Recent progress on femtosecond laser micro-/nano-fabrication of functional photonic structures in dielectric crystals: A brief review and perspective

The last decade has witnessed the tremendous development of femtosecond (fs) laser precision engineering, which has revolutionized material processing/functionalization in a broad spectrum of areas associated with academic research and engineering.^1–10 Nonthermal energy deposition (as a result of the ultrashort timescale of laser pulses and thus optical energy transfer) and three-dimensional (3D) nanofabrication (due to efficient confinement of nonlinear interactions within the focal volume) are the major keys to the great success of fs laser material processing, enabling high-precision, high-quality micro-/nano-fabrication applicable to a wide range of materials, including metals, semiconductors, and dielectrics.¹ With a view to the multi-functionalization of photonic structures/devices/systems, dielectric crystals stand out because of their diverse crystalline structures and related optical properties, offering multi-functional material platforms promising for various classic and quantum photonic applications.

Fabrication of optical waveguides is usually the first thought when it comes to linking a micro-/nano-processing technology to designing functional photonic devices/systems. In the last decades, fs laser micro-/nano-fabrication has been well-developed for the direct writing of channelized waveguide architectures with flexible geometries.^11–17 Pioneering research on fs-laser direct writing of waveguides in dielectric crystals is carried out first in lithium niobate (LiNbO₃)^18,19 and titanium-doped sapphire (Ti:Sapphire),²⁰ which are important media for photonic applications in electro-optic modulation and lasers, respectively. Since then, fs laser micro-/nano-fabrication has immediately delivered stunning performance in defining 3D crystalline objects with a footprint compatible with lab-on-a-chip applications, opening up great opportunities and possibilities in integrated photonics based on dielectric crystals. Recent progress on fs-laser-defined crystalline waveguides and their photonic applications can be referenced elsewhere.^11–14 

Beyond waveguides, diverse functional structures with 2D and 3D geometries based on dielectric crystals have been demonstrated by fs laser micro-/nano-fabrication.^21–25 According to the used laser parameters on different lattice structures, the laser-induced localized features can either modify/enhance the original optical properties or introduce additional photonic functionalities within the target crystals at desired positions, largely enriching the potential application domains of crystalline nanostructures. This article aims to present a brief overview of the most recent research progress on fs laser micro-/nano-fabrication of functional photonic structures in terms of domain engineering for nonlinear optics (Sec. II), color centers and rare-earth-doped waveguides for integrated quantum photonics (Sec. III), and surface processing for integrated photonics (Sec. IV) in dielectric crystals. Before moving forward into the conclusion/outlook section, the impact of optical aberration during laser fabrication is discussed (Sec. V). In Sec. VI, we discuss the remaining challenges and potential opportunities on the above-mentioned topics and provide an outlook.

Precise engineering of ferroelectric domain structures in nonlinear optical crystals, such as LiNbO₃, in a facile manner is highly desired for high-efficiency frequency conversion based on quasi-phase-matching (QPM).^26–30 Conventional domain engineering techniques rely on electric-field poling (EFP), which applies a large external electric field (exceeding the coercive field E_c along the spontaneous polarization direction) via patterned electrodes. While EFP has been researched and well-developed for decades, its practicability in terms of defining advanced nanophotonic components/platforms with flexible geometries has, however, encountered bottlenecks since EFP domain structures are limited to 2D patterns with a resolution of only a micrometer.

In contrast to the electric field, the light field can be manipulated more accurately with a resolution up to the diffraction limit. This is the main reason why all-optical poling (AOP) using focused fs laser has become particularly interesting in recent years.^{23,24,31–34} Pioneering AOP studies use focused ultraviolet (UV) fs laser beam with a view to inducing a modulation of the E_c or a pyroelectric/thermoelectric field.^31,35 However, due to the strong linear absorption of UV light by ferroelectrics, the inverted domain size is restricted to submicrons. In this regard, near-infrared (NIR) AOP is considered to be more promising since ferroelectrics are generally transparent in the NIR spectral region, and thus on-demand domain length can be flexibly defined.^36–40 In this process, the focused NIR pulses cause local heating by multiple photon absorption, leading to local domain inversion by inducing an internal field that exceeds the reduced E_c at the laser focus [see Figs. 1(a) and 1(b)]. By further combining with proper thermal treatment, focused NIR fs laser pulses are able to define domain structures in LiNbO₃ crystals with a high length/width aspect ratio of 800/1 µm.⁴⁰ 

In terms of AOP-based 3D domain engineering, it has been a great challenge to directly pole the domains inside a dielectric crystal for decades. In 2018, two breakthrough studies first developed 3D domain/nonlinearity engineering in nonlinear optical crystals, namely LiNbO₃³² and barium calcium titanate (Ba_0.77Ca_0.23TiO₃)³³ (BCT, which has a much lower E_c than that of LiNbO₃), using NIR fs laser pulses, opening up new possibilities and opportunities for fabrication of nonlinear photonic crystals (NPCs) with 3D configurations. In these two separate studies, the AOP parameters are optimized to either induce localized domain inversion via exceeding the E_c [i.e., χ⁽²⁾-inversion approach] of BCT or to selectively erase the nonlinearity [i.e., χ⁽²⁾-erasure approach] of LiNbO₃, enabling NPCs with desired 3D χ⁽²⁾ nonlinearity patterns for nonlinear frequency conversion; see the Čerenkov-type^41,42 and general second harmonic (SH) confocal microscopic results in Figs. 1(c) and 1(d). It is worth noting that the laser-induced χ⁽²⁾-erasure and χ⁽²⁾-inversion approaches rely on different physical mechanisms, i.e., crystallinity reduction and light-induced electric field poling, respectively; their optimized experimental parameters and applicable material platforms are, therefore, quite different.²³ Interestingly, by further systematically optimizing NIR laser parameters, both χ⁽²⁾-erasure and χ⁽²⁾-inversion effects can be realized in LiNbO₃, depending on the laser-writing direction, as recently demonstrated [see Fig. 2(b)].²⁴ Briefly, when the laser beam moves along the -z direction of LiNbO₃, inverted domain structures can be defined. While the laser-writing direction is along +z, which is the original spontaneous polarization orientation of LiNbO₃, no domain inversion can be identified. Without adjusting the laser writing parameters, furthermore, the already inverted domain structures can be effectively erased by the +z-direction laser writing. The origin of such a nonreciprocal feature is considered to be the different orientations of laser-induced localized thermoelectric-field components and thus different electric field-LiNbO₃-domain interactions [see Fig. 2(a)], offering functional tools for the fabrication of arbitrary 3D domain structures with extremely fine features [see Fig. 2(c)]. For example, the linewidths of fabricated domain structures can be continuously varied by adjusting the laser-writing direction to a certain angle to the −z orientation, offering a reconfigurable way to fabricate 3D nanosized domain structures with a resolution [30 nm as experimentally demonstrated, see Fig. 2(c)] far beyond the optical diffraction limit.²⁴ We believe this milestone study has inspired a new wave of research on high-quality 3D integrated nanophotonic devices for advanced nonlinear optical and optoelectronic applications.

Integrated quantum photonics, through integrating different components with complementary functionalities on a single chip, offers a particularly desirable platform for generation, manipulation, and detection of quantum states of light, featuring functional scalability, stability, and integrability in contrast to traditional photonic quantum systems that involve “bulk optics” technology and, in the meantime, revolutionizing the field of quantum information in applications from communications to computing.^43–47 In this new wave of quantum revolution, fs laser micro-/nano-fabrication plays an important role in terms of the definition of photonic structures with flexible geometries and diverse functionalities.^46,48 A typical example of crystalline devices is non-degenerate photon pair generation using fs-laser-written waveguides in periodically poled LiNbO₃ (PPLN), which, by the way, can be realized in the future by employing fs laser domain engineering as discussed in Sec. II, based on spontaneous parametric down conversion (SPDC).⁴⁹ In waveguide structures, such a nonlinear optical effect can be well enhanced, and the generated photons can be flexibly routed in a facile manner, offering integrated solutions to single photon sources. Beyond this, here we focus our discussion on laser-induced color centers in diamonds and laser-written waveguides in rare-earth-doped crystals, which are promising integrated solutions for applications in single photon sources and quantum memories, respectively.

Single negatively charged nitrogen-vacancy (NV⁻) centers, i.e., optically active point defects that form in diamonds by the binding of a lattice vacancy with a substitutional nitrogen impurity, have shown enormous potential in quantum applications, such as quantum information and sensing, due to their long spin coherence times (≈1 ms) even at room temperature.^50–53 Fabrication of NV⁻ centers in diamonds is never an easy task because of the diamond’s extreme hardness and chemical resistance. Conventionally, this can be realized by pure physical micromachining techniques, such as energetic ion and electron beam irradiation methods, followed by post-thermal annealing treatment to stimulate diffusion of vacancies through the lattice, whereupon random vacancy-impurity binding and thus an NV⁻ center occur.^54,55 However, ion/electron beam irradiation processes inevitably introduce residual subsurface lattice damage and degrade the color center quality.

The 3D precision micro/nano-machining nature of fs laser nanofabrication renders it an ideal solution to high-quality NV⁻ centers creation in a controlled way.^{22,46,56–62} In contrast to ion/electron beam irradiation, focused fs laser is able to fabricate localized lattice vacancies at any depth in the diamond without damaging the overlaying material. By further combining it with appropriate aberration correction, the placement of lattice vacancies, or NV⁻ centers, can be precisely controlled with a spatial resolution beyond the optical diffraction limit. Furthermore, desired NV⁻ center density, from single color centers to high-density ensembles, can also be well controlled by adjusting the laser parameters. Pioneering experimental demonstration of fs laser writing of NV⁻ centers is realized in diamond thin film with nitrogen concentration <5 ppb.⁵⁸ In the laser fabrication process, a spatial light modulator is used for aberration correction, enabling a spatial resolution of 120/500 nm (in-plane/-depth resolution). In order to facilitate the random diffusion of vacancies and subsequent NV⁻ centers formation in the lattice, post-thermal annealing treatment is carried out, boosting the NV⁻ centers formation probability up to 45% ± 15%, which is consistent with the Poisson distribution optimum of 37%.⁵⁸ A further step is taken in a subsequent study by using fs laser to introduce not only localized vacancy ensembles (using the “seed single pulse” with an energy of 27 nJ) but also localized thermal annealing (using the “annealing pulse train” with an energy of 19 nJ) for vacancy diffusion [see Fig. 3(a)], enabling near-deterministic (up to 96%) NV⁻ centers formation with a positioning accuracy of around 33/200 nm (in-plane/-depth accuracy).⁶² The near-unity yield therein is reached by using an in-line confocal microscope for online fluorescence measurement, which provides feedback to allow active control of the fabrication process [see Figs. 3(b) and 3(c)]. The strategy of fs-laser induced color centers can be straightforwardly extended to other material families, such as hexagonal boron nitride, silicon carbide, and gallium nitride, providing a promising tool for the optical fabrication of engineered materials and devices for quantum applications.

Cryogenically cooled rare-earth-doped crystals feature exceptionally high coherence times offered by the dopant ions and a considerably large bandwidth due to the inhomogeneous broadening of the optical transitions of interest. They are, therefore, very promising as excellent solid-state platforms for the development of quantum storage devices compatible with integrated photonics.^16,63–65 For example, rare-earth-doped yttrium silicate (RE:Y₂SiO₅) crystals have been demonstrated to offer coherence times exceeding six hours and are widely used for efficient quantum information storage.^64,66

As one of the most promising platforms for integrated quantum memories, RE:Y₂SiO₅ waveguides fabricated by femtosecond laser direct writing (FsLDW) have been experimentally demonstrated to be an on-demand storage strategy since their pioneering work in 2016.^46,67 Based on the enhanced light–matter interaction in FsLDW fabricated Pr³⁺:Y₂SiO₅ double-line waveguides therein and single-line waveguides in the subsequent studies,^67–69 quantum memory performance [with a storage time of up to 15 µs⁶⁷ and a storage efficiency as high as 21%,⁶⁸ see Figs. 4(a) and 4(b)] surpassing all previously reported single-photon storage in waveguides has been demonstrated. Moreover, storage fidelity characterizations of FsLDW waveguides have been recently carried out in ¹⁵¹Eu³⁺:Y₂SiO₅ surface ridge waveguides and depressed cladding waveguides, demonstrating reliable coherent optical memories [featuring an ultrahigh qubit storage fidelity of >99%, see Fig. 4(c)] with an on-demand retrieval capability.^70–74 Furthermore, at telecom wavelengths, a high storage fidelity of >98% has been achieved recently in an ¹⁶⁷Er³⁺:Y₂SiO₅ depressed cladding waveguide with both ends directly connected with fiber arrays [see Fig. 4(d)].⁷⁵ Therein, the fiber-like geometry of the FsLDW depressed cladding waveguide enables such a fiber-integrated design, which shows great potential for applications in fiber-based quantum networks.

In addition to defining micro-/nano-photonic structures buried in the substrates, fs laser pulses are also capable of processing/structuring dielectric surfaces via similar nonlinear absorption processes to, e.g., waveguide fabrication. In this section, research on laser-induced periodic surface structures and selective etching is briefly reviewed.

With the tremendous development of material science and nanofabrication technology in the last decades, periodically nanostructured surfaces, e.g., plasmonic and dielectric metasurfaces, made of artificial subwavelength elements have been widely studied in advanced photonics technology and applications.^76–83 Due to their powerful capability to efficiently manipulate light, tailored photonic functionalities, such as structure colors, data storage, enhanced absorption, and luminescence, based on conventional and emerging materials can be realized, holding great promise for enriching the designer’s toolbox with a view to addressing key performance features not available in existing planar photonics.^77,81–83 For processing such periodic surface structures (PSSs), lithographic techniques, such as photolithography and electron beam lithography, followed by chemical or reactive ion beam etching and focused ion beam milling, are the commonly used methods. However, these nanofabrication techniques often call for complex and expensive apparatus, and the whole processing flow is generally quite complicated and time-consuming.

Alternatively, focused fs laser offers a powerful solution for direct preparation of PSSs within a single-step process, which is furthermore, widely applicable to a variety of materials, including metals, glasses, crystals, and polymers.^{4,13,25,84–92} In contrast to lithographic methods, laser-induced periodic surface structure (LIPSS) provides a mask-free, high-efficiency, and high-flexibility processing method for defining subwavelength features. Depending on the period/wavelength (Λ/λ) ratio, the LIPSSs can be divided into low- (Λ/λ > 0.5) and high-spatial-frequency LIPPSs (Λ/λ < 0.5).^25,84 In addition, different Λ/λ values as well as periodic nanostructure features can be realized by adjusting fs laser parameters. Since it was first demonstrated on semiconductor surfaces, a great amount of effort has been put into elucidating the formation mechanism of LIPSSs.^25,92 Several models concerning, e.g., scattering light field,^93–97 propagating surface plasmon polariton,^87,98,99 and nanoplasmonic near-field enhancement⁸⁶ have been proposed, but a clear and precise physics picture of the LIPSS formation has not yet been finalized.²⁵ Nevertheless, promising optoelectronics applications, such as vivid structural colors,^100–104 optical birefringence based multi-dimensional data storage,^105–107 enhanced optical absorption (e.g., “black” silicon),^100,108 and improved electrical performance,^109,110 have been enabled by LIPSSs based on different material systems.²⁵ Up until now, demonstrations of LIPSS based on dielectrics have mostly been based on glasses, the formation of which is mainly attributed to the asymmetrically localized laser-field enhancement of the nanoplasma.²⁵ In fact, a similar LIPSS model/mechanism can be applied straight to the dielectric crystal family, yet, related experimental demonstrations are rather rare. One of the major causes for this is the Coulomb explosion effect, which leads to an unstable state of the material surface, and thus only crater and groove structures, instead of fine LIPSSs, can be defined on the surface of wide-bandgap transparent crystals, such as LiNbO₃.^111–113 To this end, by heating up LiNbO₃ and thus increasing its carrier density and conductivity, as well as inhibiting the Coulomb explosion effect, during fs laser processing, large-area LIPSSs with an average spacing of 174 ± 5 nm and enhanced optical absorption have been experimentally demonstrated in a recent study (see Fig. 5).¹¹⁴ Therein, it has been observed that the LIPSSs start to form when the temperature is increased above 300 °C and become uniform at around 500 °C [see Fig. 5(a)], where the optical absorption is largely enhanced [see Figs. 5(b) and 5(c)].¹¹⁴ Similar results have been subsequently demonstrated in Fe:LiNbO₃ as well.¹¹⁵ The heating strategy used in these studies can be easily applied to dielectric crystals beyond LiNbO₃, offering a promising approach to fabricating high-quality LIPSSs based on wide-bandgap transparent crystals.

In addition to localized modification of the original physical properties, fs laser pulses are also capable of altering the chemical properties of the dielectric material within the focal volume. Although highly material-dependent, the localized modification process eventually leads to a remarkable enhancement in the chemical dissolution rate of the laser-irradiated regions compared to the unmodified bulk material, enabling selective chemical etching.^116,117 Typically, the laser-induced modification regions act as the removing area or the etching mask in the subsequent etching process, which can be used for high precision and smooth 3D micro/nanostructure definition on the dielectric surfaces or deep into the substrate volume. Depending on the dimensions of the desired structures as well as the chemistry between the substrate material and chemical etchant, the etching time may vary from a few seconds up to several hours. This technique, also called FLICE (fs laser irradiation followed by chemical etching), has received a surge of interest since the first observation of this phenomenon in glasses,^116–119 opening up new opportunities for defining 3D microstructures. In particular, the FLICE technique enables direct fabrication of embedded hollow microstructures integrated with optical waveguides within dielectrics with arbitrary 3D design freedom, which is quite promising for e.g., microfluidic channels and 3D optofluidic devices.^6,116,117

The FLICE technique has been demonstrated to successfully define 3D microstructures with complex shapes in glasses.^{6,46,119–124} However, it is usually more challenging to implement laser-induced selective etching in dielectric crystals since they generally show much better chemical stability. Nevertheless, once the optimized laser irradiation parameters as well as a suitable etching recipe are met, a considerably high etching selectivity can be reached in certain dielectric crystals.^116,117 For example, in a recent breakthrough study, an etching selectivity larger than 10⁵ (the largest value ever achieved in laser-processed dielectric materials) has been experimentally demonstrated in laser-irradiated yttrium aluminum garnet (YAG) and sapphire crystals, enabling arbitrary 3D nanostructures with 100-nm feature sizes inside cm-scale laser crystals without brittle fractures.²¹ As a result, subwavelength air-pore photonic lattices [see Figs. 6(a) and 6(b)] have been fabricated by using which subwavelength diffraction gratings [see Fig. 6(c)] and microstructured optical waveguides [supporting diffraction-limited waveguide mode, see Fig. 6(d) and the inset] in YAG as well as mm-long nanopores in sapphire have been fabricated, introducing an exciting new dimension to standard fs laser micro-/nano-fabrication and offering a versatile tool to functionalizing conventional dielectric crystals toward future 3D nanophotonic integrated devices.^21,116,117

Furthermore, being a maskless, high-precision, and 3D micromachining technique, fs laser induced material modification can be used for surface lithography, which offers unique advantages in applied material universality and fabrication cost savings. In particular, most dielectric crystals suffer from significant difficulties in material integration/processing by complementary metal-oxide-semiconductor (CMOS) fabrication technology. In addition, fs laser micro-/nano-fabrication provides a promising path toward functionalization of dielectric crystals, especially in 3D devices, as well as integrated photonics. As a first step to this end, recently, the so-called fs laser assisted chemo-mechanical polish lithography (CMPL),¹²⁵ as a physical variant of the FLICE, has been successfully applied in the fabrication of high-quality on-chip waveguides (with a propagation loss as low as 0.027 dB/cm)¹²⁶ and micro-resonators (with an ultrahigh Q-factor up to 10⁸)¹²⁷ based on the thin-film LiNbO₃ on insulator (LNOI) platform.^30,65 Based on such a fs-laser-assisted CMPL fabrication process, subsequently, a reconfigurable multifunctional photonic integrated circuit (PIC),¹²⁸ which integrates a high-extinction-ratio beam splitter, a 1 × 6 optical switch, a balanced 3 × 3 interferometer, and electro-optically switchable optical true delay lines (OTDLs) with meter-scale lengths,¹²⁹ has been experimentally demonstrated, suggesting the great potential of such a fabrication recipe toward multifunctional LNOI integrated photonics. Accordingly, such a strategy can be applied to other dielectric crystal platforms, affording many “old” optical materials like LiNbO₃ new opportunities in multifunctional PIC applications.

Before moving forward into the conclusion/outlook section, we believe it is necessary to briefly discuss here the impacts of optical aberration during laser fabrication. Fundamentally, the overall optical quality and device performance of photonic structures defined by fs laser micro-/nano-fabrication are critically dependent on a sequence of laser-induced effects, such as multiphoton absorption and heat diffusion, which are closely correlated with the focal conditions. As a result, a slight disturbance in the spatial distribution of a focused light beam leads to dramatic changes in the physical properties of fabricated structures and, therefore, the overall device performance.^{123,124,130,131} Therefore, flexibly adjusting the focal beam shape to cope with diverse fabrication scenarios is crucial with a view to achieving high-quality micro-/nano-photonic devices. The refractive index mismatch, usually around 0.5–1.5 in the case of crystals, between air and processed materials introduces optical aberrations, which leads to focus distortion during fabrication and thus severely reduces the processing accuracy/efficiency.^{123,130–134} In particular, the distortion becomes even more serious as the focal depth increases, strongly limiting the widespread applications of fs laser in nanophotonic structures/devices with high-quality 3D layout.^{123,130,131,135,136}

Fs structured-light laser fabrication employing beam-shaping techniques, such as spatial light modulation (SLM), has experienced tremendous development in recent years, enabling high-accuracy aberration correction and, therefore, high-efficiency 3D processing in diverse materials.¹³⁰ In such a scenario, the spatial distribution of laser intensity and induced modified area can be finely regulated over a wide range of depths (on a scale of >mm), which is essential for multi-layer processing and true 3D fabrication deep inside transparent dielectrics.^{131,136–139} By using the analytical expression for depth-dependent aberration correction proposed by Booth et al.,^{134,136,138,140,141} e.g., single-mode waveguides with uniform cross-sections and circular symmetric profiles at different depths in glasses can be experimentally achieved. Even more accurate aberration correction is also possible by applying closed loop optics with a real-time feedback system for optimizing the spatial phase distribution of a used fs laser.^{130,142–145} Therefore, fabrication of high-performance photonic structures/devices with on-demand geometric features can be readily implemented by using aberration-corrected fs laser with optimized beam shape and intensity distribution. As of yet, such a fabrication strategy is rarely seen in the definition of photonic structures based on dielectric crystals. A typical experimental demonstration so far has been 3D waveguides fabricated in diamond.¹⁴⁶ 

In this article, notable promising paths for the future development of fs-laser micro-/nano-fabrication of functional photonic structures in dielectric crystals are briefly highlighted. These nanophotonic features, together with other fs-laser-defined chip-integrated functionalities, are rapidly revolutionizing a broad range of photonic applications, from classic optics in optical communications and nonlinear optics to quantum regimes such as quantum photonics and optomechanics.^15,46,63 To enable these, however, several important challenges need to be addressed. In this section, we discuss the key challenges ahead for the above-mentioned research topics and outline several exciting near- and long-term opportunities.

In terms of AOP-based 3D domain engineering, a sequence of research breakthroughs in recent years have enabled direct domain engineering with a resolution reduced to 30 nm inside a dielectric/ferroelectric crystal.^24,32,33 Such a high resolution paves a new way for high-performance nonlinear optical frequency conversion based on QPM structures/devices, such as NPCs,²³ and even ferroelectric-domain-wall-based nanoelectronic devices, e.g., photovoltaic devices and nonvolatile memories.^147,148 One of the major technical challenges that impose limitations on the practical applications of AOP-based domain-engineered structures is their relatively small size. For example, the typical dimensions of 3D NPCs in recent demonstrations are limited to 100 μm.²³ With a view to achieving high-efficiency nonlinear frequency conversion, an effective light–matter interaction length larger than the millimeter level is usually required. To this end, an effective approach is to employ a multifocal laser-writing strategy to facilitate fast domain engineering on a large scale. Alternatively, the combination of AOP-based domain engineering and whispering-gallery-mode (WGM) resonators, in which the guided light circulates in spheroidally shaped dielectrics, is a promising strategy for lengthening the effective light-mater interaction and thus enhancing nonlinear frequency conversion without enlarging the total footprint of the nonlinear devices.^149–151 Such a strategy has been successfully demonstrated by using radial EFP and mm-sized/chip-integrated WGM resonators based on LiNbO₃, providing a significant improvement in the nonlinear frequency conversion efficiency.^152–155 By replacing the EFP with the state-of-the-art AOP, even finer domain engineering and, therefore, ultra-high efficiency nonlinear frequency conversion in a monolithic on-chip device can be expected. Moreover, the laser-writing-direction dependency of the recently demonstrated 3D nanodomain-engineering techniques can be further applied to ferroelectric crystals, such as lithium tantalate (LiTaO₃) and potassium titanyl phosphate (KTP), beyond LiNbO₃, holding great promise for boosting the development of high-quality 3D integrated nanodevices for advanced optoelectronics applications.

In the case of color center creation, fs laser writing has established itself as a promising alternative to ion/electron beam irradiation methods.^22,56,57 However, the relevant mechanisms and theoretical models describing the fs-laser-induced color center generation have not been studied thoroughly yet. In addition, in this regard, all the laser processing parameters, including the laser wavelength, laser energy, repetition rate, laser polarization, exposure time, and pulse duration, need to be precisely adjusted and systematically researched. In addition, the relevant lattice structures and material properties have to be taken into consideration since, beyond diamond, fs-laser-induced color centers have been realized in a variety of crystals, such as silicon carbide (SiC),¹⁵⁶ lithium fluoride (LiF),^157,158 as well as some laser and nonlinear crystals.⁵⁶ Overall, the research related to the fs laser writing of color centers in dielectric crystals is still in its infancy, and the applicable material scope will expand toward different applications together with the in-depth studies of experimental principles and relative material processing techniques. In terms of photonic integration of fabricated color centers and integrated photonic structures, such as waveguides and gratings, although there have been some promising attempts, alignment accuracy and coupling efficiency are very limited.^57,61,159 For example, the combination of NV⁻ color centers and buried waveguide structures has been experimentally demonstrated in quantum grade diamonds by employing the same laser and positioning system, enabling alignment accuracy within micrometer resolution.⁶¹ Therein, the waveguide-NV⁻ coupling efficiency is estimated to be rather low. Improvement can be expected via optimization of the waveguide geometry and, therefore, the mode field. This work can be regarded as the beginning toward on-chip optical routing of single photons between NV⁻ color centers and optically integrated spin-based sensing applications, such as magnetometry, electrometry, and thermometry.⁶¹ Future attempts on this research topic will also continue to aim at accurate placement, yield enhancement, property improvement, and processing dimension expansion. Spatial light modulation technology and real-time monitoring/active control of the laser fabrication will be of great help in the process.^22,56,57

On a different front, LIPSS has emerged as a promising fabrication strategy for enrichment, enhancement, and modification of existing photonic functionalities based on dielectrics. Although the experimental demonstration of LIPSS in dielectric crystals is rather rare, the successful experience in glasses and semiconductors can be fully borrowed. Meanwhile, the corresponding formation mechanism of LIPSSs for dielectrics needs to be well developed. To this end, a lot of efforts must be made in experimental and theoretical studies, especially ultrafast dynamics.²⁵ Furthermore, 2D periodic nanoantenna arrays, or metasurfaces with subwavelength resolution and tailored functionalities in optical field manipulation, fabricated by 2D-LIPSS, are and will continue to be an exciting research topic that may evolve into an efficient alternative to lithographic techniques. Recently, the first demonstration of subwavelength 2D-LIPSSs based on dielectric crystals has been realized on diamond surfaces with a single step process, exploiting sequences of two temporally delayed and cross-polarized fs laser pulses generated by an interferometer-like configuration.¹⁶⁰ This milestone study marks the starting point toward high-throughput laser fabrication of large-scale subwavelength dielectric metasurfaces for future planar optoelectronic applications.

Moreover, as the fs-laser micro-/nano-fabrication method is fully compatible with the existing optical fabrication techniques, including fs-laser direct writing of waveguides, gratings, microresonators, and photonic integrated circuits, the newly developed nanostructures, including but not limited to the above-mentioned ones, can be easily combined with well-developed designs/devices with a view to achieving multi-functional integrated 3D photonic systems for both classic and quantum photonics.^11–15 

As a result of the constant quality improvement of fs-laser fabricated photonic structures, especially optical waveguides, in terms of waveguide loss and mode manipulation, their application scope has been smoothly extended to topological physics and quantum information processing. In particular, fs-laser direct written arrays of evanescently coupled waveguides have been established as a multi-functional platform for supporting light propagation/evolution that reveals nontrivial bulk/edge topological states or exhibits intrinsic similarities with the quantum evolution of particle wave-functions, enabling the achievement of various innovative topological states, quantum physical concepts, and quantum information applications that are not feasible with other fabrication techniques. A major challenge ever-present for practical applications of the fabricated optical waveguide devices in PIC is high-efficiency optical coupling between waveguides and other external optical components, such as standard optical fibers. To this end, specially designed coupling fibers (such as lensed optical fibers¹⁶¹ and tapered optical fibers^162,163) can be used with a view to reducing the large mode mismatch between standard optical fibers and waveguide devices.^30,164

In addition to what we have described earlier, there are many other areas in physics and applications of fs laser micro-/nano-fabrication technology, e.g., 3D nanolithography^21,165–167 and additive manufacturing,^168,169 and the fabricated functional 3D structures for, e.g., optofluidics,¹⁷⁰ bionics,¹⁷¹ and microbots,¹⁷² that we believe will undergo rapid development in the coming years. In addition, as indicated above, the possible developments in this area will benefit extensive domains from science to industry. Overall, the immediate future of fs-laser micro-/nano-fabrication is clear and bright. By further combining with dielectric crystal families, its versatility and complexity are fascinating and intriguing. We expect such a promising combination to make a significant impact on every corner of future photonics.

This work was supported by the National Natural Science Foundation of China (Grant Nos. 12174222 and 12074223), the Shandong Provincial Natural Science Foundation (Grant Nos. ZR2021ZD02 and 2022HWYQ-047), the Taishan Scholars Program of Shandong Province (Grant Nos. tspd20210303 and tsqn201909041), and the “Qilu Young Scholar Program” of Shandong University, China.

The authors have no conflicts to disclose.

Yuechen Jia: Conceptualization (lead); Funding acquisition (equal); Writing – original draft (lead); Writing – review & editing (lead). Feng Chen: Funding acquisition (equal); Resources (equal); Writing – review & editing (equal).

Data sharing is not applicable to this article as no new data were created or analyzed in this study.