Laser Floating Zone Growth: Overview, Singular Materials, Broad Applications, and Future Perspectives

Abstract

The Laser Floating Zone (LFZ) technique, also known as Laser-Heated Pedestal Growth (LHPG), has been developed throughout the last several decades as a simple, fast, and crucible-free method for growing high-crystalline-quality materials, particularly when compared to the more conventional Verneuil, Bridgman–Stockbarger, and Czochralski methods. Multiple worldwide efforts have, over the years, enabled the growth of highly oriented polycrystalline and single-crystal high-melting materials. This work attempted to critically review the most representative advancements in LFZ apparatus and experimental parameters that enable the growth of high-quality polycrystalline materials and single crystals, along with the most commonly produced materials and their relevant physical properties. Emphasis will be given to materials for photonics and optics, as well as for electrical applications, particularly superconducting and thermoelectric materials, and to the growth of metastable phases. Concomitantly, an analysis was carried out on how LFZ may contribute to further understanding equilibrium vs. non-equilibrium phase selectivity, as well as its potential to achieve or contribute to future developments in the growth of crystals for emerging applications.

1. Introduction

Growth of crystalline solids can be achieved by a large variety of methods driven by thermal or chemical potential gradients. Crystal growth in nature is achieved under many different geological conditions, in some cases under extremely high temperatures and pressure and in others by very slow solution processes [1]. Many current laboratory methods are nothing but transcripts of those that we can find in nature, as in the case of hydrothermal processes involved in the preparation of many functional materials [2]. A comprehensive classification of crystal growth techniques can be found in the classical work of Pamplin [3]. In this work, four main categories of growth methods are established: Growth from solid (S→S), melt (L→S), vapor (V→S), and solution (sol→S). Melt growth techniques are the choice methods to obtain bulk single crystals of inorganic materials and they could be divided into four categories: crystal pulling, directional solidification, floating zone, and Verneuil techniques [4]. Crystallization techniques such as Verneuil, Bridgman–Stockbarger, and Czochralski methods are usually employed for large and conventional production of gemstones (Al₂O₃ allomorphs and doped crystals, TiO₂, SrTiO₃), semiconductor single crystals (Si, Ge, GaAs), or metals (Pd, Pt, Ag, Au). In these methods, a material of approximately the correct composition is melted congruently, i.e., the crystalline phase is maintained before and after melting, being solidified in a carefully controlled fashion causing the formation of a single crystal from a well-oriented seed material [5]. However, they present some disadvantages when the materials are highly refractory or exhibit incongruent melting. Likewise, they usually imply the use of a large amount of starting precursor materials, long processing times, and the use of crucibles that could introduce impurities on crystals with consequences in their physical properties. In contrast, the study of new materials implies limited quantities of precursors, the required highest purity possible to present the best performance, and, considering these competitive times, their production in the shortest time possible. Thus, the Laser Floating Zone (LFZ) technique, namely, Laser-Heated Pedestal Growth (LHPG), has been revealed as a suitable prototyping technique that enables the production of high-quality crystalline materials in a simple, fast, and crucible-free method with low consumption of precursor materials. In addition, several studies have allowed us to easily process materials presenting incongruent melting, as well as obtaining non-equilibrium phases.

The origin of this technique may be assigned to Haggerty in 1972 [6], who defined a Laser-Heated floating zone growth process of Al₂O₃ (sapphire) fibers for NASA (National Aeronautics and Space Administration; Cleveland, OH, USA). Feigelson later applied the method to the growth of single-crystal fibers with potential use in solid-state lasers [7]. A good number of materials have been obtained by LFZ since then, for a wide range of applications taking into account the continuous improvement of the technique along these years, considering both laser source and optical setup [7]. Thus, materials for photonics and electrical applications can be highlighted as the most produced by LFZ together with the industry interest on environmentally friendly processes and materials, reducing energy consumption and enhanced properties of materials. It must be noted that, since the reviews published by Feigelson in 1989 [7], Rudolph and Fukuda in 1999 [8], and Andreeta et al. in 2010 [9], there is no other review about materials processed by the LFZ (or LHPG) technique. Taking into account the materials produced during the past number of years, as well as the latest apparatus advances for enhancing the quality of the molten zone and the consequent solidification, this call for a review focused on the technique and its materials’ development. Concomitantly, most promising fields are noted, highlighting the corresponding materials produced, aiming to define the future perspectives of the LFZ technique.

Before describing in detail the LFZ technique, focusing on materials developed for photonic or electrical applications, however, it is convenient to mention the analogous Optical Floating Zone (OFZ) technique [10,11,12,13,14,15]. In fact, a number of interesting papers have been published regarding the growth of different kinds of single-crystal materials. These include, for example, high-quality rubies [16], europium-doped, yttria-stabilized hafnia (YSH) [17], multifunctional BaZrO₃ [18], the Ba₂PrFeNb₄O₁₅ ferroelectric relaxor [19], or rare-earth disilicates of Er, Ho, and Tm with a rich range of interesting magnetic properties [20]. Technically, the difference between OFZ and LFZ is in the optical radiation sources used to attain heating and melting. Usually OFZ makes use of ellipsoidal mirrors with a secondary focal point at the center of the growing rod or crystal, while the primary focus contains a powerful halogen or xenon lamp. This setup allows achieving high melting temperatures with a significantly lower energy consumption, as compared to conventional crystallization methods described above [13,14]. However, the LFZ technique allows better control of the temperature gradient compared to OFZ, as well as higher temperatures [21] since the light focus in OFZ is broad and the temperature gradient at the interface between the solid and the liquid is less abrupt. This makes the melt seriously attack the feed rod and spill over to the crystal, eventually making the growth unstable [22]. Likewise, the use of lasers allows implementation of a high-strength metal growth chamber, permitting high pressures, up to 1000 bar [23]. Meanwhile, in conventional mirror-based designs, namely, OFZ chambers, apparently the maximum pressure achieved is in the proximity of 300 bar [15]. In both cases, high pressures are desired to enable the growth of highly volatile and metastable materials.

2. LFZ Technical Developments

The first equipment developed by Haggerty in 1972 [6] was proposed for the production at ambient atmospheres of Al₂O₃:Cr, TiC, and Y₂O₃ fibers, noting that there were no available crucibles for the melting of the first two materials. Thus, these materials were produced using an apparatus provided with a small 10-W CO₂ laser, coupled to a basic optical system composed of (1) the beam expanding and pointing optics, (2) the beam splitter, and (3) the beam splitting and focusing optical bench. The beam splitter consisted of a water-cooled, coated GaAs window and a front surface mirror. Meanwhile, the two furnace windows—after beam splitter—were made of NaCl. After that, beams were intercepted by semicircular and spherical mirrors and focused at the position of the molten zone. Finally, fiber withdrawal and feed-rod insertion mechanisms were installed inside the furnace as pulling heads.

Andreeta et al. [9] summarized the technical advances of existing prototypes through 2010. Among all advances, the introduction of the reflaxicon by Fejer et al. [24,25] may be highlighted. It enabled a circular, crown-shaped laser beam focus and, therefore, uniform radial heating. This term describes a setup of a two-stage pair of reflective linear axicon surfaces [21], which was invented by Martin in 1948 [26] and later improved by Nubling and Harrington in 1997 [27] (Figure 1). Likewise, processing under high-vacuum conditions, achieved for the first time by Brueck et al. in 1996 [28], must also be highlighted. Another remarkable advancement was achieved by Carrasco et al., in 2004 [29], through the application of an electrical current during processing. By establishing the Electrically Assisted Laser Floating Zone (EALFZ) (Figure 2) they demonstrated impressive preferential texture and the consequent enhancement in electrical properties of high Tc superconductors. In the same year and after previous experiments employing floating zone combined with a YAG (Y₃Al₅O₁₂) laser, Geho et al., in 2004 [30], reported the design of a hybrid laser floating zone machine for the successful growth of incongruently melting Tb₃Al₅O₁₂ (TAG) single crystals (Figure 3). This new device combined four CO₂ lasers and four halogen lamps, aiming to reduce temperature gradients during solidification. Likewise, Sekijima and Geho, in 2004 [31], patented an apparatus composed of two CO₂ lasers for growing TAG single crystals.

Outstanding advances were achieved during the last 10 years, also focused on improving radial heating uniformity. Most of these implied the use of laser diodes both to improve radial heating, envisaging the growth of incongruently melting materials, as well as to substantially increase total laser power. The first study was reported by Ito et al., in 2013 [22], who successfully grew incongruent materials such as BiFeO₃ and (La,Ba)₂CuO₄ by developing a laser–diode-heated floating zone (LDFZ) apparatus. The latter made use of five laser diodes emitting at a wavelength of 975 nm, with a total laser power output of 350 W and without use of a reflaxicon (Figure 4). In addition, these authors studied the effects of the number of laser diodes (3 to 8) on the quality of radial heating. A similar apparatus was recently employed by Kaneko NS Tokura, in 2020 [32], to grow refractory (Al₂O₃:Cr, SmB₆), incongruent melting (Ba₂Co₂Fe₁₂O₂₂) and volatile (Nd₂Mo₂O₇, SrRuO₃) materials by LDFZ, employing five laser diodes emitting at 940-nm wavelength with a total power of 1 kW. Thus, the use of various lasers enables uniform irradiation intensity distribution on the periphery of the raw material. Moreover, a vertical irradiation intensity can be designed to exhibit a flat or bell-shape distribution to improve relaxation of residual thermal strain in the grown crystal [22,32]. Nowadays, the Crystal Systems Corporation (Hokuto, Yamanashi, Japan) commercializes a laser floating zone furnace that provides a total laser power output of 5 kW, employing five laser diodes emitting at a wavelength of 808 nm [33]. In addition, similar equipment from Quantum Design International (San Diego, CA, USA) allows monitoring the temperature in the range 1173–3273 K while achieving a maximum laser output power of 2 kW from five laser diodes [34].

Finally, Schmehr et al., in 2019 [23], developed a high-pressure LFZ (HP-LFZ) apparatus provided with seven laser diodes, emitting at a wavelength of 810 nm with a total laser output power of 700 W. They particularly aimed at avoiding outgassing/volatily effects inherent to materials that exhibit high vapor pressures at high temperatures. In addition, they also addressed the melting of highly refractory materials with volatile component loss reduction, by applying high processing gas pressures. Thus, Cu₂O, Nd₂Zr₂O_7, and LiCuO₂ crystals have been successfully grown using this last apparatus.

3. Experimental Procedure (Standard)

LFZ growth requires precursor powders in the form of cylindrical rods to be used as feed and seed and there are three typical processes to produce these precursor rods:

(1)

The extrusion process is the most common way to prepare the precursor rod cylinders for the LFZ process, since it is a simple method, not requiring special equipment or additional hands [21,29]. Thus, the commercial raw oxide powders are mixed, according to the desired stoichiometry, and reduced in grain size with an agate ball mill or similar equipment. The purity of the precursors depends on the desired application. For example, the use of powders of 5–6 N of purity should be envisaged for photonic applications. Aiming to bind the powder mixture for the extrusion process, polyvinyl alcohol (PVA, 0.1 g/mL) is added, mashing the powders until a compact and plastic paste is achieved. The obtained clay is then extruded into cylindrical rods, with diameters that can reach up to 5 mm, depending on the material’s nature and its application. After extrusion, the cylindrical rods are dried in air and ready to be used as feed and seed materials (also known as green rods). However, it is important to emphasize that for support the extruded bars should have slots, aiming to guarantee their alignment during the drying process.

(2)

Alternatively, single crystals or dense ceramics appropriately cut can also be used as seed or feed rods instead of green rods [26]. The use of bulk-grown crystal seeds favors the formation of single-crystalline fibers. This approach helps laser processing and allows enhancing the structural characteristics of the single-crystal fiber produced. Similarly, in the last 10 years, cladded, single-crystalline fibers, mainly used as amplifiers, have been produced from bulk crystal seeds covered by Sol-Gel or embedded into silica or borosilicate hollow tubes, among other coating approaches [35,36,37,38,39]. Furthermore, some works from Rutgers University (Piscataway, NJ, USA) reported the growth from Pt wires together with the use of seed crystals or presintered ceramics as feed rods [40,41,42].

(3)

Likewise, precursor rods can be also prepared by cold isostatic pressing [43]. Through this compaction method, both mixed raw and presintered powders with the desired composition are enclosed in a flexible mold. This flexible bag is introduced into a perforated support inside a pressure container. Once this setup is sealed, fluid pressure is exerted over the outside surface of the container, allowing the container to press all around the bag and inducing uniform compaction of the powder and, consequently, a uniform density within the compacted rod [44].

The LFZ equipment usually comprises a CO₂ laser coupled to a reflective optical setup which includes a reflaxicon, described in the previous section. In the case of LDFZ, this optical setup is not present. Once seed and feed fibers are placed on the respective holders, a molten zone is formed by irradiating the densified rods with the CO₂ laser and a different optical configuration. Fibers can be grown in the ascending or descending direction from this molten region at a defined growth or pulling rate, in air or special atmospheres, depending on the materials’ physico-chemical properties. Likewise, feed and seed rods should be rotated, favoring the mixing of precursors in the melt, homogenizing the temperature of the molten material, and contributing to reduce unsymmetrical thermal stresses. Concomitantly, occurrence of constitutional supercooling is reduced. Thus, the melt rotation stabilizes the heat and mass flow against undesirable temperature perturbations [45]. The growth process can end abruptly or by reducing the laser power gradually. This procedure is very important to reduce the thermal stresses and, therefore, to avoid crack formation [9,21].

4. Materials for Photonic and Optical Applications

Several high-quality crystals for photonic and optical applications have been obtained via the LFZ technique. These include laser garnets, luminescent, fluorescent, optoelectronic, and photorefractive materials. Indeed, Haggerty, in 1972 [6], produced ruby and yttria when the technique was reported. Likewise, Nd:YAG fibers were obtained by Stone et al., in 1976 [46], from a Nd:YAG preformed rod with a CO₂ laser. This was not the conventional LFZ method described here since a platinum wire was dipped into the melt and raised slowly to pull the desired crystal. The first notice about the production by LFZ (or LHPG) of crystals for laser applications was reported by Fejer et al., in 1984 [25], when the reflaxicon was introduced. The suitability of the LFZ apparatus developed was thus tested, producing sapphire, ruby, LiNbO₃, and Nd:YAG laser media.

Due to the crystal requirements for optics and photonics [47], defect-free, transparent, and highly pure materials with lower thermal conductivity [48] have been produced since the technique’s origin [6]. An excellent compilation of active and passive photonic materials was conveniently reported by Rudolph and Fukuda in 1999 [8]. Likewise, Maxwell et al., in 2017 [49], reported the LFZ production of neodymium- and ytterbium-doped single-crystalline YAG cylinders (SCF) with diameters under 1 mm and, in addition, processing in combination with the Sol-Gel method for obtaining cladded fibers. Moreover, the increasing interest of industry on efficient optical processes required an updated revision from that reported by Rudolph and Fukuda in 1999 [8]. Indeed, it is remarkable the great interest that has arisen recently on crystals for ultrashort pulsed laser systems [50,51] and for LED devices applied for indoor plants’ growth [52], as a consequence of the increasing food requirements derived from the human population increase.

One of the most LFZ-developed materials is single crystal Y₃Al₅O₁₂ (YAG) as a laser medium. Despite the fact that it was produced by Fejer et al. in 1984 [25], a considerable amount of work has appeared during the last 20 years (

Table 1. Y₃Al₅O₁₂ (YAG) garnets grown by LFZ since the review by Rudolph and Fukuda in 1999 [8].

Year	Authors	Dopant	Application	Ref.
2000	Ishibashi & Naganuma	Cr4+	Diode-pumped garnet	[53]
2002	Shen et al.	Cr4+	Garnets for Q-switch lasers	[54]
		Nd3+
2003	Boulon et al.	-	Laser media	[55]
2003	Yoshikawa et al.	Yb3+	Laser media	[56]
2005	Ye et al.	Cr3+	Temperature sensor	[57]
2006	Bufetova et al.	Nd3+	Laser media	[58]
2009	Chen et al.	-	Laser media	[59]
2009	Lai et al.	Cr4+	Cladded Laser media	[60]
2010	Yi et al.	Cr4+	Cladded Multi-pass ring laser	[61]
2011	Chang et al.	Cr3+	Laser media	[62]
2011	Zhu et al.	-	Laser amplifier	[40]
2012	Lautsen & Harrington	-	Laser media	[41]
2012	Wang et al.	Cr4+	Cladded fiber amplifier	[35]
2012	Kim et al.	Ho3+	Cladded High power laser media	[63]
		Yb3+	Cladded High power laser media
2013	Hsu et al.	-	Cladded Laser gain media	[64]
2013	Hsu et al.		Multi-mode media	[65]
2013	Kim et al.		Cladded laser media	[36]
2013	Lai et al.	Cr4+	Double-cladded fiber amplifier	[66]
2013	Maxwell et al.	Er3+	Cladded laser media	[67]
		Nd3+
		Yb3+
2014	Harrington	-	Laser media	[42]
2014	Wang et al.	Nd3+	Waveguide	[68]
2015	Nie et al.	Er3+	Garnets	[37]
		Ho3+
		Nd3+
		Tm3+
		Yb3+
2016	Kim et al.	-	High power fiber lasers	[50]
		Ho3+
		Yb3+
2016	Oliete et al.	-	YbAG comparative study	[69]
2017	Liu et al.	Cr4+	Fiber transmission systems	[70]
2017	Maxwell et al.	Er3+	Laser performance of cladded-core	[49]
		Nd3+
		Yb3+
2019	Bufetova et al.	Er3+	Laser media	[71]
2019	Kim et al.	Ho3+	High power fiber lasers	[72]
		Yb3+	High power cladded fiber lasers
2019	Wang et al.	Yb3+	High power laser	[51]
2020	Bera et al.	Nd3+	Garnets	[38]
		Nd3+/Ho3+

) reporting its growth [35,36,37,38,40,41,42,49,50,51,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72], improving the production efficiency [59,71], and also enhancing its mechanical and optical characteristics [38]. Among those studies, the work of Nie et al., in 2015 [37], who produced YAG single crystals doped with Er³⁺, Ho³⁺, Tm³⁺, Nd³⁺, and Yb³⁺ may be highlighted here. They studied the dopant distribution following the ideal concentration for optimal photonic properties, together with the enhanced mechanical and microstructural characteristics. Likewise, diode-pumped garnets based on YAG crystals doped with metallic ions such as Cr⁴⁺ for Q-switched laser operation were also developed [53,54]. Focusing on this composition, both Lai et al., in 2009 [60], and Yi et al., in 2010 [61], developed laser gain media and multi-pass ring laser devices, respectively, from single crystal fibers (SCF) with the double-clad structure. The double-clad structure fabrication has been generally achieved by covering the single-crystal fiber through Sol-Gel and later sintering [49] or basically inserting a diameter-reduced grown rod into silica- or glass-based capillary tubes [35,50,63,66,68]. These kinds of materials, namely, the cladded, single-crystal fibers, have been largely applied as fiber amplifiers of the laser radiation [40,67,70]. Recently, Kim et al., in 2019, produced Yb:YAG core/YAG-clad fibers by a three-step process, combining an initial growth of doped YAG fibers, followed by acid etching and, finally, an epitaxial hydrothermal growth of the undoped YAG cladding [72]. On the other hand, Ye et al., in 2005 [57], grew an excellent YAG:Cr³⁺ crystal to be used as a fiber thermometer. This was based on fluorescence lifetime and found applications in temperature monitors for microwave treatments and medium-voltage substations. This application for temperature sensing has been also successfully explored in work from Zhejiang University (Hangzhou, Zhejiang, China), reporting the growth of Y₂O₃:Er³⁺/Yb³⁺ [73] and Y₂O₃:Ho³⁺/Yb³⁺ [74] single-crystal fibers with up-conversion luminescent characteristics.

Other complex oxides have also been developed by LFZ. Harrington, in 2014 [42], reported, for example, the growth of single-crystal fiber garnets of MgAl₂O₄ (spinel) from commercial YAG seeds and Pt wires. In 2004, Romero et al. [75] produced multiwavelength-laser, single-crystal aluminates of YAlO₃ doped with Nd³⁺ (0.5–1.5 mol%). Recently, single crystals of orthosilicates (RE₂SiO₅) also doped with rare earths (RE) have been produced by LFZ two to three times faster than conventional methods like Czochralski [76,77,78], maintaining or even improving laser performance by doping with Nd³⁺ or Yb³⁺ [48]. Other typical garnets usually employed in commercial laser resonators are based on lanthanum vanadates. Such is the case for LaVO₄, Gd_1−xLa_xVO_4, and Y_1−xLa_xVO₄, produced by Andreeta et al. in 2006 [79] from mixtures of raw oxide powders. These were processed in air at relatively high growth rates (9–18 mm/h), as compared to those applied on Czochraslki. From the research group of Andreeta et al. (Universidade Federal de São Carlos; São Carlos, Brasil), erbium-doped and undoped CaNb₂O₆ and CaTa₂O₆ single crystals, suitable as laser active media, were also developed and produced [80]. The optically transparent CaTa₂O₆ single crystals were also grown by Almeida et al., in 2013 [81], exploring the three polymorphic modifications that this material exhibits at room temperature. Focusing on niobates, the production of promising optical active elements for lasers such as lithium niobates (LiNbO₃) [55,82,83,84,85] or incongruent melting lithium-potassium niobate (K₃Li_2−xNb_5+xO_15+2x, KLN) [86,87,88] (

Table 2. Niobates grown by LFZ since the review by Rudolph and Fukuda in 1999 [8].

Year	Authors	Material	Application/Study Aim	Ref.
2001	Reyes-Ardila et al.	LiNbO3	LHPG technique development	[82]
2001	Matsukura et al.	KLN 1	Optical applications	[86]
2002	Andreeta et al.	LiNbO3	LHPG technique development	[83]
2003	Boulon et al.	LiNbO3	Laser media	[55]
2003	Bourson et al.	LiNbO3:Fe	Photorefractive material	[89]
2003	Cochez et al.	LiNbO3:Fe	Photorefractive material	[90]
2004	Guo et al.	KLN 1:Zn2+	SHG 2 blue laser	[87]
2004	Nagashio et al.	LiNbO3	Nonlinear optics and acoustic wave	[91]
2005	Chen et al.	LiNbO3	Optical material growth & study	[84]
2005	Lee et al.	LiNbO3:MgO	Wavelength modulator	[92]
2007	Chen et al.	LiNbO3	Optical material growth & study	[85]
2008	De Camargo et al.	CaNb2O6	Laser media	[80]
		CaNb2O6:Er3+	Laser active media
2011	Maxwell et al.	KLN 1	Laser media	[88]
2013	Graça et al.	EuNbO4	Dielectric devices	[94]
2015	Kashin et al.	LiNbO3	Frequency doubling	[93]

¹ KLN: K₃Li_2−xNb_5+xO_15+2x. ² SHG: second harmonic generation.

) must be also highlighted. In addition, more niobates were also grown by LFZ for holographic devices, as photorefractive materials by doping with Fe [89,90], for nonlinear optics and acoustic wave devices [91], for wavelength and frequency modulators [92,93], and, in the case of the EuNbO₄, for optoelectronics as light emitters [94].

Following the production of laser garnets, gadolinium- and gallium-based compositions have also been produced during the last several years. Taking into account the ease of evaporation of gallium oxide (Ga₂O₃), an excess of the same must be introduced, aiming to obtain the desired compositions [95]. Thus, Harrington reported, for the first time in 2014 [42], the single-crystal production of Gd₃Ga₅O₁₂ (GGG) laser garnet. Recently, Rey-García et al., in 2020 [21], produced, in air atmosphere, gadolinium oxyorthosilicate single crystals, suitable to be employed as laser host matrices.

Typical sapphire (Al₂O₃), ruby (Al₂O₃:Cr), and rutile (TiO₂) crystals have been continually produced by this technique, improving both production efficiency and material properties, as can be deduced from work published in the last few years [22,23,32,39,42,96,97,98,99,100,101,102,103,104] (

Table 3. Sapphire (Al₂O₃) and ruby (Al₂O₃:Cr) fibers produced by LFZ since 1999 [8].

Year.	Authors	Material	Application	Ref.
2001	Seat	Al2O3:Cr3+	High temperature sensing	[96]
		Al2O3:Er3+
		Al2O3:Er3+:Yb3+
2003	Seat & Sharp	Al2O3:Er3+:Yb3+	High temperature sensing	[97]
2006	Liu et al.	Al2O3:Mg	Strengthened sapphire garnet	[89]
2010	Carvalho et al.	2Al2O3-SiO2:Nd3+	Scintillator	[111]
2012	Dragic et al.	Al2O3	Fiber sensors & high energy lasers	[102]
2012	Mesa et al.	Al2O3-ErAG 1-ZrO2	Selective emitter for Si TPV cells	[117]
2013	Ito et al.	Al2O3:Cr3+	LDFZ technique development 2	[22]
2014	Harrington	Al2O3	Laser media	[42]
2014	Lai et al.	Al2O3:Ti4+	High power cladded fibers	[103]
2016	Lai et al.	Al2O3:Ti4+	Light sources for endoscopy	[39]
2017	Bufetova et al.	Al2O3	Laser media	[100]
2017	Ren et al.	Al2O3-ErAG 1-ZrO2	Selective emitter for Si TPV cells	[116]
2018	Lai et al.	${\mathrm{Si}}^{4+}:\gamma$-Al2O3	Second Harmonic Generation	[104]
2019	Liu et al.	Al2O3	Laser media	[101]
2019	Schmehr et al.	Al2O3:Cr3+	HP-LFZ furnace development 3	[23]
2020	Kaneko & Tokura	Al2O3:Cr3+	LDFZ technique development 2	[32]

¹ ErAG: Er₃Al₅O₁₂; TPV: thermal photovoltaic cells. ² LDFZ: Laser Diode Floating Zone. ³ HP-LFZ: High-Pressure Laser Floating Zone.

). Indeed, doping of sapphire with Cr³⁺ and Er³⁺, or even codoping with Er³⁺/Yb³⁺, was achieved through LHPG by Seat in 2001 [96] and Seat and Sharp in 2003 [97], demonstrating the suitability of the fibers produced to be successfully employed for high-temperature sensing, as Ye et al., in 2005 [57], also similarly tested later for YAG:Cr³⁺. Likewise, Lai et al. in 2016 grew borosilicate-cladded, single-crystalline-core sapphire fibers suitable to be employed in biomedical applications, such as light sources for endoscopy [39]. It must be highlighted that Lai et al., in 2018, also demonstrated a facile and scalable approach for the transformation of a centrosymmetric sapphire (α-Al₂O₃) crystal to a large-scale 3D metamaterial (Si⁴⁺:γ-Al₂O₃) with sub-wavelength fine structures [104]. Coming back to garnets for laser operation and revising the work mentioned up to this point, we could metaphorically denote this section as “Laser Kindergarten: laboratories in those laser parents produced laser children”. Additionally, due to their optical properties and their efficiency, it is also interesting to remark on the works focused on the growth of laser garnets based on disordered crystals like sesquioxides, fluorides, or chalcogenides (

Table 4. Crystalline fibers of sesquioxides and other binary oxides produced by LFZ since 1999 [6].

Year	Authors	Oxide	Dopant	Application	Ref.
2001	Laversenne et al.	Gd2O3	Yb3+	Laser media	[106]
		Lu2O3	Yb3+
		Y2O3	Yb3+
2002	Laversenne et al.	Y2O3	Er3+	Laser media	[107]
			Ho3+
			Yb3+
2003	Boulon et al.	Lu2O3	-	Laser media	[55]
		Sc2O3	-
		Y2O3	-
2010	Saggioro et al.	Ta2O5	Eu3+	Electro-optical applications	[108]
2012	Santos et al.	ß-Ga2O3	Eu3+	Electro-optics (TCO) 1	[95]
2012	Kim et al.	Lu2O3	Ho3+	Laser media	[63]
			Yb3+
2012	Rodrigues et al.	TiO2	Cr3+:Fe3+	Electro-optics	[99]
2016	Kim et al.	Lu2O3	Ho3+	Laser media	[50]
			Yb3+
2019	Schmehr et al.	Cu2O	-	Dye, fungicide, …	[23]
2019	An et al.	Y2O3	Ho3+:Yb3+	Temperature sensing	[74]
2019	Bao et al.	Y2O3	Ho3+:Yb3+	Temperature sensing	[73]
2019	Kim et al.	Lu2O3	Ho3+	Laser media	[72]
			Yb3+

¹ TCO: Transparent Conductive Oxide.

) [105]. Thus, the last 20 years have also witnessed the LFZ growth of crystals like Gd₂O₃ doped with Yb³⁺ [106], Lu₂O₃ doped with Yb³⁺ or Ho³⁺ [50,72,106], Sc₂O₃ [55], Y₂O₃ undoped and doped with Yb³⁺, Er³⁺, or Ho³⁺ [55,106,107], and Ta₂O₅ doped with Eu³⁺ [108], and fluorides like CaF₂:Yb³⁺ [109,110] or KY₃F₁₀:Yb³⁺ [22]. It must be highlighted that fluorides have been grown under Ar atmosphere.

Likewise, other complex oxides based on 2Al₂O₃-SiO₂ (mullite), Bi₄Ge₃O₁₂ (BGO), or Lu₂SiO₅ (LSO) suitable to be used as scintillators have been grown in the last decade by this technique [111,112]. Visible light emission, highly efficient, solid-state, light-emitting materials allowing the fabrication of devices with low energy consumption, high brightness, and environmentally friendly characteristics, are presently of great interest for many researchers [52]. Thus, the LFZ or LHPG technique has allowed obtaining a wide production of phosphors and luminescent materials in the form of doped single crystals or polycrystalline and eutectic materials [52,73,74,112,113,114,115] in a fast manner. Despite this vast production, considering that this section is focused on single crystals applied as laser media or other specific optical engineering applications, we will not delve into a topic that would probably need a chapter by itself.

Other interesting kinds of materials produced, as a response to the increasing energetic demand, are those in which the optical properties are modulated by electrons, namely, electro-optics. This, together with the previously commented gallium trend to be evaporated, make the work of Santos et al., in 2012 [95], scientifically very important. They were capable of growing fibers of the transparent conductive oxide (TCO) ß-Ga₂O₃ doped with Eu³⁺, which is suitable to be employed on solar cell devices. Thus, bulk fibers were produced at relatively high speeds (10 to 30 mm/h) in air at atmospheric pressure from raw oxide powder mixtures of Ga₂O₃ and Eu₂O₃ with an excess of the former compound. More recently, Ren et al., in 2017 [116], grew polycrystalline Al₂O₃/Er₃Al₅O₁₂/ZrO₂ eutectic ceramics under a nitrogen atmosphere suitable to be used as selective thermal emitter for thermal photovoltaic (TPV) generation [117]. One year before, these researchers also grew Al₂O₃-Yb₃Al₅O₁₂ (YbAG) eutectic rods to be selective emitters for thermophotovoltaic (TPV) devices [69].

Following this study, one of the most remarkable advantages of this technique is the possibility to grow incongruent melting materials with relative ease and within short times. This way, an incongruent melting material, like TAG, suitable to be used as optical isolator, was successfully produced in a modified LFZ process. The latter entailed combining laser melting with four halogen lamps employed to reduce thermal gradients and favor TAG crystallization [30,31,118]. The first approach of this modified LFZ was presented by Sekijima et al., in 1999 [119], reporting the growth of cerium-doped Y₃Fe₅O₁₂ crystals (YIG) that can also be employed as optical isolators. Likewise, working on the LFZ apparatus development, Ito et al., in 2013 [22], were able to grow incongruently melting materials such as BiFeO₃ and (La,Ba)₂CuO₄, making use of a LDFZ apparatus that enables melting at moderate laser power values.

Finally, it is very important to highlight a couple of works reporting the LFZ growth of crystalline L-arginine phosphates (LAP), in essence, the production of crystalline structures of organic molecules [120,121]. Singh et al. in 2008 [120] reported the production of ~1-mm-diameter transparent crystalline fibers of rhodamine-6G (Rd6G)-doped L-arginine by LFZ, applying a very low laser power of 4 W at 36 mm/h growing speed over pressed rods of fully reacted precursor compounds. It is remarkable that despite the fact that LFZ is ideal for high-temperature melting materials, some authors have achieved crystalline growth for a material that presents weight losses starting at 473 K, with complete decomposition at about 523 K when it is grown by a solution crystallization method. Additionally, the fibers produced maintained transmission for up to four weeks, despite being stored without oil in normal laboratory conditions. A couple of years later, the same authors also produced other LAP crystalline fibers by LFZ [121]. Thus, potassium di-hydrogen phosphate (KDP)-doped L-arginine phosphate crystalline fibers were grown, applying ~5.4 W with a pulling speed of 194 mm/h. It must be noted that the optical quality of these phosphate crystalline fibers, however, is not comparable to those obtained for inorganic materials.

5. Materials for Electrical Applications

LFZ technique has been also extensively explored to produce materials for electrical applications, where metal oxide superconductors, thermoelectrics, and magnetic materials (

Table 5. Materials for electrical applications since 1999 [8].

Year	Authors	Material	Tech. 1 Doping	Improved Properties/Results	Ref.
2008	Silva et al.	La0.7Ca0.3MnO3	EALFZ/-	Field freezing, supercooling transition: planar to a cellular/dendritic S/L interface	[124]
2011	Amaral et al.	CaCu3Ti4O12	LFZ/-	Dielectric properties single & polycrystalline CCTO 2	[156]
2011	Sergeenkov et al.	La0.7Ce0.3MnO3	LFZ/-	Three ferromagnetic transitions at 126 K, 180 K and 300 K	[157]
2012	Ferreira et al.	Bi2Ca2Co1.7Ox	EALFZ/-	PF = 0.088 mW/K2m @ 923 K	[123]
2013	Constantinescu et al.	Bi2Ba2Co2Ox	LFZ/-	Grain alignment decrease secondary phases PF = 0.4 mW/K2m @ 923 K	[141]
2013	Madre et al.	Ca3Co4O9	LFZ/-	PF = 0.42 mW/K2m @ 1073 K, (annealing 72 h)	[142]
2013	Sotelo et al.	Bi1.6Pb0.4Sr2Co1.8Ox	LFZ/Pb & Ag	PF = 0.42 mW/K2m @ 923 K (3wt% Ag)	[143]
2014	Rasekh et al.	Bi2Ca2Co1.7Ox + Ca3Co4O9	EALFZ/-	PF = 0.18 mW/K2m @ 298 K PF = 0.3 mW/K2m @ 923 K, samples grown at 300 mA	[147]
2016	Madre et al.	Bi2−xPbxBa2Co2Oy	LFZ/Pb	ZT = 0.53 @ 650 °C, 0.2 Pb-doped	[136]
2017	Madre et al.	Bi1.6Pb0.4Ba2Co2Oy/Ag	LFZ/Ag	PF = 0.46 mW/K2m @ 923 K (annealed)	[146]
2018	Çetin Karakaya et al.	Bi2Sr2Co2Oy	LFZ/Na	Grain alignment decrease secondary phases	[144]
2018	Ferreira et al.	BaFe12Ox	LFZ/-	Fe2O3 and BaCO3 precursors: high abundance and low costs	[155]
2019	Ferreira et al.	CaMnO3	LFZ/-	PF = 0.39 mW/K2m @ 1073 K, (10 mm/h)	[154]
2019	Ferreira et al.	Ca0.9La0.1MnO3	LFZ/La	(no data)	[154]
2019	Ferreira et al.	CaMn0.93Nb0.05O3	LFZ/Nb	(no data)	[154]
2019	Özçelik et al.	Bi2Sr2Co2Oy	LFZ/K	Grain orientation decrease secondary phases ZT = 0.029 @ 400 K (x = 0.10)	[145]
2020	Carreira et al.	Ca1−xPrxMnO3	LFZ/Pr	PF = 0.303 mW/K2m @ 1123 K for (100 mm/h, air, annealing 1573 K)	[153]
2020	Ferreira et al.	Bi2Sr2Co1.8Oy	EALFZ/-	Grain orientation decrease secondary phases ZT = 0.09 @ 923 K (+300 mA)	[148]

¹ Tech.: Technique employed. ² CCTO: CaCu₃Ti₄O₁₂.

) [122,123,124,125,126,127,128,129,130,131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,146,147,148,149,150,151,152,153,154,155,156,157] particularly stand out. Indeed, LFZ is a recognized technique to grow textured materials due to the very localized heating that generates steep thermal gradients at the growth interface, keeping undercooling conditions for dendritic growth. This is particularly interesting to grow materials that exhibit crystalline structure where the growth rate is highly anisotropic. The directional growth (Figure 5), besides the favorable contribution of oriented grains to current transport, also contributes to reduce the number of grain boundaries, as the misaligned orientations are blocked during crystallization by the well-aligned dendrites [125]. Besides, it is well known that the solidification rate has a strong influence on the crystallization degree, particularly on the phase nature, size, and orientation of crystals. This is even more important in the LFZ growth, a non-equilibrium process (Figure 5) [9,126,127]. Bearing in mind the influence of the morphological characteristics on the transport performance, many studies were dedicated to the influence of growth rate on the fiber’s microstructure [128,129], despite the fact that the optimum pulling rate also depends on the thermal gradients in the melt [130,131,132]. Moreover, with the growth conditions being far from equilibrium, multiphasic materials are usually obtained and, accordingly, many studies were dedicated to the effect of post-growth thermal treatment [133,134], as well as cation substitutions [135,136], among other effects.

In the last decade, thermoelectric (TE) materials have been intensively explored by LFZ. This is because they have attracted huge attention as a very promising technology, considering their energy-conversion ability. Thermoelectric materials can directly transform heat energy into electrical power as a consequence of the well-known Seebeck effect. The transformation of a thermal gradient into electrical energy enables waste heat recovery and reduction of CO₂ production. Nowadays, the commercial applications are based on alloys and/or intermetallic materials, with high thermoelectric performances at relatively low temperatures. However, there are many limitations for their use in several applications, such as their low abundance, high costs, toxic nature, and their relatively low working temperatures associated to their intrinsic thermal degradation. These limitations were overcome by the discovery of thermoelectric properties in Na₂Co₂O₄ [137]_. This material can work at high temperatures in ambient atmosphere conditions and is free of heavy or toxic elements. Following this discovery, great efforts were made to find similar characteristics materials. Other oxides, such as Ca-Co-O, Bi-Sr-Co-O, or Bi-Ca-Co-O, with attractive thermoelectric properties were thus studied [138,139]. These layered cobaltite materials exhibit a strong crystallographic anisotropy and tend to crystallize with stacked plate-like grains, a suitable structure to be explored by a directional growth. Following this idea, the LFZ growth without and with applied electrical current (EALFZ) was executed, aiming to increase the preferential grain orientation and to produce ceramics with very high density to reduce electrical resistivity and, consequently, raise their thermoelectric performance (Figure 5).

Among the several studies carried out in p-type ceramic oxides, it is important to emphasize the work performed by Sotelo and co-workers (INMA, Instituto de Nanociencia y Materiales de Aragón; Zaragoza, Spain) that extensively explored the effect on thermoelectric performances of growth and annealing conditions [141,142], doping and co-doping [136,143,144,145], and composites’ formation [146] (Table 5). All textured samples prepared by LFZ showed remarkable increase of power factor (PF) values as compared with conventional, sintered ceramics due to the preferential grain alignment along the growth axis. Nevertheless, it is important to emphasize the high amount of secondary phases that is usually produced due to the incongruent melting behavior of these oxide materials, as well as the generation of oxygen vacancies associated to the fast solidification process [142]. The optimization of microstructural characteristics through the growth and annealing conditions are reflected in the thermoelectric performance, with a significant increase in the PF values, reaching much higher values than the ones typically obtained in misfit cobaltites processed by conventional solid-state reaction [141,142]. The highest power factor values (0.42 mW/K²m at 923 K) observed for Ca₃Co₄O₉ are a consequence of phase content, high apparent density, grain alignment, and Co³⁺/Co⁴⁺ relationship [142]. The thermoelectric performance of misfit cobaltites can be controlled not only by the grain orientation but also by cation substitutions. Following this approach, many dopants were studied in the CoO-based families, namely, by Pb incorporation [136,143,144,145]. An important improvement on the resistivity and thermopower was obtained for Bi_2−xPb_xCa₂Co_1.7O_y with 0.4 Pb substitution [143]. Likewise, exposing the as-grown samples to a post-annealing step induced an increase of thermoelectric phase content and, consequently, an improvement in PF by a factor of two at room temperature [136]. The thermal conductivity of these samples was measured and an estimated ZT value of 0.53 for the 0.2 Pb-doped samples was obtained, which is much higher than the ones reported in the literature [136]. The effect of K aliovalent substitution for Sr on the Bi₂Sr₂Co₂O_y system enabled an increase in ZT values up to 0.10 of K-content, reaching 0.029 at around 400 K, pointing to a positive effect on the formation of a spin-density wave state when K substitutes Sr in the crystalline structure [136]. Considering the importance of electrical connectivity through the grain boundaries on thermoelectric properties, the addition of metallic Ag was explored [143,146]. The performance was significantly improved after post-thermal treatment optimization (3 wt% Ag annealed samples at 923 K, 24 h), reaching a PF value of 0.32 mW/K²m at 1023 K [146].

Considering the excellent results obtained when an electrical current was applied during the crystallization of superconducting materials using the EALFZ method [122], the same approach was explored by the group of Costa et al. (Universidade de Aveiro; Aveiro, Portugal) in p-type TE materials, considering the similarity of both types of crystalline structures [123,147,148] (Table 5). The current application in the EALFZ strongly modified the phase development, the crystal morphology, and the grain alignment [29,124]. If a direct current (DC) is applied, by connecting the positive and negative poles, respectively, to the seed and feed rod samples, the system approaches equilibrium, since it drastically changes the solidification features, namely, the solute ion distribution, phase nature, and crystal growth kinetics [29,123,124,148]. These modifications arise from the electromigration phenomena introduced by the electrical current application during the growth process [29]. Consequently, the ion flux follows on the balance of the constitutional undercooling and the electromigration effect. Accordingly, this electrical field through the solid–liquid interface accelerates the solute ions until a drift velocity is reached. Thus, the effective segregation coefficient (the solute ratio concentrations in the crystallized solid at the level region and in the bulk liquidus) is significantly modified [29,127,149]. An equivalent phenomenon was observed in other systems when an electric current was applied, namely, the refinement of Bi-Sn and Au-Ge alloys [149,150].

The polarity effects of an external current applied during the growth process (EALFZ) was investigated in the Bi₂Ca₂Co_1.7O_x system [123]. A significant improvement of power factor was obtained for samples grown with the positive pole connected to the seed rod, around 50% higher than those measured on samples grown by LFZ but without current application. The main reason for this improvement was the grain orientation and the crystal size of the thermoelectric phase. In contrast, the reverse current (negative electrode connected to the seed rod) favors the non-thermoelectric phase’s development. The same behavior was observed in Bi₂Sr₂Co_1.8O_y, with the best thermoelectric performances obtained for samples grown under +300 mA, reaching a ZT value of 0.09 at 923 K [148]. Exploring the same technique, a new thermoelectric composite was successfully grown by EALFZ with a 300 mA applied current, promoting the formation of Bi₂Ca₂Co_1.7O_x + Ca₃Co₄O₉ thermoelectric phases (Figure 5) [147]. Under these conditions the Seebeck coefficient is higher than that obtained for single crystals due to the formation of oxygen vacancies and PF values of 0.18 mW/K²m at room temperature and 0.3 mW/K²m at 923 K being obtained, which are the highest reported values for Bi₂Ca₂Co_1.7O_x polycrystalline materials.

TE devices are made by joining at least two semiconductor materials together (TE module), one n-type (negative thermopower and electron carriers) and the other p-type (positive thermopower and hole carriers), wired electrically in series and thermally in parallel. This way, in addition to the development of p-type materials, n-type oxide materials have also been explored in the last decade. Transition metal oxides are gaining increasing attention for their TE properties due to their high thermal stability and tunable electronic and phonon transport properties. Among the n-type materials, SrTiO₃, CaMnO_3, and ZnO are the best candidates for TE power-generation modules operating at mid- to high-temperature ranges [151]. The high values of their Seebeck coefficients arise from either high carrier effective masses due to electronic correlations or from electron–electron interactions [152]. In the case of calcium manganite-based thermoelectrics, among the materials processed by conventional routes, Pr substitution shows high ZT values [152]. Accordingly, Pr-substituted calcium manganite was selected as a model system to study the relevant impacts of LFZ processing under different growth atmospheres on the thermoelectric performance [153]. Also, Nb and La were explored [154], putting in evidence that PF is significantly affected by the growth conditions and that additional heat treatment of the laser-processed fibers could improve TE performance. The guidelines suggest that LFZ is a suitable technique for processing thermoelectric perovskite-type manganites requiring, however, an optimization of growth and thermal annealing conditions (Table 5).

Recently, a new thermoelectric material composed of environmentally friendly elements, with nominal BaFe₁₂O_x composition, was also processed by LFZ [155] (Table 5). Its highest power factor measured at 1073 K (0.2 mW/K²m) is comparable to the best observed so far in oxide ceramic materials, with additional advantage of high abundance and low costs of Fe₂O₃ and BaCO₃ precursors.

LFZ was explored also to produce other materials for electrical applications, taking advantage of the LFZ ease to produce single crystals and polycrystalline ceramics. Thus, aiming to better understand the main mechanisms of the colossal dielectric constant of CaCu₃Ti₄O₁₂ (CCTO), several samples were grown by LFZ under different growth conditions [156] (Table 5). The results suggest that a polarization mechanism is present and should contribute toward a further increase of the dielectric constant. Moreover, the dielectric properties of this perovskite-structured material measured at microwave frequency (2.7 GHz) by the resonant cavity method confirm the high dielectric constant (56.7) and relatively low values of tan δ (0.04), putting in evidence that this material is potentially interesting for microwave device applications.

The LFZ technique was also used to produce manganites of La_0.7Ce_0.3MnO₃ (LCMO), a fascinating material that exhibits colossal magnetoresistance due to the presence of the mixed valence of manganese as Mn³⁺ and Mn⁴⁺ [157] (Table 5). The magnetic properties of the LCMO fibers grown by LFZ were evaluated and three ferromagnetic transitions were detected at 126, 180, and 300 K. The possibility to synthesize bulk electron and hole doped manganites by LFZ will have a significant impact on designing new hybrid devices that take advantage of both spin and charge degrees of freedom. As referred above, one of the major effects of the EALFZ technique is the control of the effective distribution coefficients, which approximate to the equilibrium coefficient values when the current is applied [29]. This field-modified segregation effect was demonstrated in perovskite-based manganese oxides (Ln_0.7Ca_0.3MnO₃) with colossal magnetoresistive (CMR) effect [121]. Indeed, the application of an electric field decreased the macrosegregation phenomenon, even at low cooling rates. A transition of planar to a cellular solid–liquid interface due to constitutional supercooling conditions due to field freezing effects was observed. Uda and co-workers (Institute of Materials Research, Tohoku University; Japan) also imposed an external electric field during oxide crystal growth, putting in evidence that it is possible to change an incongruent solidification into a congruent one [158]. The same behavior was observed in the case of La₃Ga₅SiO₁₄, as well as in LiNbO₃, due to the modifications of chemical potentials under the extrinsic electric field. In addition, the group of T. Duffar (University of Grenoble; Grenoble, France) put in evidence recently a significant modification of the eutectic phase diagram induced by an external field for the Al₂O₃-YAG-ZrO₂:Y system [159].

6. Materials for Applications in Superconductivity

As introduced in the previous section, texturing has been promoted in high-temperature superconducting materials, aiming to enhance superconductivity. The development of power applications based on these materials compel strong requirements on these anisotropic materials. This is more evident for the Bi-Sr-Ca-Cu-O families because of their small coherence length and a large crystallographic cell that are responsible for an extreme two-dimensional behavior. Traditional ceramic processing technologies are not adequate for obtaining bulk materials with the minimum level of performance in order to be used in large-scale electrical applications.

Early works were carried out in the YBa₂Cu₃O_x system using CO₂ lasers as heat sources [160,161,162,163]. These initial works showed that LFZ processing is not an adequate technique to obtain textured rods with high critical current values. Apart from the fact that a substantial evaporation takes place when Y₂O₃, BaCO_3, and CuO are used as precursors, it was observed that the directional solidification process was performed using a semisolid zone instead of a complete molten zone, due to the peritectic reaction that controls solidification in this material. In all the cases, growth rates lower than 40 mm/h were required to maintain a stable growth habit and the process had to be performed in controlled atmospheres, usually with oxygen partial pressures in the range between 6 and 10³ Pa.

Different studies showed that the Bi superconducting phases were more appropriate for using LFZ as a texturing technique, because they exhibit high stability during incongruent melting and low evaporation and very low deviations from the material stoichiometry [160,164,165,166,167,168]. LFZ produces a conglomerate of crystals with their crystallographic c-axis perpendicular to the sample growth direction, and the a-b planes aligned in the direction of the fiber axis, improving electrical transport current flow [125]. Bi superconductors present two phases with critical temperatures above 77 K: Bi₂Sr₂CaCu₂O_8+δ (Bi-2212) with T_c ≈ 80 K and Bi₂Sr₂Ca₂Cu₃O_10+δ (Bi-2223) with T_c ≈ 110 K.

The Bi-2223 phase has lower stability and, despite greater interest due to its higher critical temperature, it was very complicated to obtain textured materials with this composition. Initial works [167,169,170] explored the strategy of substituting some amount of Bi by Pb and the LFZ process was performed in an 8% O₂ in Ar mixture flowing through the growth chamber. In the textured material, the main phase is the Bi-2212 phase and additional annealing is required to partially recover the Bi-2223 phase. Due to the high density of the textured materials obtained by LFZ, this annealing was longer in comparison with ceramic pellets. A second alternative was to modify the stoichiometry of the precursors aiming to enhance the formation of the Bi-2223 phase. Larrea et al., in 1994 [171], used the Bi_1.87Pb_0.35Sr_1.87Ca₄Cu₆O_y composition adding an extra amount of Ag. Costa et al. proposed the use of Bi₂Sr₂Ca₂Cu₄O₁₁ [125,133] and Miao et al., in 1997 [172], proposed precursors with different proportions of Bi-2212, Bi₂Sr₂CuO₆ (Bi-2201) and CaCuO₂.

Due to the higher stability of the Bi-2212 phase, an important amount of work reported was focused on this phase [132,135,164,168,173,174,175,176,177,178,179,180,181,182,183]. During these works, the Nd:YAG radiation started to be used as an alternative to the CO₂ lasers [168,175], observing that, with similar focusing optics, 1.06-μm radiation is more adequate to process textured Bi-2212 superconductors because these oxides absorb the 10.6—μm radiation mostly in the surface, creating larger radial temperature gradients and larger radial inhomogeneities due to constitutional supercooling. These studies [125,173,174,177] showed that in as-grown samples, the superconducting phase only develops for thin samples (Φ < 250 μm) and with very low growth rates (<5 mm/h). When the growth rate increases, growth is dominated by constitutional supercooling. Under these conditions, the first solid phases formed during solidification are Sr-Ca-Cu oxides, normally the (Sr,Ca)CuO₂ (1/1 phase) and, in some cases, the (Sr,Ca)₁₄Cu₂₄O₄₁ (14/24 phase), and the melt becomes enriched in Bi₂O₃, generating some superconducting grains with intergrowths of the Bi-2212 and the Bi-2201 phases and an amorphous matrix of frozen bismuth-rich liquid phase between them. The growing habit of the 1/1 phase has the b-axis parallel to the growing direction and the c-axis along the shorter dimension in the transverse cross section [125] and this determines the texture of the superconducting material. Size and radial distributions of these phases are strongly dependent on the thermal gradients induced during the LFZ process and they can be controlled by adjusting the processing parameters, mainly the growth rate [177], the laser power, or the initial precursor stoichiometry [178]. The strong correlation between the effects of the different processing parameters complicates the optimization process. A Simplex optimization method was proposed [179] to process, at the same time, four growth parameters (laser power, the growth rate, and the precursor and textured rod rotation speeds) and four annealing parameters, taking as control variables to obtain high critical current values and short processing times. Main conclusions of the optimization process are that high laser powers originate drastic reductions in the T_c values of the superconducting grains that grow within the central part of the rod. The second main conclusion is that a two-step annealing is needed. During the first step, at 1143 K, cation diffusion takes place to form the superconducting phase. In the second one, at 1073 K, the oxygen content required to optimize the critical temperature value is established. Critical current densities of the order of 5500 A/cm² at 77 K in 1-mm-diameter rods were reached after these optimization studies. Also, it was observed that critical current values decreased linearly with sample diameter, with the magnetic field generated at the sample surface being the limiting factor.

Several works have analyzed the influence of modifying the composition of the Bi-2212 precursor in the final microstructure of the textured material. Results showed that moderate Ag additions, with contents up to 3 wt%, reduce porosity and increase electrical connectivity because Ag fills the holes between superconducting grains [180]. Also, the substitution of some cations has been explored. For instance, Ca substitution by Y ions [181] generates an increase in the critical temperature values but with a deterioration of the grain alignment, while substitution by Rb [182] or Cs [183] leads to an increase of the pinning force at 10 K by about 20%.

Aiming to increase the level of current that can be transported by these materials, some studies have been performed, aiming to increase the cross section of the textured material. Vieira et al., in 2012 [130], analyzed the thermal gradients induced in rods with diameters between 1.75 and 2.5 mm and how they affect the phase distribution in the cross section of the sample and in the superconducting properties. Natividad et al., in 2001 [184], used LFZ to induce texture in hollow cylinders with external diameters up to 10 mm. The external part of the cross section is similar to that observed in bulk samples. These samples were used to fabricate a current lead with a coaxial configuration, showing an increase in the critical current value due to the reduction of the generated magnetic fields. Laser power limits the maximum thickness of the hollow cylinder. Angurel et al., in 2009 [185], proposed a new alternative with the aim to be able to texture rods with any diameter. In this case, samples are placed on a support that contains a series of Al₂O₃ rods in constant rotation, on top of which superconductor cylinder preforms are placed. The proposed laser system was a diode laser and allowed to process several pieces in parallel.

As it has been mentioned in previous sections, the group from Universidade de Aveiro (Costa et al.) introduced a modification in the LFZ process, called EALFZ, in which a DC electrical current is applied through the solid–liquid interface during the solidification process [29,122,127,128,134,186,187]. This modification was applied to the Bi-2212 and to the Bi-2223 phases, showing excellent results in both cases. When the electrical field is applied in the direct polarization configuration (positive pole connected to the seed) solidifications deviate from metastability and the primary solidification phase is mainly the 14/24 phase, yielding thinner and longer grains in the as-grown material and an improved texture in the superconducting rod. By contrast, when a reverse polarization is used, the dendritic microstructure disappears, generating a globular structure. Applying an electrical current of 300 mA in the processing of Bi-2212/2.9 wt% Ag rods, it was possible to reach critical current values at 77 K of the order of 5800 A/cm² in samples of 2.3 mm in diameter, ~2.5 times higher than the values measured in similar samples processed without applying the electrical current during the solidification [122]. A review of the evolution of the superconducting families and the wavelength of the lasers used to process these materials is presented in

Table 6. Superconducting materials produced by LFZ.

Year	Authors	Material 1	Laser	Ref.
1988	Feigelson et al.	Bi-2212	CO2	[164]
1989	Qiao et al.	YBaCuO, Bi-2212	CO2	[160]
1989	De la Fuente et al.	Bi-2223	CO2	[166]
1989	Carillo-Cabrera et al.	Bi-2223	CO2	[169]
1989	Gazit et al.	Bi-2212	CO2	[173,174]
1992	Cima et al.	YBaCuO	CO2	[161]
1992	Snoeck et al.	Bi-2223	CO2	[170]
1994	Figueredo et al.	YBaCuO	CO2	[162]
1994	Larrea et al.	Bi-2223	CO2	[171]
1995	De la Fuente et al.	Bi-2212, Bi-2223	CO2, Nd:YAG	[175]
1997	Costa et al.	Bi-2223	CO2	[123]
1998	Diez et al.	Bi-2212, Bi-2223	Nd:YAG	[176]
1998	Angurel et al.	Bi-2212	Nd:YAG	[177]
1999	Costa et al.	Bi-2223	CO2	[133]
2004	Carrasco et al.	Bi-2223	Nd:YAG	[29]
2004	Costa et al.	Bi-2223	Nd:YAG	[186]
2004	Natividad et al.	Bi-2212	Nd:YAG	[179]
2005	Sotelo et al.	Bi-2212	Nd:YAG	[180]
2015	Costa et al.	Bi-2212	CO2	[122]
2016	Özcelik et al.	Bi-2212	Nd:YAG	[181]
2020	Özcelik et al.	Bi-2212	Nd:YAG	[182]

¹ YBaCuO: YBa₂Cu₃O_7−δ; Bi-2212: Bi₂Sr₂CaCu₂O_8+δ; Bi-2212: Bi₂Sr₂Ca₂Cu₃O_10+δ.

.

These superconducting materials were used in the development of the superconducting part of current leads [188]. In the framework of a collaboration between the INMA, the CEDEX-CIEMAT Laboratory of Applied Electromagnetics (CEDEX: Centro de Estudios y Experimentación de Obras Públicas, CIEMAT: Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas; Madrid, Spain) and the company ANTEC (Antec Inc.; Freemont, CA, USA), the company designed and fabricated a set of 600-A current leads for the LHC (Large Hadron Collider) correction magnets at CERN (European Organization for Nuclear Research) [189,190]. Each module consisted of four Bi-2212 textured bars, with a diameter of approximately 0.8 mm, connected in parallel. The module has to work at temperatures close to 50 K, in which the expected critical current can reach values close to 1200 A. This requirement revealed that one of the main limitations to developing technological applications with these materials was to obtain good electrical contacts to inject the electrical current. This called for an effort to deposit on the oxide surface a stable, uniform, and reproducible metal layer. Painting, sputter deposition, or electrodeposition processes from non-aqueous solvents [191] have been used. In all the cases, best results were obtained when the coating was fabricated in the as-grown material and the annealing of the superconducting rod was used to consolidate the electrical contact. Reproducible surface contact resistivity values lower than 10⁻⁹ Ω cm² were reached, values that allow us to use these materials in electrical applications where high transport currents are required.

7. Incongruent Melting and Volatile Materials for Additional Applications

Researchers working on LFZ have shown every so often the possibility to obtain incongruent melting materials in a more or less controllable, direct, and fast way. In most cases, this was possible thanks to high temperatures provided within controlled atmospheres and fast speeds. Lately, as was mentioned in Section 2, equipment setup advances providing high-pressure [23] or uniform radial heating by using various laser diodes [22,32] has allowed us to synthesize volatile (Li₂CuO₂, Nd₂Mo₂O₇, SrRuO₃) and incongruent melting (BiFeO₃, (La,Ba)₂CuO₄, Ba₂Co₂Fe₁₂O₂₂) materials. For example, Schmehr et al., in 2019 [23], were able to grow volatile Li₂CuO₂ by introducing a 80:20 Ar:O₂ atmosphere with a total pressure of 98.69 atm (100 bar) for avoiding volatility. In contrast, the absence of special atmospheres when processing an incongruent melt system like Nd₂O₃:SiO₂ has allowed us to synthesize unexpected materials. Thus, from stoichiometric mixtures of Nd₂O₃ and SiO₂ processed in air, a biphasic Nd₂SiO₅:Nd_9.33(SiO₄)₆O₂ material has recently been produced by Rey-García et. al., observing how the dielectric character of the stoichiometric phase rules electrical properties, making it suitable for microwave dielectric devices [192]. Likewise, for growing crystals of the pyrochlore-type Nd₂Mo₂O₇ (NMO), suitable for Hall effect sensors, only a slight pressure of 1 atmosphere is needed, as recently reported by Kaneko and Tokura, in 2020 [32], to promote the re-evaporation of MoO₂ by rapidly increasing the temperature of the molten zone. These authors have also grown SrRuO₃ (SRO), an archetypal perovskite ferromagnetic metal, which also presents the deposit of RuO₂ near the melting point. Thus, by applying 10 atmospheres of Ar with 10% O₂ at a growth rate of 7 mm/h from a raw material with a 0.05 at% excess of Ru, a stable crystal growth without fluctuations of the material rod or the molten zone was achieved. These two volatile materials were synthetized through LDFZ technique, in which the use of various lasers allows a uniform irradiation intensity distribution on the periphery of the raw material. In addition, a vertical irradiation intensity distribution can be designed to have flat-shape or bell-shape irradiations for relaxation of residual thermal strain in the grown crystal [22,32].

The LDFZ technique has been also recently applied to produce incongruent melting materials like the Y-type hexaferrite Ba₂Co₂Fe₁₂O₂₂ conventionally grown by a flux method [32]. So, single crystals of the multiferroic compound were successfully grown at 10 atmospheres and a slow growth rate of 1 mm/h from a seed crystal having the desired composition and crystal structure. Indeed, crystals of BiFeO₃ and La_2−xBa_xCuO₄ were obtained [22], as was also mentioned in Section 2, avoiding the stability problems caused by the gentle temperature gradient when growing in conventional LFZ. However, conventional LFZ has allowed us to produce other incongruent melting materials through the years [192,193]. For example, Andreeta et al., in 1999 [193], were able to grow single-crystal fibers of the dielectric SrHfO₃ in air from green rods. Its properties were found to be influenced by incongruent melting since, depending on the growth rate, surface inclusions and compositional variations were observed. Some years after, Tb₃Al₅O₁₂ was also successfully produced in a modified LFZ setup, adapted with four halogen lamps to reduce thermal gradients, favoring TAG crystallization [30,31,118], as mentioned earlier in Section 4. Likewise, iron aluminate garnet (Y₃Fe₅O₁₂) is another incongruent melting material, which was easily grown in air using conventional LFZ by Sekijima et al., in 1999 [119], and by Lim et al., in 2000 [194]. It is useful for monitoring unique magneto-optical properties in the near infrared, making it suitable as an infrared isolator, optical switch, and spatial light modulator, as well as in other sensor applications. In the beginning of the 21st century, researchers grew through the LFZ or LHPG technique other examples of incongruent materials. For example, Chen et al., in 2002 [195], successfully grew metastable BaTiO₃ crystal fibers doped with calcium oxide, namely, Ba_1−xCa_xTiO₃ (x = 0.1, 0.2). Taking advantage of their tetragonal structure, the growth from both SrTiO₃ and Ba_0.8Ca_0.1TiO₃ seeds was demonstrated. The growth of potassium lithium niobate (KLN) single crystal suitable for second harmonic generation with blue laser emission [86,87,88] may also be cited once more here, as an important example.

8. Future Perspectives

The Laser Floating Zone (LFZ) technique reviewed in this paper, also known as Laser-Heated Pedestal Growth (LHPG), could be considered as an ideal, environmentally benign method to obtain high-quality single crystals within a very short time span, while minimizing the use of starting precursors and increasing overall energy efficiency with respect to other bulk-crystal growth methods. From these points of view, it is ideal to fulfill the long-sought desire to advance the understanding of fundamental physico-chemical phenomena at a fast pace. In essence, it complies perfectly with the message transmitted in a recent sentence by Schmehr et al., in 2019 [23]: “The availability of pristine single crystals is essential to the discovery of new physical phenomena in condensed matter physics”. Perhaps the latter is a good indication that LFZ is essential toward the development of an important part of the future awaiting solid-state chemistry and physics, as well as materials science and the technological developments that follow from these.

Taking into account exclusively the growth of congruently melting materials in single-crystal cylinder or fiber form, LFZ provides, to a first approximation, important advantages regarding the use of small quantities of materials, crucible-free melt containment, and fast growth rates. The quality of the crystals obtained thus far has proven sufficient for their use in demonstrators and in high-technology devices. Crystals with good optical quality for laser resonators are a good example of their demonstrated use and future potential. For example, ultra-fast mode-locking lasers producing picosecond (10⁻¹² s) and femtosecond (10⁻¹⁵ s) high-energy pulses at multi-GHz repetition rates [196] can be applied in a wide range of applications that include nonlinear optics, telecommunications, nanomachining, medical surgery, and environmental monitoring [197,198,199,200]. Likewise, this method is adequate to achieve high-quality crystals of small dimensions for compact and portable instruments, as discussed in Section 4. LFZ has enabled fabrication of high-quality photonic materials suitable to be used as optical isolators, amplifiers, or modulators, waveguides by embedding into adequate cylinders, scintillators, luminescent materials, temperature sensors, electro-optics, photorefractive materials, or selective emitters for silicon photovoltaic cells, among others.

More recent developments related to the use of different types of lasers, particularly diode lasers that may be deployed in various focusing configurations, have provided significant improvements on heating homogeneity and melt uniformity [32]. These affect not only the crystal quality, but potentially reduce the undesired appearance of microcracks, which make the crystal useless, or fail under further stress during use. In addition, these novel configurations enable the growth of incongruently melting phases, such as those considered in previous sections above, as well as phases that require particular atmosphere control. Examples of these include incongruently melting Ba₂Co₂Fe₁₂O₂₂, grown under O₂ atmosphere [32], and highly volatile Nd₂Mo₂O₇ and SrRuO_3, grown under Ar and Ar:O₂ atmospheres, respectively. An extreme case is the growth of the known volatile Li₂CuO₂ compound, achieved within a high-pressure (HP-LFZ) apparatus [23] and an 80:20 Ar:O₂ atmosphere. Thus, the demonstrated growth of non-congruent melting [26,27,28,29,82,83,114,188,189,190,191,192] or volatile [22,28,29,188] materials in a controlled, direct, and fast manner paves the way to explore the fabrication of crystal elements for already relevant, expected, and unforeseen applications in high technology.

Focusing on difficult-to-grow materials, we must highlight the recent work of Lai et al., in 2018, who studied how to design a nonlinear, hybrid, crystal-glass 3D metamaterial fiber with sub-wavelength fine structures [104]. They grew a large-scale harmonic crystal (Si⁴⁺:γ-Al₂O₃) from a sapphire (α-Al₂O₃) crystal monolithically integrated into silica tubes. A hybrid crystal-glass metamaterial fiber was thus produced in a two-step process. Laser irradiation was followed by rapid cooling, to induce an inter-diffusion process where nucleation of large-sized harmonic crystals in anisotropic faceting was achieved. These open the possibility for attaining monolithically integrated dendrites for intracavity and resonant second harmonic generation (SHG) [104].

In the last decade, new, clean, and sustainable power-generation methods have been intensively investigated in attempts to control global warming. Innovative sustainable development is now accepted as a way of progress toward an improved environment, for which the design and development of novel materials able to be used in a plethora of applications becomes an urgent necessity. In this sense, a breakthrough in the efficiency of thermoelectric (TE) materials will have an important technological and economic impact on the global energy balance. This may include novel approaches to develop advanced thermoelectric oxides, by in situ defects’ engineering and nanostructuring, promoted by controlled redox reactions. These innovative aspects can be explored by LFZ processing under strongly non-equilibrium conditions for designing desired functional thermoelectric properties [29,122,123,124,137,138,139,140,141,142,143,144,145]. Moreover, the use of an externally applied current during the growth process (EALFZ) is a powerful tool to design and control microstructure, thus to improve transport properties. Particular attention needs to be given to the fabrication of prototypes based on LFZ-processed thermoelements, considering not only the present geometry achieved for the LFZ-obtained TE materials.

High-temperature superconducting ceramics have been among the most commonly studied materials with LFZ [29,122,125,133,160,161,162,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182], probably because their anisotropic structure is ideal for directional solidification processing. As reviewed above, LFZ has demonstrated excellent microstructural transformations that result in high critical current elements in rod and tube geometries. These have been demonstrated to be competitive in several applications, particularly in superconducting current leads. These are used to connect electric power between room temperature and cryogenic temperatures, necessary for high-magnetic-field superconducting magnet operation, present for example in MRI machines, among other devices.

A most relevant advancement for superconductor processing is based on the EALFZ technique [29,122,127,128,134,186,187], which provides an additional control of texture, yields excellent results in terms of critical current values, and opens new frontiers to study the fundamental phenomena that stand behind controlled diffusion during directional solidification at large.

Future developments of LFZ within the superconductor materials domain should probably center at least around three aspects. First of all, on applying further and improving our understanding of electrically assisted phenomena during melting and solidification. Secondly, in exploring texturing of large-area, high-T_c coatings and films, which may find a number of attractive applications, including in attractive devices driven by levitation. Thirdly, modifying precursor, heating, and solidification geometries in order to improve the scalability of the method for industrial production.

A least explored aspect of LFZ entails the possibility of producing organic complex molecules in the form of crystalline fibers [120,121]. As suggested by the results reviewed above, a wide range of opportunities to explore bio-compatible applications may be envisioned. In addition, the production on geometries suitable to be employed, for example, in photonic devices, highly reduces the engineering, mechanical, and economical cost associated when an organic-based compound must be integrated into a commercial photonic device. These types of materials need intensive studies in order to establish a proper knowledge base for LFZ-processed organic compounds, and future efforts will determine the potential of the technique to address the intrinsic scientific and technical problems involved, particularly with respect to competing fabrication methods.

Finally, a universal advantage that may always be considered for the future evolution of LFZ is, as commented earlier, its facile, low cost, and turnaround productivity of an ample collection of materials in single-crystal form and with convenient size for physical property characterization and demonstration prototypes. These would apply to wide-open disciplines of science and technology. The technique may gain relevance if consideration is given to miniaturization of apparatus and the achievement of higher mechanical precision. Together with air-cooled, small-size, and high-beam-quality lasers recently available at lower cost and high durability, LFZ may end up being incorporated into many laboratories worldwide for quick turnaround and convenient materials’ processing exploration purposes. It will certainly contribute to the advancement of both fundamental understanding of physical phenomena and developments of new devices.