R$), the beam expands linearly with a half-angle divergence:

This inverse relationship between waist size and divergence is a direct consequence of diffraction: a tighter waist diffracts more strongly and diverges faster. The Rayleigh range defines the boundary between the near field (quasi-collimated) and the far field (diverging) [1, 2, 6].

5.2Beam Parameter Product and M²

The beam parameter product (BPP) is the product of the beam waist radius and the far-field half-angle divergence:

For an ideal Gaussian beam, the BPP takes its minimum possible value:

Real laser beams deviate from the ideal Gaussian. The beam quality factor $M^2$ (pronounced "M-squared") quantifies this deviation:

By definition, $M^2 = 1$ for an ideal Gaussian beam. Real beams have $M^2 > 1$. Single-mode fiber lasers typically achieve $M^2 < 1.1$, while high-power multimode industrial lasers may have $M^2 > 10$. A real beam with quality factor $M^2$ propagates as:

The beam quality factor is conserved through ideal optical systems (lenses, mirrors) and can only degrade through aberrations, scattering, or gain non-uniformities [1, 2, 6].

5.3Intensity Profile

The intensity distribution of a Gaussian beam at any plane z is:

where $I_0 = 2P/(\pi w_0^2)$ is the peak on-axis intensity at the waist for a beam of total power P. At the beam radius $r = w(z)$, the intensity has fallen to $1/e^2 \approx 13.5\%$ of its peak value. The diameter $2w$ therefore contains approximately 86.5% of the total beam power [1, 2].

The transition from spontaneous emission to laser oscillation occurs at a well-defined threshold. Below threshold, the gain medium amplifies spontaneous emission noise but the losses prevent sustained oscillation. At threshold, the round-trip gain exactly equals the round-trip loss, and above threshold the output power grows linearly with pump power [1, 2].

6.1The Threshold Condition

For a Fabry-Perot cavity with mirrors of reflectivity $R_1$ and $R_2$, gain medium length $l_g$, and distributed internal loss coefficient $\alpha_i$ (per unit length), the round-trip condition for threshold is:

Solving for the threshold gain coefficient:

The second term represents the mirror loss — the fraction of the intracavity power that escapes through the mirrors (or is absorbed by them) per unit length of gain medium per pass. The threshold gain must compensate both the internal losses and the useful output coupling. Reducing the cavity losses or increasing the gain medium length lowers the threshold [1, 2].

6.2Output Power and Slope Efficiency

Above threshold, the output power through the output coupler increases linearly with pump power:

where $\eta_s$ is the slope efficiency and $P_{\text{th}}$ is the threshold pump power. The slope efficiency can be decomposed into several contributing factors:

where $\eta_p$ is the pump efficiency (fraction of pump photons absorbed), $\eta_q = \lambda_p/\lambda_l$ is the quantum defect efficiency (ratio of pump to laser photon energy), $\eta_a$ is the mode overlap efficiency, and $\eta_c$ is the output coupling efficiency [1, 2, 5].

The overall wall-plug efficiency relates the optical output to the total electrical input:

Modern fiber lasers achieve wall-plug efficiencies exceeding 30%, semiconductor lasers can exceed 50%, while gas lasers are typically below 1%. The wall-plug efficiency determines cooling requirements, operating costs, and the system's overall energy footprint [1, 5].

Laser light is distinguished from all other sources by its extraordinary spectral purity and coherence. The linewidth of a laser — the spectral width of its output — determines its coherence properties and is critical for applications ranging from interferometry to spectroscopy to telecommunications [1, 2].

7.1Linewidth Fundamentals

The linewidth of a single longitudinal mode is determined by the cavity losses. The passive cavity linewidth (without gain) is:

where $\mathcal{F}$ is the cavity finesse, defined as the ratio of the FSR to the linewidth of the cavity resonance. For mirrors of reflectivity R, $\mathcal{F} \approx \pi\sqrt{R}/(1-R)$. A cavity with R = 99% has a finesse of about 313 [1, 2].

In an operating laser, the fundamental linewidth limit is set by spontaneous emission into the laser mode. The Schawlow-Townes linewidth is:

This remarkable result shows that the fundamental linewidth decreases with increasing output power — more photons in the cavity mode suppress the phase fluctuations caused by spontaneous emission. For typical visible lasers, the Schawlow-Townes limit is well below 1 Hz, but in practice, technical noise (mechanical vibration, temperature fluctuations, pump noise) broadens the linewidth to kHz–MHz levels [1, 2, 7].

7.2Single-Mode Operation

Many applications require the laser to oscillate on a single longitudinal mode. Three principal techniques achieve this:

1. Short cavity. If the cavity is made short enough that the FSR exceeds the gain bandwidth, only one longitudinal mode falls within the gain curve. This is the approach used in microchip lasers and vertical-cavity surface-emitting lasers (VCSELs) [1, 4].

2. Intracavity etalon. A thin etalon (a parallel-sided plate with partially reflecting surfaces) placed inside the cavity introduces additional frequency selectivity. The etalon's transmission peaks can be adjusted to select a single longitudinal mode from the cavity [1, 2].

3. Distributed feedback (DFB). A periodic refractive index variation (Bragg grating) is embedded in or adjacent to the gain medium. The grating provides wavelength-selective feedback with a bandwidth narrower than the mode spacing, enforcing single-mode operation. DFB lasers are standard in telecommunications [1, 4].

7.3Coherence Length

The temporal coherence of a laser is characterized by its coherence length — the maximum path difference over which the beam can produce interference fringes with good visibility. For a source with Lorentzian line shape of linewidth $\Delta\nu$:

For a Gaussian line shape:

The coherence length spans an enormous range across laser types: a multimode He-Ne laser has $L_c \sim 20\;\text{cm}$, a single-mode He-Ne reaches $L_c \sim 300\;\text{m}$, and a frequency-stabilized narrow-linewidth laser can achieve $L_c > 100\;\text{km}$ [1, 2].

Lasers are most commonly classified by their gain medium, which determines the fundamental wavelength, power scaling capability, efficiency, and practical operating characteristics. Each major laser family has a distinct set of strengths that suit it to particular applications [1, 3, 4, 5].

8.1Major Laser Families

Gas lasers (He-Ne, Ar⁺, CO₂, excimer) operate on transitions between discrete energy levels of atoms, ions, or molecules in the gas phase. They are characterized by excellent beam quality (low divergence, often TEM₀₀), narrow linewidth, and long coherence length. The He-Ne at 632.8 nm remains a standard for alignment and metrology. The CO₂ laser at 10.6 μm delivers high CW power (up to multi-kW) for industrial cutting of non-metals. Excimer lasers (ArF at 193 nm, KrF at 248 nm) produce high-energy UV pulses for photolithography, corneal surgery (LASIK), and micromachining [1, 3].

Solid-state lasers (Nd:YAG, Nd:YVO₄, Ti:sapphire, Er:YAG) use rare-earth or transition-metal ions doped into a crystalline or glass host. Nd:YAG at 1064 nm is the workhorse of industrial and scientific laser systems — it can be Q-switched to produce nanosecond pulses and frequency-converted to 532 nm (green), 355 nm (UV), or 266 nm (deep UV). Ti:sapphire provides broad tunability (650–1100 nm) and supports mode-locked pulses as short as a few femtoseconds, making it essential for ultrafast spectroscopy and attosecond science [1, 3].

Semiconductor lasers (laser diodes, VCSELs, quantum cascade lasers) convert electrical current directly into coherent light with wall-plug efficiencies exceeding 50%. They span the spectrum from the UV (GaN, ~405 nm) through the visible and near-IR to the mid-IR (quantum cascade lasers, 3–12 μm). Laser diodes are by far the highest-volume lasers manufactured, used in telecommunications, optical storage, barcode scanners, laser printers, fiber-optic pump sources, and LIDAR [1, 4].

Fiber lasers (Yb-doped, Er-doped, Tm-doped) confine the gain medium within a doped optical fiber, providing an extremely long interaction length and excellent thermal management. Yb-doped fiber lasers at 1.0–1.1 μm dominate industrial metal cutting and welding at power levels from watts to tens of kilowatts with superb beam quality ($M^2 < 1.1$). Er-doped fiber amplifiers (EDFAs) at 1550 nm are the enabling technology of long-haul fiber-optic telecommunications [1, 5].

Dye and other liquid lasers use fluorescent organic molecules in solution. Their chief advantage is continuous tunability over a broad spectral range (typically 30–50 nm per dye, spanning the visible and near-IR). Though historically important for spectroscopy and photochemistry, dye lasers have been largely replaced by tunable solid-state systems (Ti:sapphire, OPOs) that offer superior convenience and stability [1, 3].

Lasers can operate in continuous-wave (CW) mode or produce pulses spanning durations from milliseconds down to attoseconds. The operating regime profoundly affects the peak power, spectral characteristics, and suitability for specific applications [1, 2, 3].

9.1Continuous-Wave Operation

In CW operation, the laser output is constant in time (steady-state). The pump continuously maintains the population inversion, and the intracavity field reaches a stable equilibrium where the round-trip gain exactly equals the loss. CW lasers are used when a steady beam is needed: materials processing (cutting, welding), optical trapping, spectroscopy, telecommunications, and laser pointers. CW output powers range from microwatts (single-mode diodes) to hundreds of kilowatts (industrial fiber lasers) [1, 3].

9.2Pulsed Operation and Key Definitions

Pulsed lasers concentrate energy into short bursts, achieving peak powers far exceeding the CW capability of the same gain medium. Key relationships between pulse parameters are:

where $E_p$ is the energy per pulse, $P_{\text{avg}}$ is the average power, $f_{\text{rep}}$ is the pulse repetition rate, $\tau_p$ is the pulse duration, and D is the duty cycle. A laser with 10 W average power at 10 kHz repetition rate delivers 1 mJ per pulse; if the pulse duration is 10 ns, the peak power is 100 kW — four orders of magnitude above the average power [1, 2].

9.3Q-Switching

Q-switching produces pulses in the nanosecond range (typically 1–100 ns) by modulating the cavity quality factor (Q). While the Q is held low (by an intracavity shutter, electro-optic modulator, or saturable absorber), the pump builds up a large population inversion without laser oscillation. When the Q is suddenly switched high, the stored energy is released in an intense, short burst. Q-switched Nd:YAG lasers routinely deliver millijoule to joule pulses with peak powers of megawatts to gigawatts [1, 2, 3].

Q-switching is widely used for laser ranging (LIDAR), material marking, laser-induced breakdown spectroscopy (LIBS), and nonlinear frequency conversion (harmonic generation, optical parametric oscillation) [1, 3].

9.4Mode-Locking

Mode-locking produces ultrashort pulses (picoseconds to femtoseconds) by locking the phases of many longitudinal modes so they interfere constructively at one point in the cavity, forming a circulating pulse. The minimum achievable pulse duration is set by the transform limit:

where $\Delta\nu_g$ is the gain bandwidth and K is a numerical constant that depends on the pulse shape (K = 0.44 for Gaussian pulses, K = 0.315 for sech² pulses). For Ti:sapphire with a gain bandwidth of ~100 THz, the transform limit is about 4.4 fs — pulses containing only a few optical cycles [1, 2, 7].

Mode-locking is achieved by active means (an intracavity modulator driven at the cavity round-trip frequency) or passive means (a saturable absorber, Kerr lens, or semiconductor saturable absorber mirror — SESAM). Mode-locked lasers are essential tools for ultrafast spectroscopy, frequency comb generation, multiphoton microscopy, and attosecond pulse generation [1, 2, 7].

9.5Gain-Switching

Gain-switching produces short pulses by modulating the pump source itself (typically the drive current of a laser diode). When the pump is turned on abruptly, the gain builds rapidly, overshoots the threshold, and produces a relaxation oscillation pulse before the system can reach steady state. The pulse terminates when the population inversion is depleted below threshold. Gain-switched pulses are typically 10–100 ps in duration for semiconductor lasers. The technique is simple and cost-effective, but pulses are longer and have more timing jitter than mode-locked pulses. Gain-switching is used in telecommunications, LIDAR, and seeding of amplifier chains [1, 4].

Selecting the right laser for an application requires balancing multiple interrelated parameters. No single laser type excels at everything; each represents a set of trade-offs between wavelength, power, beam quality, coherence, cost, size, and reliability [1, 3, 5].

10.1Wavelength Selection

The wavelength must be matched to the application physics. Material absorption spectra determine the optimum wavelength for materials processing — metals absorb efficiently at 1 μm (fiber lasers), while plastics and organics absorb strongly at 10.6 μm (CO₂). Biological tissue absorption, water absorption windows, and atmospheric transmission bands constrain medical, sensing, and free-space communication applications. Detector availability and sensitivity are also wavelength-dependent [1, 3].

10.2Power and Energy Requirements

CW applications are specified by average power; pulsed applications by pulse energy and/or peak power. Thin-film deposition requires milliwatts; industrial cutting requires kilowatts; laser fusion experiments require megajoules in nanosecond pulses. The pulse duration, repetition rate, and average power must all be considered together — the same average power can represent vastly different peak powers depending on the temporal format [1, 3].

10.3Beam Quality Requirements

Applications requiring tight focus (micromachining, laser surgery, single-mode fiber coupling) demand low $M^2$ (ideally $M^2 < 1.3$). Applications where the beam illuminates a broad area (heat treatment, display projection) can tolerate much higher $M^2$. Note that beam quality and power are often in tension — higher power typically degrades beam quality due to thermal lensing and multimode operation [1, 2, 5].

10.4Coherence Requirements

Interferometry, holography, and coherent LIDAR require long coherence length (narrow linewidth, single longitudinal mode). Spectroscopy requires narrow linewidth for resolving closely spaced transitions. Materials processing and most medical applications have no significant coherence requirement, and a broader linewidth may actually be preferable to avoid speckle and interference artifacts [1, 2].

10.5Environmental and Cost Factors

Practical deployment considerations include operating temperature range, vibration tolerance, cooling requirements (air vs. water), electrical power consumption (wall-plug efficiency), physical size and weight, maintenance interval (lamp/diode lifetime, gas replenishment), and total cost of ownership. Fiber and semiconductor lasers excel in most of these criteria, which drives their growing dominance across applications [1, 5].

10.6Selection Workflow

A systematic laser selection process follows five steps:

Step 1: Define the wavelength. Identify the required wavelength or wavelength range from the application physics (material absorption, detector response, atmospheric window, safety classification).

Step 2: Determine the temporal format. Decide whether CW, pulsed, Q-switched, or mode-locked operation is needed, and specify the pulse duration, repetition rate, and energy requirements.

Step 3: Specify power and beam quality. Define the average power (or pulse energy), peak power, and $M^2$ requirements based on the application geometry (spot size, working distance, depth of focus).

Step 4: Assess coherence and linewidth. Determine whether single-mode operation, narrow linewidth, or long coherence length is required.

Step 5: Evaluate practical constraints. Consider size, weight, cooling, power consumption, reliability, maintenance, and cost. Select the laser type that satisfies all technical requirements at the lowest total cost of ownership.