High-Power Lasers

Abstract

High-power lasers play an important role in modern science, industry, and medicine. A significant milestone was reached on 5 December 2022, when Inertial Confinement Nuclear Fusion (ICF) achieved scientific breakeven, releasing more energy than the input laser energy. Additionally, Extreme Ultraviolet Lithography (EUVL) has enabled the development of microchips with 3 nm process nodes, marking a leap in semiconductor technology. These examples, together with the recent achievement of 10 PW (10¹⁵ W) laser output, herald remarkable advancements in technology and science. Laser systems are broadly classified based on their operating regimes into two main categories: Continuous Wave (CW) operation, where the laser is continuously pumped and emits a steady beam of light, and the pulsed regime, in which the laser produces single or multiple pulses at various repetition rates. This review will primarily focus on pulsed laser systems, exploring their various types and recent technological advancements.

1. Introduction

It is hard to overstate the crucial role that lasers play in modern life. From welding, cutting, and drilling in subtractive manufacturing [1] to 3D printing in additive manufacturing [2], particularly in metal processing, lasers are indispensable. In addition, they accelerate charged particles to high energy [3], produce hard radiation [4] crucial for various scientific and medical applications, and have recently demonstrated breakthroughs such as achieving breakeven in Confinement Nuclear Fusion [5] and enabling Extreme Ultraviolet Lithography [6]. These are just a few examples showcasing the immense capabilities of high-power lasers.

Laser systems are broadly classified based on their operating regimes into two main categories: Continuous Wave (CW) operation, where the laser is continuously pumped and emits a steady beam of light, and the pulsed regime, in which the laser produces single or multiple pulses at various repetition rates.

In the CW regime, high-power fiber lasers are created from active optical fibers pumped by semiconductor laser diodes, a merger between two of the most innovative and advanced laser technologies. They can reach a remarkable output power of up to 100 s kW [7], feature a wide range of operating wavelengths, and record wall-plug efficiency of up to 50% [8]. Multi-kW carbon oxide (CO₂) gas lasers represent another example of greater reliability and higher output consistency, which allowed them to become widespread in industry and medicine [9].

Nevertheless, this paper will primarily focus on ultra-high-power pulsed laser systems, exploring their various types and recent technological advancements. High-power laser systems possess the ability to concentrate energy into extremely small space–time volumes. The output pulse power of the laser beam that recently reached 10 petawatts (10¹⁵ Watt-PW) was focused on a spot of a few microns, which allowed it to achieve one of the highest intensity approaches to 10²⁴ watts per square centimeter (W/cm²) in the observable universe [10]. With a duration of approximately 20 femtoseconds (10⁻¹⁵ s-fs), this pulse occupies a mere ~7 μm of space during its propagation. Such a short pulse duration enables the attainment of tremendous power and intensity with a modest energy input ranging from 300 to 500 Joules. To put this into perspective, consider that the total power output of all of the world’s power plants is around 2 terawatts (10¹² watts-TW), while the sunlight reaching the Earth amounts to approximately 90 PW. Achieving an intensity of 10²⁴ W/cm² is feasible by concentrating this power into a spot with a diameter of just 30 μm.

The light produced by cutting-edge technologies boasts remarkable specifications, revolutionizing various fields of science and industry. Take, for instance, the Extreme Light Infrastructure for Nuclear Physics (ELI-NP) 20 PW laser, crafted from sapphire crystals doped with titanium ions (Ti:Sa). This system generates 600 J of output energy with a pulse duration of 30 fs across two converging channels [10]. In a different league, the NIF Nd: Glass laser system, pivotal in achieving thermonuclear fusion breakeven, wields a staggering 4 MJ of energy at a wavelength of 1.053 μm, converted in 1.8 MJ of third harmonic at 0.35 μm through 192 channels. Its pulse duration spans a nanosecond, with an output power of 0.5 PW [11]. Furthermore, ASML employs a 1 MW CO₂ gas laser within their Extreme Ultraviolet Lithography system, revolutionizing microchip production [12] and showcasing just a glimpse of the transformative capabilities of these cutting-edge light sources.

Figure 1 demonstrates the chart of the historical development of laser systems with high pulse power (intensity). Since Richard Maiman’s groundbreaking invention of the laser in 1960 [13], the power of laser technology has undergone rapid growth, fueled primarily by the discovery of two pivotal operational regimes: Q-Switching and Mode Locking (a detailed explanation is provided below) [14,15].

These breakthroughs, illustrated on the accompanying chart, marked a turning point in laser development. By implementing these regimes in laser oscillators, scientists and engineers achieved a remarkable feat: the gradual reduction of pulse durations from microseconds to nanoseconds, and later, to femtoseconds, all without significant energy loss. The peak power is determined by dividing the energy by the pulse duration, thus there are two primary pathways to increase peak power. The first is extensive, involving increasing the energy output of the laser by developing a Master Oscillator + Power Amplifiers (MOPA) scheme. The second is intensive, focusing on reducing the pulse duration. Intensive improvements, such as pulse duration reduction, are often favored due to their cost-effectiveness. By enhancing the efficiency and precision of laser systems, researchers can achieve higher peak powers without significantly increasing energy consumption or production costs. This approach underscores the importance of refining pulse duration technologies, which not only boost parameters but also promote accessibility in laser applications across various new fields.

Following the period of gradual intensity growth post-1970, a significant technological breakthrough emerged in 1985: Chirped Pulse Amplification (CPA) [17,18]. This innovation provided a substantial boost to laser capabilities, propelling the power growth trajectory forward, a trend that continues to evolve to this day and beyond. The development of Chirped Pulse Amplification (CPA) involved the collective efforts of numerous scientists over the years. While it is true that two individuals, namely Jerard Mourou and Dona Strickland, were awarded the Nobel Prize in 2018 for their contributions to CPA, it is essential to recognize the collaborative nature of scientific progress. Countless researchers worldwide have dedicated their knowledge, expertise, and resources to advancing CPA technology, each playing a vital role in its development and refinement. In this paper, we will delve into a comprehensive explanation of the aforementioned technologies, elucidating their principles, applications, and contributions to the evolution of laser science and industry.

2. Ideas and Methods for Increasing Laser Power

Let us begin with a brief overview of how a laser oscillator operates. For a deeper understanding, readers can refer to [19]. In essence, a laser oscillator comprises three primary components: a resonator or cavity, typically consisting of two mirrors (one of which is usually partially transparent, serving as the output coupler), a Gain Medium (GM) that facilitates the amplification of light within a specific spectral range, and a pump system responsible for energizing the gain medium to enable light amplification (refer to Figure 2a (top) for an illustration).

Some of the spontaneous photons generated by the GM and reflected by the mirrors in the cavity can produce standing waves or resonance frequencies (longitude modes), which are able to exist for a long time in the resonator, for example, as standing waves on the guitar string. The condition for that is the equivalence of the number (q) of half-waves (λ/2) to resonator length (L) multiplied by the index of refraction (n) of the media nL = qλ/2. Similarly, the resonance frequencies can also be expressed as υ = qc/2L, where (c) represents the speed of light.

Initially, the frequency amplitudes are uniform within the empty cavity, depicted in Figure 2a at the bottom, with equidistant spacing calculated by the formula Δυ = c/2L, resulting in an infinite array of frequencies. However, the resonator is not truly empty; it contains a GM capable of amplifying the frequencies suited to its luminescence spectra. In reality, the laser medium (GM) exhibits specific spectral bandwidths for light amplification, determined by its structure and condition, as illustrated in Figure 2b (blue curve). Only a handful of frequencies align with this curve, as other frequencies experience limited or no amplification. Additionally, losses occur within the cavity due to imperfections in optical surfaces, finite apertures, and output energy (R₁, R₂). Consequently, only a select few modes persist within the resonator under steady-state conditions, matching the amplification spectra and surpassing the losses. This condition means that the lasing threshold can be found by the formula R₁R₂ exp(−aL) exp(bL) = 1, where a is the coefficient of losses in the cavity material and b is the coefficient of amplification.

The lasing threshold is shown by the green line in Figure 2b and only five mod amplitudes were bigger than it, meaning that the laser is able to generate only these mods. Therefore, a gain medium with a broader amplification spectrum allows light to be generated with larger spectral bandwidths (more modes). The spectral bandwidth is directly linked to the pulse duration produced by the laser, as dictated by the uncertainty principle. In this context, it can be expressed as Δυ × Δτ = Const, where Δυ and Δτ represent the spectral bandwidth and pulse duration typically measured at the Full Width Half Maximum (FWHM) of their intensities. Consequently, the product constant implies that a larger spectral bandwidth results in a shorter pulse duration. At the same time, it should be noticed that the bandwidth dictates a lower limit of pulse duration in the case of a flat spectral phase, which is called the Fourier transform limit. The actual pulse duration can be longer due to dispersion; therefore, dispersion control is paramount for short pulse generation close to the Fourier transform limit, especially in the case of broadband pulses with few cycles. The time–bandwidth product constant also depends on the temporal/spectral shape which defines its value (~0.44 for a Gaussian FWHM) [22].

In Figure 3a, three spectral widths are depicted alongside their corresponding pulse durations. Notably, the pulse duration of 50 fs requires a much larger spectral width compared to 5 ps (10⁻¹² s). Furthermore, Figure 3b,c illustrates the luminescence spectra of two laser crystals, Nd:YAG and Ti:Sa. Oscillators utilizing these crystals typically generate 4 × 10³ and 3 × 10⁵ longitudinal modes, respectively, resulting in spectral bandwidths of 0.5 nm and 30 nm measured at FWHM.

Now, let us explore the practical methods of receiving laser pulses. The simplest, albeit not the most effective, method involves positioning an electromagnetic or mechanical interrupter within the CW laser output beam. In this scenario, the continuous wave power will match the pulsed power, despite losing energy. However, as evidenced by the chart in Figure 1, following the invention of the laser in 1960, its peak power experienced rapid growth over the subsequent decade, up until 1970. This advancement can be attributed to the development of two key technologies: “Q-switching” [14] and “Mode-locking” [15].

In the first method, unlike the previously discussed approach, the switcher was positioned inside the laser resonator to alter its Q-factor (quality factor), as depicted in Figure 4a (top). The Q-factor in this context can be defined as the ratio of stored energy to energy dissipated per oscillation cycle. Essentially, a high Q- factor corresponds to low energy losses in the cavity, and vice versa. When the Q-switch is closed (low Q-factor), it prevents light propagation within the cavity during the pumping of the GM, as illustrated in Figure 4a (bottom). Once sufficient energy is accumulated, the Q-switch is opened, allowing the light photons spontaneously born in the GM to pass between the resonator mirrors and undergo amplification in the gain medium by stimulating identical photon emission, thereby increasing pulse energy.

The duration of the generated pulses is influenced by several factors, including the resonator round-trip time, the initial gain, and the resonator losses. A greater initial gain accelerates pulse formation, resulting in increased energy extraction and typically a more rapid decay of power after reaching peak pulse intensity. Generally, pulse durations are at least on the scale of the resonator round-trip time and frequently exceed 1 nanosecond (10⁻⁹ s-ns). Q-switch oscillators typically produce pulses with a repetition rate in the range below megahertz, boasting pulse powers up to megawatts (10⁶ Watts-MW).

In the resonator, longitudinal modes exhibit random phases, as depicted in Figure 4b (top). However, laser pulses can be significantly shortened by synchronizing the cavity modes, as illustrated in Figure 4b (bottom). This synchronization, known as the Mod-lock operation regime in laser oscillators, ensures that the maximum amplitudes of all longitudinal modes coincide at the same point and time within the cavity [26]. This constructive summation results in a substantial increase in intensity at this point, manifesting as a pulse with a very short duration, down to a few optical cycles in the best scenario (a few fs in the case of a Ti:Sa oscillator). Furthermore, as more modes concentrate the energy in a smaller area, the pulse duration becomes even shorter (see Figure 3). As the soliton traverses between cavity mirrors, its energy escalates to the nano–microjoule level upon passing through the GM. A significant number of modes or a broad spectral bandwidth necessitates a gain medium with a large fluorescence spectrum, along with devices to compensate for dispersion in optical elements, such as pairs of prisms [27]. In practice, like Q-switching, achieving the Mode-locking regime can be realized through external driving modulators within the laser resonator (active mode locking) [28] or by exploiting nonlinear effects in the cavity elements, such as the Kerr effect or absorption saturation (passive mode locking) [29]. Advancements in these techniques, coupled with the integration of laser oscillators with amplifiers, led to a remarkable increase in output laser pulse power within a decade following the invention of lasers, achieving peak powers of up to terawatts (TW) [30]. However, further increases in power necessitated the significant enlargement of the aperture of optical elements and gain media. This was due to the approaching of these elements to their damage threshold and/or the onset of undesired nonlinear effects, which can distort the laser beam and degrade pulse quality.

An effective method for increasing pulse power involves reducing its duration rather than simply increasing energy, as expressed by the equation P = E/τ, where P is power, E is energy, and τ is pulse duration. However, it is essential to consider the damage threshold, which is proportional to the square root of pulse duration. For instance, when pulse durations approach tens of femtoseconds, the damage threshold can be as low as 100 millijoules per square centimeter (mJ/cm²) of energy fluence. Moreover, severe nonlinear effects such as self-focusing [31] can occur even earlier, typically when energy fluence approaches 1 mJ/cm², with critical power levels reaching gigawatts (10⁹ Watts-GW) in most cases. A direct method to mitigate these undesirable phenomena is by increasing the aperture size, as pointed out above. However, starting from the TW energy level, this approach becomes exceedingly expensive and even impractical. For instance, to achieve PW energy levels, a Ti:Sa crystal with a diameter of 3 m would be necessary for the final amplifier in the laser system.

A more creative approach was proposed back in 1985. Initially, a short pulse with low energy is temporally stretched, resulting in a longer duration suitable for further amplification. This allows for a significantly reduced beam aperture. After accumulating a certain amount of energy in a chain of amplifiers, the long pulse is then compressed back to its original short pulse duration, yielding very high power. This groundbreaking method is known as the Chirped Pulse Amplification (CPA) technique [17,18]. CPA has proven immensely fruitful in the development of ultra-high-power lasers, leading to a gradual increase in power over the years (see Figure 1). Recently, it has even enabled the achievement of power levels of up to 10 petawatts. The general scheme of the CPA laser system is presented in Figure 5a.

Initiating from the Mod-lock oscillator, which produces pulses ranging from 10 s nJ to 1 μJ with durations spanning from a few ps to a few fs, the beam is directed towards a stretcher. This device elongates the pulse duration by several orders of magnitude, extending it up to the nanosecond level. The stretcher comprises two diffraction gratings with an optical telescope positioned between them. This specific configuration of the stretcher, proposed by Oscar Martinez in 1984 [34], has proven highly effective in facilitating the implementation of the CPA method. This optical setup disperses the broad spectrum of the short pulse into multiple narrow harmonics and routes them along different optical paths. Specifically, it arranges for harmonics with longer wavelengths (red) to travel shorter paths compared to harmonics with shorter wavelengths (blue). Upon undergoing double passes through the stretcher, these harmonics reassemble into a single, elongated pulse with what is termed a positive chirp, where the red wavelengths precede the blue ones. In reality, the discretization of the continuous spectrum into separate harmonics was carried out for the sake of simplicity. For a more thorough understanding of the stretcher’s operation, refer to [35]. As demonstrated earlier, the longer pulses possess a higher damage threshold for optics, enabling a significant reduction in the aperture of the optical elements to a reasonable level, even for high-energy applications. These elongated pulses are then directed, following stretching, to a chain of multi-pass amplifiers where they accumulate the required energy. Subsequently, from the final amplifier, the high-energy long pulses are directed to a compressor comprised of either two or four diffraction gratings, as depicted in Figure 5a. The compressor, as described in [36], executes the reverse procedure of the stretcher. It guides the red harmonics with longer wavelengths along longer paths and the blue ones with shorter wavelengths along shorter paths, culminating in the compressor output as a short pulse. Ideally, this pulse equals the initial one but now possesses significantly higher energy, leading to an exceptionally high pulse power.

Each element of this scheme has undergone modifications and improvements since 1985. Here, we will outline the main trends of this technology. The general modifications applied to two key components of the CPA system, namely the stretcher and compressor, primarily involved enhancing pulse stretching [37], improving the quality and size of the gratings [38], and achieving better alignment and conjugation between the stretcher and compressor [39,40,41].

Concerning the oscillator and amplifiers, their initial designs were centered around Nd:glass and Nd:YAG crystals [42,43], which typically exhibited a luminescent spectral bandwidth of 0.5–15 nm FWHM. These oscillators could generate pulses with durations of several hundreds of femtoseconds. However, to achieve the desired output power of 1 PW from these laser systems, amplification to the kilojoule energy level was necessary.

The adoption of Ti:Sapphire crystals [23] as the gain medium for CPA brought about a significant reduction in pulse duration, thanks to their much broader luminescent spectra spanning up to 200 nm, capable of supporting pulse durations as short as 5 fs. Additionally, these crystals possess several unique properties such as high thermal conductivity for effective heat extraction, low thermal expansion, and impressive hardness, rendering them suitable for use in laser oscillators and amplifiers. With the capacity to produce large-aperture crystals above 20 cm [44] in diameter (Figure 6a), they have become the workhorse for amplifiers in CPA laser systems, ultimately facilitating the achievement of the current record output power of 10 PW [10].

However, achieving this extraordinary result demanded numerous technological improvements in laser amplifiers. For instance, generating the energy required for 10 PW pulse power necessitates several hundreds of joules, requiring crystal apertures of tens of centimeters in diameter. Consequently, the gain in the transverse direction far exceeds that in the longitudinal direction, considering the typical crystal thickness of a few centimeters. Thus, extracting stored energy using the amplified pulse (seed) requires a multi-pass optical scheme and takes considerably longer compared to the pulse duration (approximately 100 ns for a 1 ns pulse). A typical scheme of a multi-pass Ti:Sa amplifier is illustrated in Figure 6b.

The initial method of amplification involved full energy crystal pumping before the arrival of the seed pulse, resulting in substantial energy losses due to amplified spontaneous emission (ASE) and/or transverse parasitic generation.

One of the methods employed to mitigate these substantial losses was cladding the side surface of the crystal with a material with a refractive index equal to that of the crystal, thereby preventing light reflection back into the crystal material and minimizing further losses [46]. However, this approach proved insufficient for truly large crystal apertures due to the significant transverse gain. In 2003, a new method was proposed to address this challenge, namely the optimization of timing between crystal pumping and pulse extraction [47,48,49]. This method, known as Extraction During Pumping (EDP), enabled the optimization of the energy required for pumping before and after each seed pulse pass, thereby reducing losses. Typically, a high-energy amplifier crystal is pumped by several laser sources. The temporal pump profile can be adjusted by delaying the arrival of pump pulses from different arms to meet the specific energy requirements of the amplifier. The first time this method was tested was with the Hercules laser system in the Center for Ultrafast Optical Science (CUOS) at the University of Michigan. Figure 7 shows the final amplifier of The Hercules during the shot. The laser system reached the record intensity of 2 × 10²² W/cm² in 2008 [50]. With EDP, it became possible to extract up to 0.6 kJ from large-aperture Ti:Sa crystals with diameters of up to 25 cm [48].

Another type of amplifier utilized in high-power CPA systems is the Optical Parametric Amplifier (OPA) or Optical Parametric Chirped Pulse Amplifier (OPCPA) [51]. Unlike traditional laser amplifiers, OPAs leverage nonlinear interactions between pump and seed pulses within crystals. In the parametric amplification process, the high-energy quantum state of the crystal, excited by pump photons, undergoes decay stimulated by a seed photon (see Figure 8). This decay leads to the emission of another seed photon and an idler photon, facilitated by the virtual energy level between the excited and lower states. All three photons adhere to the laws of energy and momentum conservation, necessitating their efficient amplification within a certain distance of the crystal [52]. To amplify pulses with broad spectral bandwidths, the initially collinear incidence of pump and seed beams on the crystal surface has been replaced with a noncollinear approach (NOPA), enabling the maintenance of synchronization conditions for these broad spectra [53]. NOPAs offer several advantages over traditional laser amplifiers for their applications in CPA systems, such as improved temporal contrast and energetic efficiency. However, OPCPA systems face challenges, including the necessity for high-quality pump pulses and precise synchronization with the seed pulse, which can be considered disadvantages. It is worth noting that the output power of OPCPA systems now approaches that of laser CPAs [54]. Used for OPA, the KDP and DKDP crystals can be grown with large-size clear apertures up to 500 mm in diameter, allowing the amplification of nanosecond stretched broad-bandwidth laser pulses at the kJ energy level. Looking ahead, there are ambitious plans to further increase power output, with projects aiming for up to 50–100 PW, as demonstrated by initiatives such as the OPCPA system development project referenced in [55,56].

Industrial, medical, and numerous scientific applications of CPA lasers necessitate a substantial increase in the pulse Repetition Rate (RR), or in other words, the average power. However, the RR of existing high-peak power laser systems is typically low, often limited to a few Hz or less [58].

The primary obstacle to significantly increasing RR is escalating heat generation in amplifiers, attributed to the quantum defect, which creates an energy difference between pump and seed photons.

The Thin Disc (TD) geometry of laser amplifier active media, such as Yb:YAG, has proven highly effective for heat dissipation, enabling remarkable achievements in laser power. Continuous Wave (CW) regimes have achieved output powers of up to 10 s kW [59], while in master oscillator power amplifier (MOPA) systems, several Joules of energy per pulse and high average power have been attained [60].

The typical scheme of a TD amplifier is shown in Figure 9a. Heat is efficiently and uniformly extracted through the largest face of the crystal, contrasting with conventional side surface heat extraction methods. The Ti:Sa crystal’s exceptionally high thermal conductivity, combined with TD technology, enhances this effect, preventing laser crystals from overheating.

This eliminates thermal distortions in the beam and potential crystal damage, even during operations at extremely high average power levels [62,63,64,65]. Figure 9b,c consequently demonstrates a TD 35 × 3 mm Ti:Sa crystal and holder with both direct contact with large surfaces and water chilling.

However, the TD crystal’s shape represents just one possible geometry for the final amplifier in laser systems, and it may undergo changes depending on the desired output parameters. In one paper [44], simulations were conducted to determine the maximal achievable pump fluence of the amplifier crystal, which was found to depend on the repetition rate (RR). These simulations also highlighted the effect of pump fluences on the selection of crystal shapes.

Figure 10 illustrates the relationship between the required output average power and the repetition rate (RR), delineating preferred areas of usage for different Ti:Sa crystal configurations such as Thin Disc (TD), Rectangular Thin Crystal Plate (RTCP), and Cross Thin Slab (XTS). As shown, the TD crystal configuration is ideal for systems with an RR below a few hundred Hz, particularly when aiming for an output pulse energy exceeding 10 Joules (highlighted in red) [62,63]. In contrast, the RTCP geometry excels in Ti:Sa amplifiers with an RR above 1 kHz and an energy range of 0.1–10 J (indicated in violet). Lastly, the XTS crystal shape emerges as the preferred option for amplifiers at high RR levels (above 1 kHz) with low pulse energy, typically below 100 mJ (shown in blue) [64,65].

The laser systems utilizing the CPA technique have recently achieved remarkable results and continue to show significant potential for further development. Several facilities have been constructed or are currently under construction. One of the pioneering and notable projects in this field is the Extreme Light Infrastructure (ELI).

Under this project, three laser centers were developed in Eastern Europe (see Figure 11): ELI-Beamlines in the Czech Republic [66], ELI-ALPS in Hungary [67], and ELI-NP in Romania [68]. These three pillars of ELI include several petawatt-class lasers: an ELI-ALPS (HF) Ti:Sa system with a peak power of 2 PW and a pulse duration of 17 fs; an ELI-BEAMLINES (L3) Ti:Sa laser with pulse energy of 30 J, 1 PW peak power, and a pulse duration of ~30 fs; ELI-BEAMLINES (L4) Nd:Glass lasers of 2 kJ/10 PW/130 fs; and two-channel ELI-NP Ti:Sa lasers of 300 J/10 PW each.

Two other notable high-power laser systems are the Chinese system with an output power of 10 PW [69] and the South Korean system with an output power of 5 PW. The latter recently achieved the highest intensity of 2 × 10²³ W/cm² [70].

3. Prospects and Conclusions

The next milestone for high-power lasers is reaching 100 PW, with several systems currently under design and development. A Chinese OPCPA system aims to achieve 100 PW by 2030 [71], and the University of Rochester’s system in the USA is expected to reach 50 PW [56].

Several designs incorporating CPA techniques promise to further increase laser power to 10¹⁸ W-exawatt (EW) and 10²¹ W-Zettawatt (ZW) levels. One of them is the post-compression method, where intense pulses from a standard CPA laser undergo spectral broadening via self-phase modulation (SPM) in a nonlinear material, followed by further compression (see Figure 11) [72,73].

This technique, which allows for the compression of femtosecond pulses into near-single-cycle pulses lasting just a few femtoseconds, achieves spectral broadening in a variety of materials, ranging from gases to solids. In 2010, a method for compressing laser pulses with energy levels at the joule scale and higher was both proposed and demonstrated [74,75]. However, SPM in solid materials faces challenges such as low damage thresholds and potential heating problems, limiting improvements in both peak and average output power. In contrast, gas-filled Multi-Pass Cells (MPCs) for spectral broadening via SPM offer a promising solution to substantially mitigate these limitations and are now under intensive development [76,77].

Another approach is an amplification method, weaving the three basic compression techniques, CPA, OPCPA, and Plasma Compression by Backward Raman Amplification (BRA), in plasma. This technology efficiently compresses kilojoule to megajoule nanosecond laser pulses into femtosecond pulses and is referred to as C3, which stands for Cascaded Conversion Compression [78]. The modification of pulse compression in plasma was presented in [79]. The simulation demonstrates that laser pulse compression by a density gradient plasma allows the achievement of exawatt to zettawatt pulse power.

Compression greater than five orders of magnitude from a stretched pulse 20 ns in duration to a 100 fs Fourier transform-limited pulse after propagating through a six-grating compressor arrangement allows the production of exawatt-scale pulses. This paper [80] presents this based on simultaneous proof of this ability by an ideal spatial-temporal pulse structure that requires a chirped beam and chirped pulse amplification in an Nd:Mixed-glass laser system.

As seen from the presented review, laser technologies are experiencing rapid progress, inspired by promising recent applications such as confinement nuclear fusion, extreme ultraviolet lithography, and industrial manufacturing. The deployment of these new devices and systems in industry and medicine has enabled significant advancements in technological capabilities, which can greatly enhance human life. In scientific research, these new systems will facilitate experiments in the new regime of ultra-relativistic optics and even in nonlinear quantum electrodynamics (QED), with laser intensities for matter interaction exceeding 10²⁵ W/cm², leading to a deeper understanding of nature.