Laser Light Characteristics

The process of coherent amplification imbues laser light with a very unique set of characteristics. Typically, only one or a subset of these characteristics are most critical for a particular application, therefore each will be described separately; however, many of these characteristics are interrelated. While not exhaustive, the most common laser output characteristics include: wavelength, gain bandwidth, monochromaticity, spatial and temporal profiles, collimation, output power, coherence and polarization.

A large portion of the electromagnetic radiation spectrum is covered by a wide range of existing lasers. The wavelength range extends from the ultraviolet (UV) to the mid-infrared (MIR) and does not account for other more exotic systems that provide access from the soft-X-ray spectral region (< 10 nm) to the far-infrared (FIR, > 100 µm). The lasing wavelength (or frequency ν₀) is determined by the laser gain medium, which provides the optical transition. The wide range of wavelengths possible is attributable to the large variety of available gain media. Furthermore, nearly all laser wavelengths can be converted or shifted to an alternative wavelength (see Spectral Tunability) and thus can reach from the UV to the MIR spectral region. This spectral agility is the reason laser systems can be employed for short-wavelength applications like lithography for semiconductor processing and long-wavelength applications like material processing and molecular spectroscopy.

The bandwidth of the laser gain medium (B) determines the range of wavelengths over which amplification can occur. This bandwidth is determined primarily by the bandwidth over which spontaneous emission occurs. While various processes contribute to the broadening of the transition linewidth (Δν), several electronic transitions (which are also affected by rotational and vibrational motions) can overlap in frequency, leading to significantly wider bands, particularly for molecular or solid-state systems. Typical bandwidths for select gain media are shown on the right side of Figure 1. Gas lasers, like the HeNe laser, typically have very narrow bandwidths on the order of 1 GHz owing to their atomic transitions. Conversely, solid-state lasers, such as the Ti:Al₂O₃ (sometimes referred to as Ti:Sapphire or Ti:Saph) laser, can have extremely wide bandwidths exceeding 100 THz. The gain bandwidth is also dictated by the total loss in the system (α_r) since a net gain is required for lasing (see Figure 1, left). Consequently, the actual gain bandwidth may be different than the spontaneous emission bandwidth. For instance, modulating the intracavity loss is a means for achieving laser wavelength tuning (see Spectral Tunability). Furthermore, the gain bandwidth is not necessarily the same as the bandwidth of the exiting laser beam since that will also depend on the laser resonator as discussed below.

Monochromaticity refers to color purity or, in the case of a laser, the spectral bandwidth of the laser (sometimes referred to as the laser linewidth). Figure 2 shows how a combination of the laser gain bandwidth and the laser cavity properties determines the bandwidth of the emerging laser beam. Any number of longitudinal modes can lase provided they lie within the window where the gain exceeds the loss. The number of these lasing modes (N) is given by the gain bandwidth divided by the resonator frequency spacing given by Equation (5):

For example, if the entire spontaneous emission spectrum for a Ti:Al₂O₃ laser shown in Figure 11 is available for gain, N can exceed 200,000. This is the basis for implementing a process called mode-locking for generating short-pulse lasers in these types of systems (see Methods for Pulsed-Laser Operation for details). Alternatively, many gas lasers have a sufficiently narrow gain bandwidth for which only a few longitudinal modes are supported. Reducing the cavity length (or equivalently increasing the mode spacing) is a means of achieving lasing with a single longitudinal mode in any laser system. However, this would also place an upper limit on the length of the active medium and therefore limit the achievable gain. Alternatively, techniques exist for introducing frequency-selective elements inside the cavity that allow for single-mode selection without affecting the overall gain. An example is shown in Figure 2 where an etalon, i.e., a type of resonator, of length d₁ is placed inside the laser cavity and only one etalon mode fits within the gain bandwidth. The etalon mode is made to overlap with one of the laser cavity modes and only a single longitudinal mode lases. The spectral width of this mode is controlled by the reflectivity and stability of the cavity which is quantified in the cavity’s quality factor Q. For more details about resonator mode linewidths and cavity Q-factors. Careful cavity designs can enable very narrow laser bandwidths (< 1 MHz) to be achieved. This highly monochromatic output is desirable for applications in remote sensing or as a frequency standard.

Beams that emerge from a laser cavity have an intensity distribution that has both a transverse spatial profile as well as a temporal profile. The spatial profile is mainly determined by the transverse cavity mode and can be rotationally symmetric. Each transverse mode has a different spatial distribution; adjustments to the laser cavity mirrors along with insertion of an aperture inside the resonator can be used to selectively attenuate undesired modes. The lowest-order transverse mode (TEM₀₀), which travels down the central axis of the cavity, is often the desired mode because it propagates with the least beam divergence and can be focused to the tightest spot (see section below on collimation). A laser that operates in the fundamental or TEM₀₀ mode will emit a transverse beam profile described by a Gaussian function. The specifics of this intensity profile as well as its evolution with distance will be described in Laser Beam Spatial Profiles.

Certain lasers operate in continuous wave (or CW) mode where the temporal profile of their output power is constant over time (see section below on output power). Conversely, other lasers operate in pulsed mode such that their output power has a transient temporal profile. These lasers are often characterized by the shape and width of their temporal profiles and, since they typically emit a series of pulses, by their repetition rate, which is given in Hz. Pulsed lasers are useful for many different applications where CW laser properties are not sufficient. There are numerous pulse generation methods (detailed in Methods for Pulsed-Laser Operation) which allow for pulse durations of microseconds (µs), nanoseconds (ns), picoseconds (ps), down to femtoseconds (fs) and below.

Due to diffraction, light emitted from any source will diverge. That is, its transverse spatial profile will get larger as it travels or propagates a greater distance. Laser beams typically possess a much smaller divergence than any other source of light, which is another way of saying that a laser beam is highly collimated. Collimated light possesses photons which are highly directional and propagate parallel to one another. A laser beam’s high degree of collimation arises from the parallelism of the cavity mirrors, which forces the beam to be perpendicular to those mirrors. The TEM₀₀ mode possesses the lowest divergence of all the transverse modes. The propagation of the Gaussian beam associated with this mode is shown in Figure 3. Due to its negligible divergence, highly collimated laser light enables a wide variety of applications, including atmospheric sensing, adaptive optics for astronomical telescopes, and even lunar laser ranging where a beam is propagated to the Moon. Interestingly, a laser’s high degree of directionality can also be used to focus it to a small spot size (see Figure 3). By sending a collimated TEM₀₀ transverse mode through a focusing element such as a lens or curved mirror, the beam diameter can be reduced to a diameter that is approximately the size of the laser wavelength. This ability to tightly focus laser light to a small spot size with a large intensity is exploited for applications in high-resolution microscopy, nonlinear optics, photolithography, and even nuclear fusion.

The average output power of a laser is often quoted in units of milliwatts (mW), watts (W) or kilowatts (kW). For a CW laser whose output power is constant over time, the meaning of this value is straightforward. For a pulsed laser, an average output power can also be defined as the product of the pulse energy and the repetition rate. Since the energy in each pulse is squeezed into a small amount of time, another quantity known as peak output power or peak power is proportional to the pulse energy divided by the width of the pulse. To see how “powerful” a pulsed laser can be, consider a 100 W light bulb whose average and peak powers are necessarily the same. For a 10 ns laser with a pulse energy of 1 Joule (J) operating at 10 Hz, an average power of 10 W is generated, but the laser’s peak power is 1,000,000 kW. CW lasers often have average powers much less than 100 W; however, it is the ability of confining that power into a collimated beam with a relatively small spatial distribution that distinguishes it from other lights sources such as lamps. Thus, quantities that account for the power per unit area (known as the intensity), or that also take into account divergence (a parameter known as radiance), are also important laser characteristics.

During the stimulated emission process, laser photons are cloned such that they possess a fixed phase relation to one another. Coherence refers to the degree to which various portions of a single laser beam are in phase. This coherence is described in terms of both temporal coherence and spatial coherence. Temporal or longitudinal coherence determines how readily different beams can interfere with each other. Interference is a phenomenon that occurs when two waves occupy the same space and are coherent. This can result in either constructive or destructive interference depending on whether the amplitude of the resultant wave is either more or less intense, respectively, than the original waves. When a beam is divided such that each one travels a different distance before recombining, such as in an interferometer (see Figure 4), they will interfere with each other provided they are still coherent. The longitudinal coherence length describes the distance over which beams stay in phase and is inversely dependent on the laser bandwidth described in the monochromaticity section above. This temporal coherence is critical for applications such as holography and general interferometry. Spatial or transverse coherence refers to the phase relationship between different spatial portions of a laser beam after it has propagated a certain distance. The impact of spatial coherence is typically observed as laser speckle (see Figure 4) when laser light is reflected off a rough surface such that the waves of light from each portion of the surface interfere with one another, resulting in a granular pattern called speckle. For more information on temporal and spatial coherence and how these wave properties impact various interference process.

Polarization refers to the direction of oscillation for the transverse waves that make up the electromagnetic fields of light. Linear polarization refers to the oscillation of the field being confined to a single plane perpendicular to the propagation direction. Several laser-based applications require a linearly polarized source, including nonlinear frequency conversion (see Spectral Tunability), certain forms of optical communication, and interferometry. Consequently, it is desirable for a laser to be linearly polarized. While this could be accomplished outside the laser cavity using a polarization-selective component, there are issues with noise arising within the laser cavity from multiple polarizations competing for amplification. A single linearly polarized beam can be generated within the laser cavity if the laser gain medium is polarization dependent, which is the case for some solid-state lasers. An alternative method involves placing a polarization-selective component within the resonator. Figure 5 shows an example where a Brewster window induces additional loss in the polarization perpendicular to the plane of incidence, while the polarization in the plane of incidence experiences no loss.