Direct synthesis of colloidal nanocrystals in a molten inorganic salt solvent

Abstract

Methods of synthesizing nanocrystals are provided which comprise combining a metal molecular precursor comprising a transition metal element or a group III element, and a group V molecular precursor comprising a group V element, in a molten inorganic salt solvent comprising a molten inorganic salt and under conditions to form a metal-V compound in the form of nanocrystals. No organic solvent is required. Also provided are colloids comprising single-crystalline nanocrystals composed of a metal-V compound. This includes single-crystalline III-V nanocrystals that exhibit band edge photoluminescence upon illumination with light under room temperature.

Description

Atty. Dkt. No.05400-0073-PCT DIRECT SYNTHESIS OF COLLOIDAL NANOCRYSTALS IN A MOLTEN INORGANIC SALT SOLVENT CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] The present application claims priority to U.S. provisional patent application number 63 / 562,873 that was filed March 8, 2024, the entire contents of which are incorporated herein by reference. REFERENCE TO GOVERNMENT RIGHTS

[0002] This invention was made with government support under 2019444 and 2318105 awarded by the National Science Foundation. The government has certain rights in the invention. BACKGROUND

[0003] Nanometer scale crystals of semiconductor materials demonstrate size-dependent physical and chemical properties which can be exploited for displays, lasers, photodetectors, and solar cells. These materials also show promise for emerging applications in quantum information science, high-performance photonics, and photocatalysis. Successful commercial applications of quantum dots have proven that colloidal synthesis, which is inherently scalable, is competitive in terms of material quality with much more expensive semiconductor wafer and thin-film technologies such as molecular beam epitaxy (MBE) and chemical vapor deposition (CVD). To date, nearly perfect control over size, shape, and luminescent properties has been achieved for colloidally synthesized II-VI semiconductors (Zn, Cd and Hg chalcogenides) and IV-VI semiconductors (Pb chalcogenides). However, there are still critical omissions in the list of success stories for colloidal semiconductors, particularly gallium pnictides such as GaAs, GaN, and gallium-containing ternary and quaternary III-V materials. SUMMARY

[0004] Provided are methods for synthesizing nanocrystals, including III-V nanocrystals and transition metal-V nanocrystals. The methods involve synthesizing the nanocrystals from molecular precursors in a molten inorganic salt solvent without requiring any organic solvent. The Examples below illustrate use of the present methods to provide highly crystallineAtty. Dkt. No.05400-0073-PCT nanocrystals (including III-V nanocrystals such as GaAs nanocrystals) exhibiting superior optical properties, including room temperature band edge photoluminescence.

[0005] An embodiment 1 is a method of synthesizing nanocrystals, the method comprising comprising combining a metal molecular precursor comprising a group III element or a transition metal element, and a group V molecular precursor comprising a group V element, in a molten inorganic salt solvent comprising a molten inorganic salt and under conditions to form a metal-V compound in the form of nanocrystals, wherein the metal molecular precursor, the group V molecular precursor, and the molten inorganic salt solvent form a reaction mixture that does not comprise an organic solvent.

[0006] An embodiment 2 is according to embodiment 1, wherein the metal molecular precursor is a group III molecular precursor comprising the group III element and the metal- V compound is a III-V compound.

[0007] An embodiment 3 is according to embodiment 2, wherein the group III molecular precursor acts as a reducing agent to reduce the group V element of the group V molecular precursor under the conditions to form the III-V compound.

[0008] An embodiment 4 is according to any of embodiments 2-3, wherein the group III molecular precursor is a molten inorganic salt.

[0009] An embodiment 5 is according to any of embodiments 2-4, wherein the group III molecular precursor has formula MmXn, wherein M is the group III element, X is a halogen, m is in a range of from 1 to 2, and n is in a range of from 1 to 5, and the group V molecular precursor is a pnictogen halide or a pnictogen hydride.

[0010] An embodiment 6 is according to embodiment 5, wherein the group III molecular precursor has formula M[MX4], wherein M is the group III element and X is the halogen.

[0011] An embodiment 7 is according to embodiment 5, wherein the group III molecular precursor has formula M[M2X6], wherein M is the group III element and X is the halogen.

[0012] An embodiment 8 is according to embodiment 5, wherein the group III molecular precursor has formula MX, wherein M is the group III element and X is the halogen.

[0013] An embodiment 9 is according to any of embodiments 1-8, wherein the molten inorganic salt solvent comprises a molten inorganic salt having formula AM’X’4, wherein A is an alkali metal, M’ is a group III element, and X’ is a halogen; an alkali metal halide; a eutectic mixture of alkali metal halides; or a combination thereof.Atty. Dkt. No.05400-0073-PCT

[0014] An embodiment 10 is according to embodiment 2, wherein the group III molecular precursor has formula M[MX4], wherein M is the group III element and X is a halogen; the group V molecular precursor is a pnictogen halide or a pnictogen hydride; and the molten inorganic salt solvent comprises a molten inorganic salt having formula AM’X’4, wherein A is an alkali metal, M’ is a group III element, and X’ is a halogen; a eutectic mixture of alkali metal halides; or both.

[0015] An embodiment 11 is according to embodiment 10, wherein M is selected from Ga, In, Al, and combinations thereof, and X is selected from Cl, Br, and I.

[0016] An embodiment 12 is according to any of embodiments 10-11, wherein the group V molecular precursor has formula PnH3, PnX”3or Pn2X2”, wherein Pn is the group V element and X” is a halogen.

[0017] An embodiment 13 is according to embodiment 12, wherein Pn is selected from N, P, As, Sb, and combinations thereof, and X” is selected from Cl, Br, and I.

[0018] An embodiment 14 is according to embodiment 10, wherein the group III molecular precursor is Ga[GaI4]; the pnictogen halide is PI3, AsI3, SbI3, or a combination thereof; the pnictogen hydride is NH3; and the molten inorganic salt solvent comprises KGaI4, KGaCl4, a eutectic mixture of alkali metal halides, or a combination thereof.

[0019] An embodiment 15 is according to embodiment 2, wherein the group III molecular precursor has formula MX3, wherein M is the group III element and X is a halogen and further wherein the group V precursor is a silylated pnictogen.

[0020] An embodiment 16 is according to any of embodiments 1-15 and 17-20, wherein the conditions include a reaction temperature of greater than 425 ºC.

[0021] An embodiment 17 is according to embodiment 1, wherein the metal molecular precursor is a transition metal molecular precursor comprising the transition metal element and the metal-V compound is a transition metal-V compound.

[0022] An embodiment 18 is according to embodiment 17, wherein the transition metal molecular precursor has formula MXn, wherein M is selected from Ti, V, Nb, Mo, Ta, and W; X is a halogen; and n is from 3 to 6.

[0023] An embodiment 19 is according to any of embodiments 17-18, wherein the group V molecular precursor is a pnictogen hydride.Atty. Dkt. No.05400-0073-PCT

[0024] An embodiment 20 is according to embodiment 19, wherein the pnictogen hydride is NH3.

[0025] An embodiment 21 is according to any of embodiments 1-20, wherein the nanocrystals of the metal-V compound are single-crystalline nanocrystals.

[0026] An embodiment 22 is according to any of embodiments 2-16, wherein the nanocrystals of the III-V compound exhibit band edge photoluminescence upon illumination with light under room temperature.

[0027] An embodiment 23 is a colloid comprising single-crystalline nanocrystals composed of a III-V compound, the single-crystalline nanocrystals exhibiting band edge photoluminescence upon illumination with light under room temperature.

[0028] An embodiment 24 is according to embodiment 23, wherein the III-V compound has Ga, and optionally, one or more of Al and In, as a III element; and N, P, As, Sb, or a combination thereof as a V element.

[0029] An embodiment 25 is according to embodiment 24, wherein the III-V compound is GaAs or GaN.

[0030] Other principal features and advantages of the disclosure will become apparent to those skilled in the art upon review of the following drawings, the detailed description, and the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Illustrative embodiments of the disclosure will hereafter be described with reference to the accompanying drawings.

[0032] FIGS.1A-1C: Molten salt redox chemistry. (FIG.1A) Relationship between molten salt redox chemistry and III-V nanocrystals controlling their chemical stability. (FIG. 1B) Qualitative electrochemical scale for key reactions governing stability of the pnictide component of the III-V nanocrystals compared with redox reactions of III-halide molten salt components. (FIG.2C) Molten salt redox chemistry for precursor activation during nanocrystal synthesis.

[0033] FIGS.2A-2G: Direct synthesis of GaAs nanocrystals in molten salt solvents. (FIG.2A) Powder X-ray diffraction patterns of colloidal GaAs NCs synthesized from Ga[GaI4] and AsI3 in molten CsI / NaI / KI eutectic at temperatures from 425-500 °C. SampleAtty. Dkt. No.05400-0073-PCT synthesized at 400 °C was prepared using a KGaI4 solvent. The (*) peaks originated from x- ray scattering from the organic ligand shell. TEM image of colloidal GaAs nanocrystals synthesized at (FIG.2B) 450 °C and (FIG.2C) 500 °C. Inset photos of the same samples as colloidal solutions in toluene. (FIG.2D) Raman spectra of colloidal GaAs nanocrystals with reference GaAs Raman modes denoted at dotted lines. (FIG.2E) Room temperature photoluminescence from GaAs nanocrystals synthesized from 425-500 °C with inset photo of GaAs nanocrystals synthesized at 425 °C under UV illumination. (FIG.2F) Calculated size dependent optical gap and PL peak position for GaAs nanocrystals compared to InP, GaP, and CdSe. (FIG.2G) Room temperature PLE spectra recorded at different emission wavelengths for GaAs nanocrystals.

[0034] FIGS.3A-3D illustrate colloidal characterization of GaAs nanocrystals. The first plot in each figure shows the small angle x-ray scattering (SAXS) with model fit; the second plot in each figure shows the extracted volume-scaled distribution with extracted size and size distribution information resulting from maximum-entropy fits to the data; the third plot in each figure shows the dynamic light scattering (DLS) size histograms. FIG.3A corresponds to the 425 °C, 1 hour sample of FIG.2A. FIG.3B corresponds to the 450 °C, 1 hour sample of FIG.2A. FIG.3C corresponds to the 475 °C, 1 hour sample of FIG.2A. FIG. 3D corresponds to the 500 °C, 1 hour sample of FIG.2A. The small peak at larger sizes in the second plots may correspond to “dynamic aggregation,” which is indicative of a well behaving nanocrystal colloid. Regarding the third plots, DLS measures the hydrodynamic size of the particles (which includes ligands), rather than the size of the inorganic core. The length of an extended oleyl group is about 2 nm, and hydrodynamic nanocrystal size is larger than the size measured by SAXS (the second plots).

[0035] FIGS.4A-4C demonstrate the estimation of the Stokes shifts of GaAs nanocrystals (as synthesized in FIG.2A at 425 °C, 450 °C, 475 °C, and 500 °C). FIG.4A shows PLE spectra collected with a detection wavelength 20 nm blue of the PL peak maximum for samples with different sizes. The lines represent the maximum of the PLE spectrum at the 1stexciton (left) and the detection wavelength (right). FIG.4B plots estimated Stokes shifts for GaAs nanocrystals as a function of size as determined by Scherrer analysis of XRD patterns. FIG.4C shows a comparison of the measured stokes shift and the stoke shift calculated using atomistic electronic structure calculations.Atty. Dkt. No.05400-0073-PCT

[0036] FIGS.5A-5B show a comparison of linear absorption and transient absorption signals in from the GaAs nanocrystals as synthesized in FIG.2A. FIG.5A shows a comparison of the bleach signal 500 fs after photoexcitation with the linear absorption spectrum. FIG.5B shows a comparison of the bleach signal 2 ps after photoexcitation with the linear absorption spectrum. Linear absorption and TA spectra were collected on identical samples for all cases, except for GaAs synthesized at 450°C. In this case, the samples were from different batches synthesized under identical conditions

[0037] FIG.6 shows another illustrative synthetic procedure for the direct synthesis of GaPn colloidal quantum dots (CQDs) in molten salt solvent.

[0038] FIG.7A shows Raman spectra of GaAs CQDs synthesized at different temperatures. FIG.7B shows UV-Vis spectrum (solid line) and fluorescence spectrum (dashed line) of in-situ HF etched GaAs CQDs synthesized at 500 ℃.

[0039] FIGS.8A-8F each show PXRD plots and TEM images of colloidal nanocrystals synthesized using the present methods, including TiN nanocrystals (FIG.8A), VN nanocrystals (FIG.8B), GaN nanocrystals (FIG.8C), NbN nanocrystals (FIG.8D), Mo2N nanocrystals (FIG.8E), and Ta3N5 nanocrystals (FIG.8F).

[0040] FIG.9A shows room-temperature photoluminescence from GaN semiconductor nanocrystals synthesized at 425° to 575°C and the inset is an image of colloidal GaN nanocrystals in methylcyclohexane under ultraviolet illumination. FIG.9B shows Raman spectra of GaN nanocrystals and CVD-grown GaN. FIG.9C shows the absorption spectra of plasmonic TiN nanocrystals. FIG.9D shows magnetic susceptibility of NbN nanocrystals as a function of temperature. DETAILED DESCRIPTION

[0041] Provided are methods for synthesizing nanocrystals. The nanocrystals comprise a metal element and a group V element. The metal element may be a transition metal element. Various transition metal elements may be used, such as a first row transition metal element, e.g., Ti, V; a second row transition metal element, e.g., Nb, Mo; or a third row transition metal element, e.g., Ta, W. The metal element may be a group III element, e.g., e.g., Al, Ga, In, or Tl. In embodiments, the group III element is Ga or In. The group V element may be, e.g., N, P, As, Sb, or Bi. In embodiments, the group V element is N, P, As, or Sb. The presentAtty. Dkt. No.05400-0073-PCT methods are further described below using III-V nanocrystals as an illustrative type of metal- containing nanocrystal that may be synthesized.

[0042] The methods comprise combining a group III precursor comprising a group III element and a group V precursor comprising a group V element in a molten inorganic salt solvent. The group III precursor is a chemical compound comprising the group III element (i.e., group III molecular precursor) and the group V precursor is a chemical compound comprising the group V element (i.e., a group V molecular precursor). The molten inorganic salt solvent comprises a molten inorganic salt. The combined group III precursor, group V precursor, and the molten inorganic salt solvent may be referred to as a reaction mixture, which is subjected to conditions to produce III-V nanocrystals therein. The conditions include those which induce covalent bond formation between the group III element and the group V element to form a III-V compound as well as conditions which allow for crystalline (versus amorphous) growth. By contrast to existing techniques for synthesizing III-V nanocrystals, the present methods synthesize the desired III-V nanocrystals directly from molecular precursors (e.g., versus pre-synthesized “seed” nanocrystals) and no organic solvents are required.

[0043] Desirably, the group III precursor does not comprise oxygen. In embodiments, the group III precursor does not comprise any carbon-hydrogen bonds. In embodiments, the group III precursor does not comprise carbon, does not comprise hydrogen, or both. In embodiments, the group III precursor does not comprise a group V element.

[0044] The group III precursor may be a molten inorganic salt. The phrase “molten inorganic salt” refers to inorganic salts that do not comprise any carbon-hydrogen bonds. The phrase further refers to inorganic salts having a melting point (Tm) below a reaction temperature being used in the present methods, e.g., below 600 ºC or below 550 ºC. This includes having a Tmin a range of from above room temperature (about 20 to 25 ºC) to less than 450 ºC, from above room temperature to less than 350 ºC, from 50 ºC to less than 300 ºC, from 65 ºC to 275 ºC, from 80 ºC to 200 ºC, and from 95 ºC to 150 ºC. The molten inorganic salt may be in its ionized form (i.e., a blend of its constituent cations and anions) at the reaction temperature being used.

[0045] The group V precursor is distinct from the group III precursor, i.e., they are different chemical compounds. Desirably, the group V precursor does not comprise oxygen. In embodiments, the group V precursor does not comprise any carbon-hydrogen bonds. InAtty. Dkt. No.05400-0073-PCT embodiments, the group V precursor does not comprise carbon, does not comprise hydrogen, or both.

[0046] Selection of the group III and group V precursors being used to form the III-V nanocrystals may be guided by a consideration of relative redox potentials, e.g., the relative reduction potentials of the group III and group V precursors. In embodiments, a group III and group V precursor pair is selected such that the group III precursor is a stronger reductant as compared to the group V precursor. The group III precursor is further a sufficiently strong reductant so that it can facilitate reduction of the group V element of the group V precursor to form the desired III-V compound. In other words, the group III and group V precursor pair is selected on the basis of the ability of the group III precursor to serve as a reducing agent for the reduction of the group V element of the group V precursor. Consideration of relative redox potentials is illustrated in FIG.1B for various group V reactions (left) and various group III molten inorganic salt reactions (right).

[0047] Illustrative group III precursors include group III halides. These include those having formula MmXn, wherein M is the group III element (e.g., any of those described herein, e.g., Ga, In, Al); X is a halogen (e.g., any of those described herein, such as Cl, Br, or I); m is in a range of from 1 to 2; and n is in a range from 1 to 6 or 1 to 5. It is noted that m and n may be, but need not be, integers; m and n may be any real number between the recited values. There may be a single group III element or more than one (e.g., 2) different group III elements. In other words, mixed metal group III halides may be used. The value of m and n depend upon the selected M and its oxidation state. In embodiments, the group III precursor has formula MX3(m = 1, n = 3, and M and X are as defined above.) In embodiments, the group III precursor is GaI3. Such group III halides (MX3) are useful for embodiments of the present methods based on dehalosilylation reactions to generate III-V nanocrystals. Another suitable class of group III halides has formula M[MX4] (i.e., M2X4, m = 2, n = 4, and M and X are as defined above). In embodiments, the group III precursor is Ga[GaI4]. Another suitable class of group III halides has formula M2[M2X6] (i.e., M2X3, m = 2, n = 3, and M and X are as defined above). In embodiments, the group III precursor is Ga2[Ga2I6]. Such group III halides (M[MX4] and M2[M2X6]) are useful for embodiments of the present methods which harness redox reactions to generate the III-V nanocrystals. Another suitable class of group III halides is the monohalides, i.e., n = 1. This includes, e.g., InX, wherein X is as defined above. Regarding mixed metal group III halides, these include, e.g., Ga[AlX4], In[GaX4], In[AlX4], wherein X is as defined above. As noted above, m and n need not beAtty. Dkt. No.05400-0073-PCT integers. This includes, for example, the group III precursor having formula M2X4.6, wherein M and X are defined above. In embodiments, the group III precursor is Ga2I4.6.

[0048] Illustrative group V precursors include silylated pnictogens, e.g., tris(trimethylsilyl) pnictogen (TMS)3Pn, wherein Pn is the pnictogen (group V element, e.g., N, P, As, or Sb). Such group V precursors are useful for embodiments of the present methods based on dehalosilylation reactions to generate the III-V nanocrystals. Another suitable class of group V precursors includes pnictogen halides, e.g., PnX3(wherein Pn and X are as defined above). In embodiments, the group V precursor is selected from PI3, AsI3, SbI3, and combinations thereof. Such group V precursors are useful for embodiments of the present methods which harness redox reactions to generate the III-V nanocrystals. Pnictogen halides also include Pn2X4, e.g., P2I4. As with the group III halides described above, pnictogen halides may include a single group V element or more than one different group V element (e.g., 2). Another suitable class of group V precursors includes pnictogen hydrides, e.g., PnH3(wherein Pn is as defined above). In embodiments, the group V precursor is NH3 (ammonia). As with NH3, in embodiments, the group V precursor may be in its gaseous phase under the conditions being used in the method.

[0049] A single type of group III precursor or multiple, different types of group III precursors may be used. Similarly, a single type of group V precursor or multiple, different types of group V precursors may be used. The use of multiple, different types of precursors, e.g., multiple, different types of group V precursors, enables the synthesis of ternary, quaternary, pentanary, etc. III-V nanocrystals, in addition to binary III-V nanocrystals.

[0050] The following group III precursors may be excluded, in embodiments: polymeric gallium imide ((Ga(NH)3 / 2)n), group III azides (Et2Ga(N3), (N3)2Ga[(CH2)3NMe2], (Et3N)Ga(N3)3), group III acetates, group III amidos (Ga2[N(CH3)]6, group III cupferrons. The following group V precursors may be excluded, in embodiments: alkali nitrides (Li3N), hexamethyl disilazane, As(NMe2)3, lithium bis(trimethylsilyl amide).

[0051] The molten inorganic salt solvent provides the medium in which the chemical reactions occur to form the III-V compound. The molten inorganic salt also provides colloidal stability to the formed III-V nanocrystals. As noted above, the molten inorganic salt solvent comprises a molten inorganic salt. Molten inorganic salts have been generally defined above. Since the group III precursor itself may be a molten inorganic salt, the group III precursor alone can provide the molten inorganic salt solvent. However, in other embodiments, otherAtty. Dkt. No.05400-0073-PCT types of molten inorganic salts (distinct from the group III precursor) are used or included in the molten inorganic salt solvent. In addition to the molten inorganic salts described above, suitable molten inorganic salts include those having formula AMX4, wherein A is an alkali metal (e.g., Li, Na, K, or Cs), M is a group III element (e.g., Ga or In) and X is a halogen (e.g., Cl, Br, or I). Molten inorganic salts also include those having formula AX, wherein A is an alkali metal and X is a halogen as described above. Eutectic mixtures of inorganic salts may be used to provide a molten inorganic salt, e.g., those based on alkali metal halides, e.g., CsI / NaI / KI, CsBr / NaBr / KBr, CsCl / NaCl, KBr / InBr3; KBr / GaBr3, etc. In the present disclosure, a eutectic mixture may be referred to as a “molten inorganic salt” which may have any of the Tmdescribed above. Other molten inorganic salts which may be used include those as described in U.S. Pat. No.11,040,323 and U.S. Pat. No.11,247,914, each of which is hereby incorporated by reference in its entirety.

[0052] Salt additives (which may be referred to as surfactants) may be included in the molten inorganic salt solvent. These salt additives may be provided by the inorganic salts of a eutectic mixture. A single type, or multiple, different types of molten inorganic salts may be used to provide the molten inorganic salt solvent.

[0053] As noted above, the present methods do not require the use of any organic solvent to synthesize the III-V nanocrystals. This includes not requiring use of a biphasic mixture comprising a molten salt phase and an organic liquid phase such as that disclosed in International Patent Publication WO 2024 / 044525. Thus, both the molten inorganic salt solvent and the reaction mixture may each be characterized as being free of an organic solvent, i.e., an organic chemical compound comprising carbon-hydrogen bonds. Organic solvents which may be excluded include, e.g., alkylamines, trioctylphosphine, octadecene, squalane, etc. However, these embodiments do not preclude those in which the group V precursor is a silylated pnictogen compound. Finally, desirably, the molten inorganic salt solvent and the reaction mixture are each free of oxygen, including oxygen-containing compounds.

[0054] In embodiments, the molten inorganic salt solvent comprises (or consists) of one or more molten inorganic salts having formula AMX4, one or more eutectic mixtures of inorganic salts, or both. In each of these embodiments, the group III precursor itself may be a molten inorganic salt, which may be considered to be part of the molten inorganic salt solvent.Atty. Dkt. No.05400-0073-PCT

[0055] It is to be understood that at an initial time point in the present methods, e.g., upon initial combination of the group III and group V precursors in the molten inorganic salt solvent, the reaction mixture does not comprise any III-V nanocrystals. In addition, the reaction mixture formed in the present methods is generally free of any pre-synthesized III-V nanocrystals, including III-V nanocrystals which may have been synthesized using other methods.

[0056] The conditions used to carry out the present methods include parameters such as selection of materials (group III precursor, group V precursor, molten inorganic salt solvent), amounts thereof, reaction temperature and time, and atmosphere. In general, these parameters are tuned to facilitate nucleation and crystalline growth of a desired III-V nanocrystal. Further guidance for selection of materials has been described above.

[0057] Regarding reaction temperature, generally the temperature is above the Tmof the selected molten inorganic salt solvent (or components thereof). Thus, the molten inorganic salt solvent is in its liquid phase during the formation of the III-V nanocrystals. In addition, as described in Examples, below, the temperature may be increased to ensure the crystallinity of the III-V nanocrystals. (See FIG.2A.) Illustrative temperatures include those greater than 400 ºC, at least 425 ºC, at least 450 ºC, at least 475 ºC, at least 500 ºC, at least 525 ºC, at least 550 ºC, at least 575 ºC, or in a range of between any of these values, e.g., from 400 ºC to 600 ºC, from 450 ºC to 600 ºC, from 425 ºC to 500 ºC, or from 525 ºC to 565 ºC. Specific illustrative reaction temperatures and illustrative reaction times are provided in the Examples, below, but include, e.g., from a few minutes to a few hours.

[0058] Regarding amounts, illustrative amounts are provided in the Examples, below. However, by way of illustration, the group III precursor and the group V precursor may be included in the reaction mixture at a ratio (in equivalents) in a range of from 3:1 to 15:1. This includes from 5:1 to 13:1, and from 7:1 to 11:1. If multiple group III precursors and / or multiple group V precursors are used, amounts of each may be selected so as to achieve a desired specific alloy composition. Assuming an additional molten inorganic salt is used to provide the molten inorganic salt solvent (versus using a group III molten inorganic salt precursor alone as the molten inorganic salt solvent), the molten inorganic salt solvent may make up from 0 weight% to 50 weight% of the reaction mixture. This includes up to 35 weight%, up to 25 weight%, and up to 15 weight%. By “weight%” it is meant (weight of the molten inorganic salt solvent (excluding any group III molten inorganic salt precursor)) / (totalAtty. Dkt. No.05400-0073-PCT weight of reaction mixture)*100. If salt additives are included in the molten inorganic salt solvent, their amounts may be adjusted to provide desired size and shape control.

[0059] Regarding atmosphere, conditions may be selected to ensure the absence of O2 / air and H2O as the steps of the method are carried out, e.g., by maintaining the reaction mixture under vacuum or inert gas (N2or Ar).

[0060] For embodiments in which the group V precursor is a gas, e.g., NH3, partial pressure is a relevant parameter which may be tuned to facilitate nucleation and crystalline growth. As described in Example 3, below, the pressure may be selected to ensure the crystallinity of the III-V nanocrystals. Illustrative partial pressures include at least 0.1 MPa, at least 0.8 MPa, at least 2.0 MPa, at least 2.5 MPa, at least 2.8 MPa, at least 3.0 MPa, at least 3.5 MPa, at least 4.0 MPa, at least 4.5 MPa, at least 5.0 MPa, at least 6.0 MPa, or a range of between any of these values, e.g., from 2.0 MPa to 5.0 MPa or from 2.0 MPa to 4.0 MPa .

[0061] The present methods need not, but may include other steps. For example, the III-V nanocrystals may be recovered from the reaction mixture and, if desired, redispersed as a colloid. Illustrative details for such recovery are provided in the Examples, below. As another example, shells may be grown over the III-V nanocrystals. Illustrative shell-growth techniques are described in U.S. Pat. No.11,040,323 and U.S. Pat. No.11,247,914, each of which is hereby incorporated by reference in its entirety. However, the methods need not include additional steps (e.g., annealing) to achieve III-V nanocrystals having the crystalline quality as described below.

[0062] As described above, rather than a group III element, the metal element of the present nanocrystals may be a transition metal element, e.g., Ti, V, Nb, Mo, Ta, W. The description above applies to such embodiments except that the group III precursor may be a transition metal precursor, the group III halide may be a transition metal halide (e.g., TiI4, VCl3, NbCl5, MoCl5, TaCl5, WCl6), and the “M” in formulas of the group III halide may be a transition metal. These embodiments of the present methods provide transition metal-V nanocrystals rather than III-V nanocrystals.

[0063] The nanocrystals produced by the present methods may be characterized by their composition, which depends upon the selected precursor(s) as well as the conditions described above. In embodiments, the nanocrystals are binary nanocrystals, i.e., having a single type of metal element and a single type of V element. However, the nanocrystals are not limited to binary nanocrystals, e.g., as multiple, different types of metal precursors and / orAtty. Dkt. No.05400-0073-PCT group V precursors may be used, e.g., to provide ternary nanocrystals, quaternary nanocrystals, pentanary nanocrystals, etc. Illustrative III-V nanocrystals include gallium- based nanocrystals such as GaAs nanocrystals, GaP nanocrystals, GaSb nanocrystals, GaN nanocrystals, GaPAs nanocrystals, GaAsSb, InGaN, AlGaN, and InGaP nanocrystals. In the ternary, pentanary, etc. nanocrystals, it is understood that the ratios of the various elements may vary, e.g., GaP1-yAsy, wherein y = 0.01 to 0.99. This includes y = 0.25 to 0.75, 0.30 to 0.70, 0.35 to 0.65, and 0.40 to 0.60. Illustrative transition metal-V nanocrystals include transition metal nitrides such as TiN, VN, NbN, Mo2N, Ta3N5, W2N, and TixV1-xN. Again, x may vary, e.g., x = 0.01 to 0.99.

[0064] The nanocrystals are generally free of oxygen, which may be confirmed using elemental analysis and / or X-ray diffraction analysis as described in the Examples, below.

[0065] The nanocrystals produced by the present methods may be further characterized by their size / shape, crystallinity, colloidal stability, and optical properties as described immediately below. Although this description is provided using III-V nanocrystals as an illustrative example, relevant description may also be applied to the transition metal nanocrystals.

[0066] The III-V nanocrystals may be further characterized by their size and shape. Regarding size, the largest cross-sectional dimension of the nanocrystals is not greater than 1000 nm and is generally significantly smaller, including sufficiently small so as to ensure quantum confinement, e.g., no greater than 100 nm, no greater than 50 nm, no greater than 25 nm, or no greater than 10 nm. This includes largest cross-sectional dimensions in a range of from 1 nm to 50 nm, from 1 nm to 25 nm, and from 1 nm to 10 nm. These dimensions may refer to the average largest cross-sectional dimension for a collection of nanocrystals. Nanocrystals exhibiting quantum confinement may be referred to as quantum dots (QDs). Regarding shape, the nanocrystal shape may be, e.g., spherical, cubic, pyramidal, tetrahedral, etc. For any of these shapes, the dimensions above may refer to a diameter of the nanocrystal or a largest width. In some embodiments, the shape is elongated, e.g., rods. For these shapes, the dimensions above may refer to a diameter or width. As noted above and demonstrated in the Examples, below, reaction temperature and time as well as use of salt additives may also be selected to achieve a desired nanocrystal size and shape.

[0067] As is clear from the term “nanocrystals,” the III-V nanocrystals are crystalline in nature, i.e., the III and V atoms are arranged in an ordered lattice, by contrast to amorphousAtty. Dkt. No.05400-0073-PCT materials exhibiting a lack of such atomic ordering. Crystalline quality may be confirmed and quantified using X-ray diffraction analysis and Raman spectroscopy as described in the Examples, below. Regarding X-ray diffraction analysis, III-V nanocrystals formed using the present methods exhibit powder XRD (PXRD) patterns such as those shown in FIGS.2A and FIGS.8A-8E. By contrast, PXRD patterns of amorphous materials lack peaks and / or have broad peaks.

[0068] Regarding Raman spectroscopy, III-V nanocrystals formed using the present methods exhibit distinct LO and TO modes and overtone peaks such as those shown in FIG. 2D (above about 425 °C) and 7A (above about 425 °C). Data such as that shown in these figures is a clear signature of the high-quality crystalline nature of samples synthesized at 425 °C and above. This is by contrast to the 400 °C curves in FIG.2D and 7A showing significantly broader peaks at the LO and TO positions (and such peaks are not distinguishable) and no overtone peaks. Moreover, data such as that shown in FIG.2D further confirms that the III-V nanocrystals are free of point defects and / or other crystal imperfections not detectable via X-ray diffraction analysis. The single-crystalline nature of the III-V nanocrystals may be further evidenced by TEM and HRTEM images. As noted above and demonstrated in the Examples, below, the conditions being used in the present methods, e.g., reaction temperature, may be selected to ensure high-crystallinity, including single-crystallinity.

[0069] Related to the properties of the III-V nanocrystals described above, is the ability of the nanocrystals to form a colloid, i.e., a stable, homogenous, and uniform dispersion of the individual nanocrystals (versus aggregates thereof) within a continuous phase, e.g., in a non-polar solvent such as dimethylformamide, toluene, etc. This includes the ability of the nanocrystals to form colloids in molten media as described in U.S. Pat. No.11,040,323 and U.S. Pat. No.11,247,914, each of which is hereby incorporated by reference in its entirety. Organic capping ligands, e.g., oleylamine, may be included in forming the colloids, but may not be necessary if molten media is used as the continuous phase. The ability to form colloids is another distinguishing feature of the III-V nanocrystals as compared to amorphous and polycrystalline III-V products. FIGS.3A-3D demonstrate techniques that may be used to confirm the colloidal nature of the III-V nanocrystals (in this embodiment, GaAs nanocrystals), including small angle x-ray scattering (SAXS) (first and second plots in FIGS. 3A-3D) and dynamic light scattering (DLS). (third plots in FIGS.3A-3D). For example, as shown in the first plots in FIGS.3A-3D, a low q scaling which asymptotically approaches 1Atty. Dkt. No.05400-0073-PCT at small q is observed. This is consistent with a structure factor of ~1 and evidence of isolated particles throughout the fluid medium, i.e., a colloid. In addition, in the second plots in FIGS. 3A-3D, the existence of the small peak at larger sizes is evidence of “dynamic aggregation,” which is indicative of a colloid.

[0070] The present III-V nanocrystals may be further characterized by their optical properties. As demonstrated in FIG.2E for GaAs nanocrystals, the III-V nanocrystals exhibit room temperature photoluminescence. The photoluminescence can even be observed when the III-V nanocrystals are excited under low power photoexcitation (4 mW 450 nm laser) by eye in a dark room.

[0071] Moreover, and as further described in the Examples, below, other experiments have revealed the remarkable finding that the photoluminescence behavior of the III-V nanocrystals is dominated by band edge carrier dynamics. These experiments include FIG. 2G showing room temperature photoluminescence excitation (PLE) spectra recorded at different emission wavelengths for GaAs nanocrystals. The shift of the first peak in the PLE spectra to longer wavelengths as the detection wavelength is increased is consistent with band edge emission. Additional evidence of band edge emission is the excellent agreement between the measured emission maximum and the calculated emission maximum from atomistic electronic structure calculations as shown in FIG.2F. By contrast, if the emission was coming from a trap state, the measured emission maximum would be at a lower energy than the calculated value. Additional evidence of band edge emission is based on the measurement of the Stokes shift using PLE spectroscopy as shown in FIGS.4A-4B. These results show that the Stokes shift ranges from 20-60 meV depending on the particle size; size dependent Stoke shift is an indicator of band edge emission. As shown in FIG.2C, the observed size dependent stokes shift also correlates with the calculated stokes shift using atomistic calculations, further confirming that the emission originates from a band edge state rather than a trap state. Finally, as shown in FIGS.5A-5B, the excellent correspondence between linear absorption spectra (dotted lines) and the photoinduced bleach signals (solid lines) provides even further evidence that photoluminescence observed from the GaAs nanocrystals originates from a band edge state. Similar photoluminescence behavior indicative of band edge emission was observed for GaN nanocrystals as described in Example 3, below.Atty. Dkt. No.05400-0073-PCT

[0072] To the inventors’ knowledge, the experiments described above provide the first known example of band edge emissive colloidal gallium-V nanocrystals. These results are believed to be directly related to synthesis of the III-V nanocrystals using the unique methods described herein. As used herein, the phrase “band edge emissive colloidal III-V nanocrystals” may be used to refer to III-V nanocrystals synthesized using the present methods that have the crystallinity (including single-crystallinity), colloidal nature, and photoluminescence behavior as described above and demonstrated using the techniques described above (and in the Examples below). Single-crystallinity and the colloidal nature of the III-nanocrystals may be confirmed using PXRD, Raman spectroscopy, TEM, HRTEM, SAXS, and DLS as described above. PLE may be used to confirm band edge photoluminescence as described above.

[0073] Regarding band edge photoluminescence upon illumination with light under room temperature, this may be confirmed by the III-V nanocrystals exhibiting one or more (including all) of: a peak in PLE spectra that shifts to longer wavelengths as detection wavelength is increased (as per FIG.2G); agreement between measured PLE emission maximum and calculated PLE emission maximum from atomistic electronic structure calculations (as per FIG.2F); a size-dependent Stokes shift (as per FIG.4B); and correspondence between linear absorption spectra and photoinduced bleach signals (as per FIGS.5A-5B). The wavelength of the light being used to induce band edge photoluminescence depends upon the nanocrystals, e.g., visible light may be used for GaAs nanocrystals; ultraviolet light may be used for GaN nanocrystals.

[0074] Regarding transition metal-V nanocrystals, some such nanocrystals (e.g., TiN) may be characterized as being plasmonic as confirmed by absorption spectra exhibiting one or more plasmonic resonance peaks (see FIG.9C for TiN nanocrystals). Thus, the phrase “plasmonic transition metal-V nanocrystals” may be used to refer to such nanocrystals synthesized using the present methods that have the crystallinity (including single- crystallinity), colloidal nature, and absorption behavior as described above and demonstrated using the techniques in Example 3, below. Other such nanocrystals (e.g., NbN), may be characterized as being superconducting (Meissner effect) as confirmed by magnetic susceptibility measurements as shown in FIG.9D for NbN nanocrystals. Thus, the phrase “superconducting transition metal-V nanocrystals” may be used to refer to such nanocrystals synthesized using the present methods that have the crystallinity (including single-Atty. Dkt. No.05400-0073-PCT crystallinity), colloidal nature, and magnetic susceptibility behavior as described above and demonstrated using the techniques in Example 3, below.

[0075] The nanocrystals synthesized using the present methods may be used in a variety of applications, e.g., optoelectronic devices, high-power electronics, and microelectromechanical systems. The nanocrystals themselves (including colloids thereof) are also encompassed by the present disclosure. EXAMPLES

[0076] Example 1

[0077] The experimental results presented in this Example serve to uncover new dimensions for colloidal chemistry in molten salts. The redox chemistry of molten salt solvents is rich and control of redox processes can be leveraged to stabilize III-V nanocrystals against decomposition. New redox chemistry was utilized to nucleate and grow monodisperse GaAs, GaP, GaP1-yAsy, and GaAs1-ySby nanocrystals directly in molten salts and shape control was demonstrated using a molten salt “surfactant”. Finally, size-dependent band edge PL was demonstrated for colloidal GaAs quantum dots.

[0078] Materials and Methods

[0079] General Considerations:

[0080] Commercial Chemicals: Gallium metal (99.99% trace metals basis), and anhydrous solvents (hexane, toluene, ethanol (EtOH), acetonitrile (MeCN), methyl acetate) were purchased from Sigma Aldrich and used as received. Oleylamine (OAm, technical grade, 70%) and Oleic acid (OA, 90%) were purchased from Sigma Aldrich and degassed under dynamic vacuum at 110 °C for several hours before storage in a nitrogen glovebox. Potassium iodide (ultra-dry, 99.998%), potassium bromide (ultra-dry, 99.9%), indium bromide (anhydrous 99.9%), indium(I) chloride (99.995%) and N,N-dimethylformamide (DMF, anhydrous 99.9%) were purchased from Alfa Aesar and used as received. Tris(trimethylsilyl) phosphine ((TMS)3P, 98%, stored frozen), tris(dimethylamino)arsine (99%), and indium(III) chloride (anhydrous, 99.999%), arsenic chloride (99.999%-As) PURATREM, antimony chloride (99.999%-Sb) PURATREM were purchased from Strem Chemicals and used as received. All quartz tubing and rods were purchased from Technical Glass Products.Atty. Dkt. No.05400-0073-PCT

[0081] Considerations for Chemical Handling: In general, salt precursors were preferred, which were delivered as ~10 mesh beads in sealed ampoules over fine mesh powders since the lower surface area minimized contamination with moisture and residual glovebox solvents. The beads were ground into fine powders as needed in oven dried mortar and pestle. Throughout this Example, Coors™ porcelain mortars and pestles were used to grind salt mixtures. The porosity of the mortars and pestles made it imperative that they were dried for at least 24 hr at 100 °C and immediately transferred into a glovebox while still hot. Throughout this Example, the salts in general were handled using gloveboxes where solvents and numerous other chemicals were stored and handled; however, all work was performed while the glovebox was under continuous nitrogen purge (< 1 ppm O2and <1 ppm H2O) to ensure minimal solvent vapor contamination, and work was not performed while solvents were in active use.

[0082] Warnings: Caution: Reactions in sealed quartz ampoules with volatile reagents (e.g., iodine, GaI3, GaBr3, InBr3, InI3, PI3, AsI3, SbI3) can build up pressure. If the reaction is heated to a high temperature too quickly, rupture of the quartz tube can occur. Appropriate safety measures including performing the reactions with appropriate wall thickness for the tube diameter, in a well-ventilated fume hood, and with safety shielding is imperative. Caution: Oxyhydrogen torches must be operated with the appropriate safety precautions including proper storage of the reactive gasses, proper installation of flash arrestors on the gas delivery tubing, and proper eye protection from the intense UV and IR light produced by the quartz. When using cryogens with oxyhydrogen flames, avoid using vacuum insulated glass Dewars since thermal shock from accidental flame exposure can lead to shattering of the Dewar. Caution: AsCl3 and PCl3, are volatile and represent significant inhalation hazards. Work must be performed in a properly exhausted glovebox and / or well-ventilated fume hood.

[0083] Synthesis of Salt Precursors:

[0084] Synthesis of GaI3: Gallium(III) iodide (ultra-dry, 99.999%) was purchased from Alfa Aesar, or Gallium(III) iodide (anhydrous, 99%) was purchased from Strem Chemicals or prepared from the elements. To prepare gallium iodide from the elements, gallium and iodine were combined in a glovebox with a 1.1:3 molar ratio in a quartz ampoule consisting of 12.75 mm OD 10.5 mm ID quartz tubing with a 10 mm OD quartz plug welded on one end. A second sealing plug was inserted, and the quartz tube was attached to a vacuum transferAtty. Dkt. No.05400-0073-PCT chuck and transferred to a vacuum line outside the glovebox. The Quartz tube was immersed in liquid nitrogen contained in a closed cell foam Dewar (Chem Glass CG-1590-F500), so the iodine was completely immersed in the cold well and sealed under vacuum. The contents of the tube were slowly (over several hours) heated to 350 °C in a vertical melting furnace and held for several hours. Periodically, the ampoule was partially inverted to mix the contents. After complete reaction, the contents were distilled in the sealed quartz tube held vertically in a 400 °C to room temperature gradient where solidified gallium iodide collected in the upper cold zone. Sometimes, the tube would become completely closed with solidified gallium iodide. When this happened, the tube was removed from the heat source and cooled to room temperature. The tube was inverted, and the purified product was melted with a torch and allowed to collect further down the tube. After this, contents were allowed to solidify again, and the distillation process was resumed. The slight excess of gallium resulted in a residue of reduced gallium halides in the bottom of the ampoule but ensured no oxidizing iodine impurities were present in the products. The products were collected by breaking the ampoule open inside a nitrogen glovebox and grinding into a coarse powder.

[0085] Synthesis of InI3: Indium (III) iodide (ultra-dry 99.99%) was purchased from Alfa Aesar or prepared from the elements by slowly heating indium metal and iodine in sealed quartz ampoules at 350 °C until complete reaction.

[0086] Synthesis of KGaI4: KGaI4 was prepared by loading “ultra-dry” KI and “ultra-dry” GaI3in a 1:1 molar ratio into an oven dried quartz ampoule (9 mm outer diameter, 7 mm inner diameter, with an indentation to support a sealing plug above the sample), and a 6 mm diameter quartz plug was inserted. The ampoule was attached to a vacuum transfer chuck to mount the ampoule on a vacuum manifold without air exposure. The ampoule was sealed under vacuum using a O2 / H2torch. Next the salts were melted in a vertical furnace at ~400 °C, and after the salt melted the ampoule was inverted several times followed by additional annealing for several hours to ensure complete reaction. The reaction was cooled to room temperature, brought into a N2glovebox and the ampoule was broken open in a mortar, the quartz pieces were mechanically separated, and the resulting salt was ground into a coarse powder and stored in a glovebox for future use. Similar procedures were used to prepare KInI4 and KInBr4.

[0087] Synthesis of Ga[GaI4]: Ga[GaI4] was prepared by reacting gallium metal and GaI3 according to the equation 2Ga+4GaI3^3 Ga[GaI4] in flame sealed quartz ampoules. TheAtty. Dkt. No.05400-0073-PCT reagents were slowly heated to 400 °C (over the course of several hours) and annealed overnight until the gallium was completely dissolved in the molten GaI3. Larger scale reaction required longer reactions times, and periodic partial inversion of the ampoule was required to effectively mix the reagents and decrease the necessary reaction time. After complete reaction, the quartz tube was brought into a N2 glovebox and the ampoule was broken open in a mortar, the quartz pieces were mechanically separated, and the resulting salt was ground into a coarse powder and stored in a glovebox for future use. Similar procedures were used to prepare Ga2I3, In2I4, In2Br4, In2Br3. Reaction temperature was carefully chosen such that the boiling point of the III-halide was not excessively exceeded. Further, slow heating to the reaction tube allowed for partial conversion (and subsequent reduction in vapor pressure) before proceeding to higher temperatures for full conversion.

[0088] Synthetic Details for Direct Synthesis of Nanocrystals in Molten Inorganic Salts

[0089] Direct Synthesis of GaAs Nanocrystals in Lewis Basic Molten Salt Solvents: In a glovebox, 1 g of Ga[GaI4] (1.55 mmol, 6 equivalents), and 2 g of CsI / NaI / KI eutectic were ground into a fine powder using a mortar and pestle. Next, 116 mg of AsI3(0.25 mmol, 1 equivalent) was added to the mortar and pestle and incorporated into the salt mixture by grinding. The finely ground powder was divided into 68-inch-long oven-dried 9 mm OD, 7 mm ID quartz ampoules with an indent in the quartz such that an added quartz plug resulted in a 6-inch-long reaction vessel. A vacuum transfer chuck was affixed to the ampoule, and the sample was transferred to a vacuum line outside the glovebox without air exposure. The sample was placed under vacuum and the quartz ampoule was flame sealed using an oxyhydrogen torch. Samples were annealed using a custom-made heating block which consisted of a 4-inch diameter cylindrical aluminum block that had eight 3 / 8-inch diameter 2.5-inch-deep holes drilled in a circle pattern around a central thermocouple hole. The block was wrapped in high temperature heating tape (Briskheat BHW) and several layers of insulation. The block was heated using a PID temperature controller. The heating block was placed atop an orbital shaker unit. The heated block was preheated to the desired reaction temperature, and the flame sealed quartz ampoules which contained the reaction mixture were placed in one of the holes, and the orbital shaker unit was turned on to provide ~150 RPM of shaking. After the desired reaction time had passed, the shaking was stopped, and the ampoule was removed from the heating block and allowed to cool to room temperature naturally. The ampoule was brought into a glovebox, broken open, and the solid salt plug wasAtty. Dkt. No.05400-0073-PCT dissolved in 5 ml of anhydrous DMF by stirring at room temperature for several hours. The suspension was centrifuged and the supernatant containing the dissolved salts was decanted, leaving behind the insoluble nanocrystals. Next, the NC solids were treated with 2 ml of a 25 mg / ml GaI3in anhydrous DMF solution for ~20 min, then 2 ml of anhydrous MeCN was added. The suspension was centrifuged to collect the insoluble nanocrystals, and the supernatant was decanted. The nanocrystals were suspended in 3 ml of MeCN and centrifuged again to collect the insoluble nanocrystals. After that, the nanocrystals were treated with 1 ml of a solution containing 1 ml of toluene and 100 µl of OAm. To this, an additional 500 µl of a 5 mg / ml solution of ZnCl2 in toluene with 100 µl / ml of OAm was added. The nanocrystals were allowed to sit in this mixture for ~20 min, and after that the nanocrystals were precipitated with 3 ml of anhydrous methyl acetate, the nanocrystals were collected by centrifugation, and the supernatant was decanted away. Finally, the nanocrystals were dissolved in 500 µl of toluene and centrifuged to remove insoluble byproducts and non- colloidal nanocrystals. Samples were stored in a nitrogen glovebox and fresh aliquots were removed for further characterization.

[0090] Large Scale Direct Synthesis of GaAs Nanocrystals in Lewis Basic Molten Salt Solvents: Identical reaction conditions to those described above were used. The only modification was 12.75 mm OD / 10.5 mm ID quartz tubing was used. The reaction performed used 3 g of Ga[GaI4], 6 g of CsI / NaI / KI eutectic, and 353 mg of AsI3 loaded in a single quartz tube. This tube was heated at 500 °C for 1 hour and recovered using identical methods as used for smaller scale reactions. For recovery from the molten salt after reaction, solvent volumes used were scaled appropriately. The large size of the solidified salt matrix resulting in slow dissolution of the salt matrix (overnight) compared to smaller scale reactions (~1 hour).

[0091] Direct Synthesis of GaAs Nanocrystals in Lewis Acidic Molten Salt Solvents: In a glovebox, 1 g of Ga[GaI4] (1.55 mmol 6 equivalents) and 2 g of KGaI4 were ground into a fine powder using a mortar and pestle. Next, 116 mg of AsI3 (0.25 mmol, 1 equivalent) was added to the mortar and pestle and incorporated into the salt mixture by grinding. The finely ground powder was divided into 6 quartz ampoules, sealed under vacuum as described above, and annealed at the desired temperature as described above. To recover the nanocrystals, the ampoule was brought into a glovebox and broken open, and the solid salt plug was dissolved in 5 ml of anhydrous MeCN by stirring at room temperature for several hours. The suspension was centrifuged, and the supernatant containing the dissolved salts was decanted, leaving behind the insoluble nanocrystals. Next, the NC solids were suspended a second timeAtty. Dkt. No.05400-0073-PCT in anhydrous MeCN. The suspension was centrifuged to collect the insoluble nanocrystals, and the supernatant was decanted. After that, the nanocrystals were treated with 1 ml of a solution containing 1 ml of toluene and 100 µl of OAm. The nanocrystals were allowed to sit in this mixture for ~20 min, and after that the nanocrystals were precipitated with 3 ml of anhydrous methyl acetate, the nanocrystals were collected by centrifugation, and the supernatant was decanted away. Finally, the nanocrystals were dissolved in 500 µl of toluene and centrifuged to remove insoluble byproducts and non-colloidal nanocrystals.

[0092] Direct Synthesis of GaAs Nanocrystals in Open Tube Reactors: In a glovebox, the desired molten salt reaction was prepared by grinding together the reagents in their appropriate ratios. In all cases reagent concentration and reaction sizes identical to those used in sealed tube reactions were used. The finely ground salt precursors were loaded into an 8 inch long, 7 mm ID, 9 mm OD oven-dried quartz tube which was sealed on one end. A valved adapter consisting of a 316-stainless steel 3 / 8 in Ultra-Torr O-ring fitting, a ball valve, and a KF25 flange was affixed to the quartz tube using the Ultra-Torr fitting and the valve was placed in the closed position. The reaction mixture was removed from the glovebox and the KF25 flange was used to attach the reaction to a vacuum / inert gas manifold (Schlenk line). The connection tubing was evacuated and refilled with nitrogen 3 times before opening the ball valve to put the sample under inert gas (N2). Next the reaction was plunged into a pre-heated aluminum heating block at the desired reaction temperature and the reaction was allowed to proceed for the desired reaction time. After reaction the sample was removed from the heating block and allowed to cool to room temperature. The ball valve was closed, the valve assembly was disconnected from the Schlenk line, and the reaction was transferred back into the glovebox. The purification procedures used were identical as those used for sealed tube reactions.

[0093] Direct Synthesis of GaAs Nanocrystals in Mixed Halide Molten Salt Solvents: In a glovebox, 1 g of Ga[GaI4] (1.55 mmol, 6 equivalents) and 2 g of CsI / NaI / KI were ground together in an oven dried mortar and pestle. Next, 116 mg of AsI3(0.258 mmol, 1 equivalent) was ground into the mixture resulting in a salt mixture that weighed 3.116 g. To this, 3 mg of a CsBr / KBr / NaBr eutectic was added to the salt mixture and fully incorporated by grinding in the mortar and pestle to create a reaction mixture which contained 0.1 wt.% of the Br- eutectic.500 mg of this mixture was loaded into an ampoule and set aside. Now the mortar contained ~2.5 g of salt which contained 0.1 wt.% Br- eutectic, and to this, 5 mg of Br- eutectic was added to achieve a total of 0.3 wt.% Br- eutectic added to the salt mixture, andAtty. Dkt. No.05400-0073-PCT again 500 mg of this mixture was loaded into an ampoule. This process was repeated to achieve 0.6 wt.%, 1 wt.%, 3 wt.%, and 6 wt.% Br- eutectic additions. A similar process was used to add Cl- to the solvents. In this case, a CsCl (0.89) / KCl (0.11) mixture was used, which had an identical “molar mass” (mmol of Cl / mass of salt eutectic) as the Br- eutectic described above. An identical progression of 0.1 wt.%, 0.3 wt.%, 0.6 wt.%, 1 wt.%, 3 wt.%, and 6 wt.% was prepared for the Cl- eutectic addition. In both cases, the samples were annealed at 450 °C for 1 hour.

[0094] Samples were recovered in a way to regain colloidal stability without additional sources of non-iodide halides so it could be ensured that all Br or Cl measured came from the molten salt annealing. After annealing, the quartz ampoules were broken open in a glovebox and the salt pellet was dissolved in 5 ml of anhydrous DMF. The insoluble nanocrystals were recovered by centrifugation and the supernatant was discarded. Next, the nanocrystal pellet was suspended in 2 ml of a 25 mg / ml solution of GaI3in DMF and was allowed to sit for several minutes. To this, 3 ml of MeCN was added, the nanocrystals were collected as a pellet by centrifugation, and the supernatant was discarded. To this solid, 1 ml of a solution containing 100 µl of oleylamine in 1 ml of toluene was added and allowed to sit for several minutes. The nanocrystals were precipitated using MeOAc, collected by centrifugation as a pellet, and the supernatant was discarded. Finally, 500 µl of toluene was added to dissolve the nanocrystals, and the sample was centrifuged to remove insoluble byproducts and non- colloidal nanocrystals, and the resulting colloid was stored in a vial in a glovebox for later use.

[0095] Direct Synthesis of GaP Nanocrystals in Lewis Acidic Molten Salt Solvents: In a glovebox, 500 mg of Ga[GaI4] (0.775 mmol 6 equivalents) 1 g of KGaI4, 200 mg of KgaCl4 and 52 mg of PI3were ground together in a mortar and pestle. The resulting fine powder was divided into two quartz ampoules which were sealed under vacuum as described above and annealed at 500 °C for 1 hour. The samples were recovered using identical methods for GaAs synthesized in Lewis acidic molten salts.

[0096] Direct Synthesis of GaP Nanocrystals in Lewis Basic Molten Salt Solvents: In a glovebox 1 g of Ga[GaI4] (1.55 mmol 6 equivalents), 2 g of CsI / NaI / KI eutectic, 200 mg of a CsCl / NaCl mixture (0.89:0.11 mol), and 102 g of PI3 (0.25mmol 1 equivalent) were ground in a mortar and pestle to form a homogenous mixture. The finely ground powder was divided into 2 quartz ampoules, sealed under vacuum as described above, and annealed at 500 °C forAtty. Dkt. No.05400-0073-PCT 1 hour. The samples were recovered using identical methods for GaAs synthesized in Lewis basic molten salt solvents.

[0097] Direct Synthesis of GaP1-xAsx (x=0.5) Nanocrystals in Lewis Basic Molten Salt Solvents: In a glovebox, 500 mg Ga[GaI4] (0.775 mmol, 6 equivalents) and 1 g of CsI / NaI / KI were ground together in an oven dried mortar and pestle. Next, 28 mg of AsI3(0.0625 mmol, 0.5 equivalents) and 25 mg of PI3 (0.0625 mmol, 0.5 equivalents) were ground into the mixture. The salt mixture was divided into 3 quartz ampoules and were flame sealed under vacuum. Samples were annealed at 425 °C, 450 °C and 475 °C for 30 minutes. Samples were recovered using identical procedures as for GaAs nanocrystals synthesized in Lewis basic molten salts as described above.

[0098] Direct Synthesis of GaSb in Lewis Acidic Molten Salt Solvents: In a glovebox, 666 mg of Ga[GaI4] (1.03 mmol, 6 equivalents) and 1.33 g of KGaI4were ground together in an oven dried mortar and pestle Next, 85 mg of SbI3 (0.169 mmol, 1 equivalent) was ground into the mixture. This salt mixture was loaded into 4 separate quartz ampoules and flame sealed under vacuum. Samples were annealed at 200 °C, 250 °C, 300 °C, and 400 °C for 1 hour. Samples were recovered using identical procedures as for GaAs nanocrystals synthesized in Lewis acidic molten salts as described above; however, the resulting materials did not make stable colloids, so the resulting insoluble materials were used for powder X-ray diffraction and Raman spectroscopy.

[0099] Direct Synthesis of GaSb in Lewis Basic Molten Salt Solvents: In a glovebox, 500 mg of Ga[GaI4] (0.775 mmol, 6 equivalents) and 1 g of CsI / NaI / KI were ground together in an oven dried mortar and pestle. Next, 62.7 mg of SbI3 (0.125 mmol, 1 equivalent) was ground into the mixture. This salt mixture was loaded into 2 separate quartz ampoules and flame sealed under vacuum. Samples were annealed at 425 °C and 500 °C for 30 minutes. Samples were recovered using identical procedures as for GaAs nanocrystals synthesized in Lewis basic molten salts as described above; however, the resulting materials did not make stable colloids, so the resulting insoluble materials were used for powder X-ray diffraction and Raman spectroscopy.

[0100] Direct Synthesis of GaAs1-xSbx (x=0.5) Nanocrystals in Lewis Basic Molten Salt Solvents: In a glovebox, 500 mg Ga (0.775 mmol, 6 equivalents) and 1 g of CsI / NaI / KIwere ground together in an oven dried mortar and pestle. Next, 28 mg of AsI3 (0.0625 mmol, 0.5 equivalents) and 31 mg of SbI3(0.0625 mmol, 0.5 equivalents) were ground into theAtty. Dkt. No.05400-0073-PCT mixture. This salt mixture was loaded into 3 separate quartz ampoules and flame sealed under vacuum. Samples were annealed at 425 °C, 450 °C and 475 °C for 30 minutes. Samples were recovered using identical procedures as for GaAs nanocrystals synthesized in Lewis basic molten salts as described above.

[0101] Colloidal Atomic Layer Deposition (cALD) of ZnSe on In1-xGaxP1-yAsyNanocrystals: Colloidal ALD was performed at high temperatures according to established procedures with modifications as described below. Lithium selenidefrom lithium triethylborohydride and selenium pellets according to established procedures. All procedures were performed in a nitrogen glovebox, all solvents were anhydrous, and octadecene (ODE) and Oleylamine (OAm) were degassed at 120 °C for several hours before they were brought into the glovebox. To heat the samples, a heating block on a hot plate was use with wells that snugly fit the vials used for reaction. The temperature was measured by a thermocouple immersed in ODE contained in an identical vial to those used for the reaction, and that vial was placed in a well of the heating block. For all layers, growth temperature of 150 °C was used. The following procedure was used to deposit a single layer and can be sequentially repeated to grow thicker layers. A vial was prepared with 5-7 mg of Li2Se, 100 µl of OAm, and 500 µl of ODE. The quantum dots dissolved in toluene were added to this vial, and the vial was placed in a pre-heated heating block on a hot plate at 150 °C. The sample was cooled to room temperature and centrifuged to remove solid Li2Se. The remaining nanocrystals in liquid were transferred to a second centrifuge tube, and 1 ml of methyl acetate and 100 µl of acetonitrile was added to precipitate the particles. The particles were collected by centrifugation and the supernatant was discarded. The nanocrystal solid was dissolved in 200 µl of toluene. A vial containing 20-30 mg of zinc acetate, 100 µl of OAm, and 500 µl of ODE was prepared, and the Li2Se treated particles in toluene were added. The vial was heated at 150 °C for 2-3 minutes. The vial was cooled to room temperature and transferred to a centrifuge tube.1 ml of methyl acetate and 100 µl of acetonitrile were added to precipitate the nanocrystals, and they were collected by centrifugation. The supernatant was discarded, and the nanocrystal solid was dissolved in toluene and stored for future characterization. Additional layers can be deposited by repeating this cycle.

[0102] Nanocrystal Structural Characterization:Atty. Dkt. No.05400-0073-PCT

[0103] Powder XRD: Powder x-ray diffraction patterns were collected on a Rigaku miniflex x-ray diffractometer. Samples were deposited on 511 Si low background substrates. Lattice parameter and Scherrer size were determined by fitting the {111}, {220}, and {311} peaks to pseudo-Voigt functions to extract the peak width and position. The gallium composition was estimated by calculating the lattice parameter using the {111}, {220}, and {311} peaks and using a linear interpolation of the lattice parameters for the parent binary compounds.

[0104] TEM: Low magnification TEM images were collected on either an FEI Tecnai T20 TEM operated with a W or LaB6filament operated at 200 kV using a Gatan RIO 16 IS camera with drift correction enabled or on an FEI Tecnai F30 microscope at 300 kV using an ultrascan 4000 camera. HRTEM images were collected on a FEI Tecnai F30 microscope at 300 kV using an ultrascan 4000 camera. Images were processed using FIJI (ImageJ) to rotate images and calculate FFT of HRTEM images. Samples were imaged on 400 mesh copper grids with amorphous carbon support (EMS CF400-Cu).

[0105] Scanning transmission electron microscopy (STEM): Aberration-corrected STEM was conducted with a Thermo Fisher Scientific Titan Themis at an accelerating voltage of 300kV. Samples were prepared by drop casting diluted solutions of the nanocrystals in toluene onto TEM grids (Ted Pella, Prod # 01824; ultrathin carbon film on lacey carbon support film, 400 mesh, Cu) followed by drying under vacuum. To minimize oxygen exposure, the samples were prepared in a glovebox and transported to the microscopy facility in a sealed container. Prior to STEM imaging, the samples were further dried under vacuum at 120°C overnight. Energy dispersive x-ray spectroscopy (EDS) data was collected with a Super-X quad EDS detector. EDS analysis was done in the Thermo Fisher Scientific Velox software.

[0106] Small Angle X-ray Scattering: SAXS patterns were collected from 2 mm diameter glass capillaries of colloidal nanocrystals dissolved in anhydrous toluene. The glass capillaries were loaded inside a glovebox and the ends were sealed with UV-cure epoxy. Samples were measured using a SAXSLAB Ganesha instrument with a Cu K alpha source.

[0107] X-ray Fluorescence: X-ray fluorescence (XRF) analysis was performed with a benchtop Rigaku Energy Dispersive NEX DE VS X-ray fluorimeter equipped with a Peltier cooled FAST SDD Silicon Drift Detector. All analyses were carried out under He atmosphere to increase sensitivity for lighter elements. Elemental ratios were determined using theAtty. Dkt. No.05400-0073-PCT standardless thin films fundamental parameters method as programmed in QuantEZ software provided by Rigaku, using the standard Rigaku calibration protocols. All samples were measured on a PTFE substrate to avoid the large Si Kα peak which interferes with the aluminum and phosphorus Kα peak.

[0108] Raman Spectra: All Raman measurements were performed using a HORIBA LabRAM HR Evolution Confocal Raman Microscope. Samples were deposited on quartz substrates by drop casting concentrated toluene solutions in a glovebox and allowing the solvent to evaporate. Samples of GaAs, GaSb, and GaAs1-xSbx were measured using a 633 nm laser, a 100X air objective, and an 1800 grooves / mm grating (resolution ~1 cm-1) and were detected using a Horiba Synapse OE-CCD. Samples of GaP and GaP1-xAsx were measured using a 532 nm laser, a 100X air objective, and an 1800 grooves / mm grating (resolution ~1 cm-1) and were detected using a Horiba Synapse OE-CCD. Samples were measured using low laser power, the signal was collected using 3 s integration times, and the signal was averaged from 100 exposures. The spectrometer was calibrated using the Horiba LabSpec software using a Si (111) reference sample for all measurements.

[0109] Dynamic Light Scattering: DLS data was collected using a Mobius (Wyatt Technology). The laser operated at a wavelength of 532 nm, and the detector angle was set at 163.5 degrees. Measurements were performed at 25 °C using toluene as a solvent using a globular protein model for data analysis.

[0110] Nanocrystal Optical Characterization:

[0111] UV-Vis Absorption Spectroscopy: UV-vis spectra were collected from colloidal solutions of QDs with a Shimadzu UV-3600i Plus UV-Vis-NIR spectrophotometer in transmission mode.50 µl of a concentrated solution was diluted with 3 ml of anhydrous n- hexane and loaded in a 1 cm path length screw cap cuvette. Samples were collected with a reference cuvette consisting of 3 ml of n-hexane with 50 µl of added toluene. For samples which absorbed in the NIR range, tetrachloroethylene was used. All spectra were normalized at an optimal wavelength (typically 400nm) for easier comparison.

[0112] NIR Fluorescence Spectroscopy on InGaAsP / ZnSe: Samples were prepared by dissolving ~20 µl of a toluene solution of the nanocrystals in 3 ml of anhydrous tetrachloroethylene and placed in a 10 mm path length screw cap cuvette. PL spectra were collected on a Horiba Fluromax 4 plus spectrometer equipped with a DSS-IGA020L InGaAs detector cooled with LN2. Samples were excited at 450 nm using the integrated xenon lampAtty. Dkt. No.05400-0073-PCT and monochromator. Excitation light was passed through a 435-500 nm bandpass filter (Thorlabs FGB7S) and the emission light was passed through an 850 nm longpass (Thorlabs FLG850) filter before the emission monochromator.

[0113] Comparison of relative PL efficiency of InGaAsP / ZnSe samples was performed on an Edinburgh Instruments FLS 980 instrument, a 450 W Xe-lamp, and a single monochromator with 1200 grooves / mm blazed at 750 nm for excitation. Emitted light was dispersed using an 830 grooves / mm grating blazed at 1200 nm. PL was detected using a liquid nitrogen cooled R5509-72 with a spectral range from 300 to 1700 nm. The sample was excited with 810 nm light with a 7 nm slit, and the power was continuously referenced using an internal reference detector. Light was detected using a 4 nm slit from 850 nm to 1600 nm. Samples were prepared with an OD of ~0.7 at 810 nm, and the OD of each cuvette was independently measured using UV-vis absorption spectroscopy. PL intensity was normalized by the OD of the cuvette at 810 nm. All data was collected during the same instrument session.

[0114] Fluorescence Spectroscopy on Directly Synthesized GaAs: Samples were prepared by dissolving ~20µl of a toluene solution of the nanocrystals in 1 ml of 3- methylpentane in a glovebox.3-methylpentane was prepared by sparging the solvent with nitrogen on a Schlenk line for several hours. Next, the solvent was transferred to a glovebox and stored over activated molecular sieves for ~1 day. Finally, the solvent was transferred to a bottle and stored over NaK for future use. The nanocrystals dissolved in 3-methylpentane were loaded in a ~8 inch long, 4 mm ID, 6 mm OD quartz tube sealed at one end with an indent ~6 inches from the open end. A 3 mm OD quartz plug was inserted, and the sample was loaded onto a vacuum chuck and transferred out of the glovebox to a vacuum line. The sample was immersed in liquid nitrogen and then evacuated. The tube was flame sealed under vacuum using an oxyhydrogen torch. PL spectra were collected using a home build apparatus consisting of a 450 nm laser excitation source (Thorlabs CPS450) aligned to hit the center of the quartz tube. The emitted light was collected at a 90-degree collection angle, passed through a 500 nm long pass filter, and focused into a 400 µm diameter fiber using a fiber collimator (Thorlabs F950SMA-A) and detected using an Ocean Insight HR4Pro spectrometer. All spectra were collected as the raw intensity using the Ocean Insight software. The collection system was corrected for spectral sensitivity by collecting the spectrum of a calibrated blackbody lamp (Ocean Insight HL-3 Plus) and generating a correction file such that the measured spectrum matched the blackbody spectrum usingAtty. Dkt. No.05400-0073-PCT custom python scripts. For measurements at 77 K, the samples in quartz tubes were immersed in a vacuum insulated spectroscopy Dewar with quartz windows and filled with liquid nitrogen. The same system described above was used to measure spectra.

[0115] Visible Photoluminescence Excitation (PLE) Spectroscopy: Excitation spectra were collected on the same samples in vacuum sealed quartz tubes using a Horiba Fluromax 4 plus spectrometer. PLE spectra were collected from a variety of points along the emission spectra using 2 nm excitation and emission band pass. PLE spectra were collected as two separate scans avoiding the Rayleigh line. For the data presented in FIG.3G, the data around the Rayleigh line was fit to a Gaussian function and was used to interpolate though the missing data points as a dashed line.77 K measurements were performed using the Dewar system described above.

[0116] Time Resolved Photoluminescence: TRPL was measured on the same samples in vacuum sealed quartz tubes using a Picoquant FluoTime 300 instrument using a PMA 175 detector, and an LDH-P-C-405 diode laser with a 407 nm excitation wavelength (50 ps pulse width) and a laser repetition rate of 1 MHz.77 K measurements were performed using the Dewar system described above.

[0117] Transient Absorption Spectroscopy: Measurements were performed using a 35-fs, 2-kHz Ti:sapphire laser with time-delayed white light probe pulses produced in a 2 mm sapphire or CaF2plate and 1-kHz pump pulses centered at 400 nm produced from second harmonic generation of the fundamental.

[0118] Molten Salt Characterization:

[0119] Ambient and High Temperature Raman Measurements: All Raman measurements were performed using a HORIBA LabRAM HR Evolution Confocal Raman Microscope. Most room temperature measurements were performed using a 532 nm laser source, an ultra- low frequency filter enabling measurement of Raman shifts as low as 10 cm-1, an 1800 grooves / mm grating (resolution ~1 cm-1) and were detected using a Horiba Synapse OE- CCD. Temperatures were measured with a 633 nm laser and a standard filter enabling measurement of Raman shifts to 50 cm-1, an 1800 grooves / mm grating (resolution ~1 cm-1) and were detected using a Horiba Synapse OE-CCD. The spectrometer was calibrated using the Horiba LabSpec software using a Si (111) reference sample for all measurements. For room temperature measurements, a sealed quartz ampoule (9 mm O.D., 7 mm I.D.) with a solid salt plug at the bottom was placed horizontally on the microscope stage. The laser wasAtty. Dkt. No.05400-0073-PCT focused on the center of the curved ampoule using a 10X objective to collect spectra. For evaluating the molten salt products after reaction with nanocrystals, Gallium iodide capped InP, InAs, or InSb nanocrystals were processed identical to cation exchange reactions except before breaking the ampoule open for the recovery process. Raman spectra were collected on the solid salt matrix with the nanocrystals still embedded in the matrix. For high temperature measurements, a custom optical path turned the laser output of the Raman microscope 90 degrees and focused the light into a sealed quartz ampoule held vertically in an aluminum heating block, which was used to heat the salt samples.

[0120] Results and Discussion

[0121] Redox Reactions of Nanocrystals in Molten Salts

[0122] Molten salt solvents can be classified using Lewis acid-base concepts where group III halides such as GaX3(X = Cl, Br, I) are Lewis acids, alkali halides such as KI are Lewis bases, and the stoichiometric reaction GaI3 + KI ^ KGaI4 (m.p.230°C) produces a Lewis neutral molten salt. It was noticed that cation exchange reactions of III-V nanocrystals in molten salts are particularly sensitive to the nature of the group-V element. For example, InP can efficiently convert into In1-xGaxP in Lewis acidic, basic or neutral molten salts, while using Lewis neutral KGaI4molten salt is key for preserving chemical stability of InAs nanocrystals and forming In1-xGaxAs phase. Finally, InSb decomposes when a GaIIIsource is added to molten salts in an attempt to synthesize In1-xGaxSb nanocrystals.

[0123] InAs nanocrystals undergo the following reaction in KGaI4 molten salt: InAs(c) + xKGaI4(l) ^ In1-xGaxAs(c) + xKInI4(l), where c and l stand for “colloidal” and “liquid” respectively. The In1-xGaxAs product can be identified by powder X-ray diffraction (data not shown) and [InI4]- ions are detected by Raman spectroscopy (data not shown). To understand the stability problem of antimonides in molten salts, InSb nanocrystals were annealed in molten KGaI4 (data not shown) and no crystalline III-V products were found (data not shown). Instead, the observation of Sb0by XRD suggests that Sb-IIIin InSb is oxidized to Sb0, demonstrating redox activity of the pnictide component of III-V nanocrystals. Raman spectroscopy shows the formation of [Ga2I6]2-ions present in the molten salt after the reaction (data not shown). These GaII-containing ions indicate that KGaIIII4 is being reduced by the InSb nanocrystals in the following reaction: InIIISb-III(c) + 2KGaIIII4(l) ^ Sb0(s) + (K+, In+)2[GaII2I6]2-(l), where s stands for “solid”.Atty. Dkt. No.05400-0073-PCT

[0124] Gallium (III) and Indium (III) are the most common oxidation states, but molten halide salts can adopt a variety of compositions containing ions which are formally GaI, GaII, InIand InIIspecies. In a molten salt, the pnictide can act as a reducing agent for KMIIIX4 (M = In, Ga) molten salt, leading to decomposition of the III-V nanocrystals as schematically described in FIG.2A. However, an analysis of redox equilibria suggests that the oxidation of Sb-IIIto Sb0could be prevented if, instead of using GaIII, a molten salt with a weaker oxidizing potential was used. Indeed, annealing InSb nanocrystals in molten GaI[GaIIII4] (often referred to as Ga2I4) at 325°C prevented oxidative decomposition of the antimonide (data not shown). After annealing in Ga[GaI4], the XRD peaks of InSb nanocrystals shifted tolarger q-values (^^ ൌ ଶగௗ , where d is the interatomic spacing), indicating a decrease in lattice constant, consistent with incorporation of gallium into the InSb lattice. By tuning the reactiontime and In1-xGaxSb nanocrystals can be synthesized from x=0 to x=1 (data not shown).

[0125] Transmission Electron Microscopy (TEM) images of resulting In1-xGaxSb nanocrystals produced by cation exchange of InSb nanocrystals using Ga[GaI4] (data not shown) show well isolated nanocrystals with narrow size distribution and well-defined spherical shapes. Similarly, In1-xGaxAs and In1-xGaxP nanocrystals can be prepared from InAs and InP nanocrystals using molten Ga[GaI4] (data not shown). This reaction methodology can be further expanded to prepare quaternary and pentanary III-V colloidal nanocrystals by reacting InAs1-ySby, InP1-yAsy, InP1-ySby, or InP1-y-zAsySbz nanocrystals with Ga[GaI4] to prepare In1-xGaxAs1-ySby(x=0-0.83, y=0.67), In1-xGaxP1-yAsy(x=0-0.7, y=0.83), In1-xGaxP1-ySby(x=0-0.8, y=0.87), and In1-xGaxP1-y-zAsySbz (x=0-0.9, y=0.42, z=0.41). Importantly, the reducing environment provided by Ga[GaI4] prevented the leaching of the more reactive pnictide and the initial pnictide ratios were unchanged while the Ga:In ratio can be varied precisely (data not shown). Raman spectra of quaternary alloy nanocrystals show well defined peaks indicating high crystal quality and gradual shifts of the phonon modes as the composition was tuned, consistent with formation of alloy III-V nanocrystals (data not shown). To test the photophysical properties of quaternary In1-xGaxP0.17As0.83 nanocrystals (x = 0 to 0.7) synthesized in reduced Ga[GaI4] molten salt, a thin wide band gap ZnSe shell was grown to passivate surface states (data not shown). All samples showed strong near-IR PL (data not shown), validating molten salt transformation can prepare high-quality quaternary III-V semiconductors. The PL efficiency increased with higher temperature and longer annealing time in the molten salts (data not shown). The ability to independently control theAtty. Dkt. No.05400-0073-PCT composition of both the metal and pnictide enables independent control over the band gap, lattice parameter, size, and absolute band energies of a III-V material, which have proven instrumental for designing advanced optoelectronic devices grown by MBE and CVD methods.

[0126] The critical parameter that controls redox stability of III-V phases against decomposition in molten salt (FIG.1A) is the relative positions of the redox potentials for the pnictide oxidation and gallium or indium halide reduction. From the Raman studies, a qualitative order of the electrochemical potentials was inferred relevant to redox chemistry of InP, InAs, and InSb in molten KGaI4(FIG.1B). The key takeaway is that the antimonide in InSb is a sufficiently strong reducing agent to reduce GaIIIin KGaIIII4 to GaIIin the form of [GaII2I6]2-ions, which explains instability of group III-antimonide nanocrystals in molten salts. From this it is implied that InSb is not a sufficiently strong reducing agent to further reduce Ga[GaI4] to GaI2[GaII2I6] (often referred to “Ga2I3”). In contrast, As-IIIin InAs and P-IIIin InP are not a sufficiently strong reducing agents to react with KGaI4 and thus the formation of [GaII2I6]2-ions was not observed after reaction with KGaI4(FIG.1C and other data not shown). In support of the role of redox potential controlling the stability of III-V nanocrystals, it was found that KInI4causes oxidative decomposition of InP, InAs, and InSb consistent with the easier reduction of InIIIto InIIor InIcompared to GaIIIto GaII(data not shown). Altogether these results demonstrate that the redox potential of the molten salt is key for controlling chemical stability of III-V phases.

[0127] Direct synthesis of GaAs nanocrystals in molten salt solvent

[0128] Understanding the role of redox chemistry in molten salts can be applied not only to prevent decomposition of III-V nanocrystals, but to directly synthesize colloidal semiconductors in molten salts as outlined in the Scheme in FIG.1C. A reduced group III- halide (e.g., Ga[GaI4]) can serve as a reducing agent for an oxidized pnictide source (e.g., AsI3) to convert it into the reduced Pn-IIIstate (FIG.1C), enabling formation of III-V semiconductor phase. For example, it was found that Ga[GaI4] can simultaneously be a reducing agent and a gallium source to activate arsenic halides and synthesize GaAs according to the reaction 3Ga[GaI4](l) + AsI3(l) ^ GaAs(c) + 5GaI3(l). Given that GaAs is electrochemically stable in Ga[GaI4], it is implied that Ga[GaI4] has sufficient reducing potential (FIG.1B) to convert AsIIIto As-III. As set forth in Materials and Methods above, molten salt solvents such as KGaI4 (m.p.230°C) or CsI / NaI / KI eutectic (m.p.412°C) wereAtty. Dkt. No.05400-0073-PCT used, providing comparable results (data not shown). After allowing the reaction to proceed for one hour at 400 to 500 °C (for T < 425 °C, only KGaI4was used), followed by cooling to room temperature, the salt matrix was dissolved in an appropriate polar aprotic solvent, such as N,N-dimethylformamide. Synthesized nanocrystals were recovered as a powder, and oleylamine / ZnCl2 ligands were installed on the nanocrystal surface to enable colloidal dispersion in toluene or other non-polar solvents. Powder XRD patterns (FIG.2A) show phase pure zinc blende GaAs with decreasing diffraction peak width, and thus increasing crystallite size, of colloidal GaAs synthesized at increasing reaction temperatures. TEM images of the samples synthesized at 450 °C and 500 °C (FIGS.2B and 2C), and other reaction temperatures (data not shown) show well separated rounded particles. The inset images in FIGS.2B and 2C show a stable colloidal solution of the nanocrystals in toluene. Further comprehensive characterizations of GaAs colloids by various methods (including SAXS, DLS, UV-Vis data not shown) confirm sub-10 nm GaAs nanocrystals with narrow size distribution (~15% relative size distribution) formed from molecular reagents in molten salts. It can therefore be concluded that colloidal synthesis in molten salts can precisely balance nanocrystal nucleation and growth kinetics but without requiring the use of conventional organic solvents.

[0129] Raman spectra of GaAs samples synthesized at different temperatures from 400 °C to 500 °C (FIG.2D and other spectra not shown) show a remarkable trend: for reaction temperatures above 425 °C, strong Raman LO and TO modes were observed, as well as overtone peaks which are an indicator of high crystal quality. In contrast, below 425 °C (and irrespective of solvent used), poorly defined Raman features were observed, suggesting the lower reaction temperatures result in poor quality crystals, likely containing vacancy and antisite defects typical for GaAs grown at low temperatures by MBE or synthesized in organic solvents. In support of the improved crystal quality, samples synthesized at 425-500 °C showed room temperature PL (FIG.2E and other spectra not shown), while samples synthesized at 400 °C or lower temperature did not show any detectable PL at room or low (77 K, data not shown) temperature. As such, it appears high temperature synthesis is desirable to produce emissive GaAs nanocrystals. As-synthesized samples show PL emission bands centered at 617-759 nm (2.0-1.6 eV). To probe the degree of inhomogeneous broadening, photoluminescence excitation (PLE) spectra of as-synthesized colloids were measured for different detection wavelengths (FIG.2G). The PLE spectra show well-resolvedAtty. Dkt. No.05400-0073-PCT excitonic features, indicating that the GaAs nanocrystals have intrinsically narrow PL linewidth and size distribution is the main contributor to the width of ensemble PL spectra.

[0130] As described in the Materials and Methods section, above, the emission photon energy of GaAs nanocrystals is compared to size-dependent optical bandgaps calculated using atomistic semiempirical pseudopotential methods combined with the Bethe-Salpeter equation. An excellent agreement between theory and experiment (FIG.2F) further corroborates the synthesized GaAs nanocrystals are emitting from a band edge state. The progressive redshift of the PLE spectra onset for different detection wavelengths (FIG.2G), small Stokes shifts (FIGS.4A-4B), and nanosecond PL decay lifetimes (data not shown) are consistent with band edge emission. Finally, the samples show long-lived photoinduced transient-absorption bleach signatures which have strong correspondence with the linear absorption onset, further demonstrating that band edge carrier dynamics dominates the PL behavior of these materials (FIGS.5A-5B). Altogether the results demonstrate the first example of band edge emissive colloidal GaAs nanocrystals. Without wishing to be bound to any particular theory, it is believed a key breakthrough is colloidal synthesis at temperatures higher than 425 °C, by contrast to failed previous attempts to synthesize GaAs nanocrystals in organic solvents. The need for high temperature synthesis is related to the rigidity of covalent Ga-As bonds. At the same time, the bond rigidity gives rise to a weak electron- phonon coupling which defines the outstanding performance of GaAs in electronic and optoelectronic devices. To demonstrate this point, the calculations show that GaAs nanocrystals have significantly smaller exciton phonon coupling than CdSe and InP quantum dots (data not shown), which initial experiments corroborate.

[0131] A high-resolution TEM image of a GaAs nanocrystal synthesized in the CsI / NaI / KI eutectic solvent salt viewed down the 〈110〉 zone axis shows a nearly spherical morphology (images not shown) consistent with a Wulff construction with similar surface energy for facets with different Miller indices. Analogous to traditional colloidal synthesis, additions of surfactants can be instrumental for precise size and shape control of colloidal nanocrystals in molten salts. A surfactant, by definition, preferentially segregates to an interface and modifies the interface energy. To probe this effect in molten salts, small amounts of heterohalides (X=F-, Cl-, Br-) were dosed into the pure iodide-based synthesis of GaAs described above (data not shown). Nearly 5-fold enrichment of Cl- or Br- was found in the recovered nanocrystals, which indicates a preferential binding of Cl- and Br- to the surface of GaAs nanocrystal. Hard Lewis basic halides (e.g., Cl-) bond stronger to hard Lewis acidsAtty. Dkt. No.05400-0073-PCT like GaIIIcompared to softer I- (Ga-I bond dissociation energy: 339 kJ / mol, Ga-Cl bond dissociation energy: 481 kJ / mol). The preferential binding shows harder halides demonstrate one characteristic of a surfactant: preferential segregation to an interface.

[0132] GaAs nanocrystals synthesized with a Cl:I molar ratio of 0.26 commonly show triangular projections when viewed down the〈110〉zone axis (images not shown) which is consistent with a tetrahedron shape truncated by four {111}-type facets, which is further supported by imaging down the other zone axes (image not shown). This shape is consistent with the Wulff construction for a crystal where the {111} surface energy is much lower than other surfaces, indicating that Cl- ions have modified the interface energy of GaAs in the molten alkali halide salts. These shape changes are consistent across dozens of particles viewed by TEM and HRTEM (images not shown) and can be observed for other halides (F-) added (image not shown). The observed shape changes show chlorides demonstrate a second characteristic of surfactants: modification of the interface energy.

[0133] Direct synthesis of other III-V semiconductors in molten salts

[0134] The generality of the molten salt redox synthesis of III-V nanocrystals can be extended far beyond GaAs. For example, GaP nanocrystals can be accessed via the reaction 3Ga[GaI4](l) + PI3(l) ^ GaP(c) + 5GaI3(l). X-ray diffraction pattern and TEM images were obtained (data not shown) of GaP nanocrystals dispersed as colloidal solutions in toluene after oleylamine functionalization. As an indirect-gap semiconductor, GaP is expected to show long excited state lifetimes. Indeed, the transient absorption studies on the colloidal GaP nanocrystals revealed a 16 ms excited state lifetime (data not shown); such a long-lived excited state of GaP may be advantageous for using these nanocrystals as a light absorber for photocatalysis. Furthermore, GaP has a unique combination of a wide band gap (Eg=2.24 eV) and high refractive index (n =3.31 at 633 nm), which is appealing for photonic applications. Computational results (FIG.2F) show that quantum confinement can further open the bandgap of GaP to as wide as 3 eV for ~2 nm nanocrystals, which enables a solution- processable material with transparency and high refractive index across the entire visible spectrum. As a proof-of-concept, films of oleylamine-capped GaP nanocrystals were spin- coated on glass, and excellent transmittance was found from 450 nm to 1200 nm (data not shown). A measured refractive index of 2.6 was found, which is higher than that of TiO2 (n ~ 2.4) currently used in metalenses and other photonic elements for visible applications. Colloidal GaP nanocrystals can be lithographically patterned using DOLFIN photopatterningAtty. Dkt. No.05400-0073-PCT chemistry. The combination of high refractive index, low temperature solution deposition, ability to mix colloidal nanocrystals with polymers, and direct photolithographic patterning, demonstrates that quantum-confined GaP nanocrystals may enable inexpensive large area patterning of waveguides, metasurfaces, and other photonic elements for visible light.

[0135] Mixed-pnictide GaP1-yAsynanocrystals can be directly synthesized using mixtures of AsI3 and PI3. A powder XRD pattern was obtained (data not shown) which was consistent with GaP1-yAsyphase, and a TEM image was obtained (data not shown) which showed discrete, well dispersed colloidal nanocrystals. X-ray fluorescence (XRF) and Raman analysis further support the formation of alloy nanocrystals (data not shown). GaAs1-yPynanocrystals can be prepared from y=0.03-0.75 and thus the band gap can be tuned from ~1.7-2.4 eV (data not shown). Finally, the reductive chemistry can be expanded to III-Sb materials via the reaction 3Ga[GaI4](l) + SbI3(l) ^ GaSb(c) + 5GaI3(l) resulting in crystalline GaSb with large (50+ nm) crystallites (data not shown). Mixing AsI3and SbI3results in GaAs1-ySbynanocrystals (data not shown). The materials show good crystallinity by TEM. Colloidal stability and Raman and XRF analysis are consistent with the alloy formation (data not shown).

[0136] In summary, the results above demonstrate the unique versatility of molten salt redox chemistry in the synthesis of colloidal III-V quantum dots. This Example reports the synthesis of nearly a dozen compositions of binary, ternary, quaternary, and pentanary colloidal III-V quantum dots, which have thus far not been achieved. The combination of cation exchange and direct synthesis of Ga-pnictide materials in molten salts presents a rational path to prepare III-V quantum dots of nearly any composition and size desired. The remarkable ability to controllably nucleate and grow isolated nanocrystals in molten salts at temperatures far above those accessible for organic solvents enables synthesis of many unprecedented colloidal nanomaterials.

[0137] Additional Discussion on Direct Synthesis Reactions and Structural Characterization

[0138] Bulk and epitaxial thin film gallium arsenide and its alloys represent the most well-developed and technologically important optoelectronic semiconductor material. Early quantum dot researchers recognized its importance and attempted to modify established reaction methodologies for CdSe quantum dots to prepare GaAs quantum dots. For example, early work attempted to form GaAs nanocrystallites in the presence of coordinating solvents.Atty. Dkt. No.05400-0073-PCT As synthesized, these colloids had good colloidal stability but poor crystallinity, and it was found that flame annealing was necessary to increase crystallinity. Attempts to refine this chemistry were stalled by difficulties in obtaining highly crystalline samples and contamination of the reaction mixtures with organo-gallium species. No reliable examples of band edge photoluminescence were achieved. Other creative synthesis reactions in organic solvents do not produce crystalline materials and result in optical properties inconsistent with a quantum confined semiconductor.

[0139] Other attempts to prepare GaAs nanocrystals involved various ion exchange reactions. For example, a transmetalation reaction between GaCl3and Mg3As2in the presence of n-BuLi was used to prepare GaAs nanocrystals which displayed high crystallinity, but luminescence properties were not reported. Cation exchange routes to prepare GaAs from Cd3As2, using GaCl3 as a gallium source and trioctylphosphine as a solvent resulted in crystalline GaAs. Unfortunately, the samples still contained a high concentration of residual cadmium which likely prevented the sample from exhibiting photoluminescence. In summary, existing colloidal synthesis techniques have been unable to produce high quality GaAs with sufficiently low defect concentrations to sustain band edge PL.

[0140] GaAs nanocrystals have also been prepared from Me3Ga and AsH3, As(NMe2)3 or TMS3As under highly reducing conditions in squalane. However, like all the previous attempts to prepare GaAs, these materials did not show band edge PL. Standard structural characterization techniques typically used for colloidal nanocrystals such as powder XRD and high-resolution TEM suggest that these materials are high quality GaAs with good crystallinity. However, Raman spectra of the organic solvent synthesized GaAs show poorly defined LO and TO modes, whereas high quality GaAs wafers typically show sharp, well- defined LO and TO peaks. Low temperature grown GaAs thin films show similar ill-defined Raman peaks, and this has been correlated with formation of various point defects including gallium vacancies and antisite defects. Indeed, EPR and EXAFS studies of organic solvent synthesized GaAs nanocrystals show characteristics of point defects in the GaAs. High temperature annealing of colloidally synthesized GaAs nanocrystal films to 500 °C or annealing in molten salt solvents has been shown to cause well-defined LO and TO Raman modes to develop. This corresponds with the samples displaying photoinduced transient bleach features indicating improvement in crystal quality but no still no detectable PL.Atty. Dkt. No.05400-0073-PCT

[0141] This Example describes an approach to directly synthesizing nanocrystals in molten salts. This approach involved the following considerations: 1) Desirably, the molten salt solvent can sustain T > 500 °C.2) Desirably, reagents which would act as heterovalent dopants for GaAs are avoided as they may impart deleterious impurity defects.3) Desirably, reagents and byproducts are thermally stable at the reaction temperatures.4) Desirably, reagents do not contain organic C-H moieties which can react with gallium at high temperatures. With these considerations in mind, it was determined that alkali pnictides and III-halides are particularly useful classes of reagents, providing both components in the final oxidation state in the III-V lattice.

[0142] Structural Characterization of Colloidal GaAs Nanocrystals Synthesized in Molten Salts

[0143] This section explains why the GaAs synthesized in Example 1 is (1) high quality GaAs on an atomic scale and (2) a good quality colloidal dispersion consisting of isolated nanocrystals. GaAs was synthesized using both CsI / NaI / KI and KGaI4as the molten salt solvents. In both cases, nanoscale GaAs crystals are recovered. Powder X-ray diffraction shows phase-pure GaAs nanocrystals can be recovered from both solvents with minimal impurities present and broad diffraction peaks that are consistent with nanoscale crystallites. Samples synthesized in KGaI4solvent are systematically larger than those synthesized in CsI / NaI / KI. For example, for a 450°C reaction temperature KGaI4 derived GaAs has a Scherrer size of 4.7 nm and CsI / NaI / KI derived GaAs has a Scherrer size of 3.6 nm indicating solvent choice can modulate reaction outcome. Despite these small differences, GaAs nanocrystals derived from KGaI4and CsI / NaI / KI solvents are very similar.

[0144] Raman spectroscopy provides additional atomic-scale insight into the structural quality of the GaAs nanocrystals with sensitivity to various point defects. (FIG.2D.) For GaAs synthesized in KGaI4, a critical threshold is found where a sample synthesized at 400°C shows poorly defined LO and TO modes and no overtone peaks whereas samples synthesized at 425°C and above all show well-defined LO and TO modes and overtone peaks indicative of high quality. For samples synthesized above 425°C, narrower Raman peak widths are observed with increasing reaction temperature. Size-dependent Raman peak widths are associated with colloidal semiconductor nanocrystals. Importantly, the sample synthesized at 400°C has much broader peaks than expected for the crystallite size measured by X-ray diffraction indicating that another factor, like point defects, are contributing to theAtty. Dkt. No.05400-0073-PCT broad, weak Raman signal. For samples synthesized in CsI / NaI / KI, sharp well defined Raman features observed are consistent with high quality.

[0145] Regarding the colloidal structure of the nanocrystals, it was observed that GaAs nanocrystal colloids synthesized in KGaI4 or CsI / NaI / KI form solutions which do not visually scatter light, indicating the colloids consist of particles much smaller than the wavelength of light, and are stable in the long term. (FIGS.2A-2B.) Transmission electron microscopy of dried colloidal solutions shows small, isolated particles with sizes which are consistent with the size determined by Scherrer analysis. Further HRTEM images of individual nanocrystals show single crystal domains for all sizes, indicating that the size reflected by Scherrer analysis is representative of the particle size in the nanocrystal colloids.

[0146] Small angle X-ray scattering was used to analyze the size of GaAs nanocrystals synthesized in CsI / NaI / KI as colloidal solutions in toluene. (FIGS.3A-3D, first plots.) A small angle X-ray scattering pattern was observed, which is evidence of colloidal nanocrystals with a relatively narrow shape and / or size distribution. Importantly, a low q scaling which asymptotically approaches 1 at small q is observed. This is consistent with a structure factor of ~1, consistent with isolated particles in solution, i.e. a colloidal gas. This critical piece of data demonstrates that the entirety of the colloidal solution consists of isolated nanocrystals. Further, dips in the SAXS intensity between q~0.2 and q~0.1 were observed for samples synthesized at 400 to 500°C. This is interpreted as the size dependent form factor of the particles, consistent with a size increase with temperature. The small angle X-ray scattering patterns were quantitatively fit using the particle size distribution module in the Irena software package in Igor Pro (https: / / usaxs.xray.aps.anl.gov / software / irena), which utilizes the model-free maximum entropy approach to fit the data. Based on TEM images, the particles’ form factor was assumed to be that of a sphere with an aspect ratio of 1. The extracted size distributions were plotted (FIGS.3A-3D, second plots). The primary peak for each sample occurred at sub 10 nm sizes, with a small shoulder peak for larger sizes. The extracted size distributions were further fit with symmetric Gaussians to the main peak to determine the average particle size and deviation. It was found that the average particle sizes determined by SAXS fitting correspond well with the sizes determined by Scherrer analysis of wide-angle X-ray diffraction patterns. For example, in GaAs synthesized in CsI / NaI / KI at 500°C, Scherrer analysis indicates a crystallite size of 6.2 nm and SAXS obtains a diameter of 7.8 nm. This indicates that most of the GaAs nanocrystals are single-crystals.Atty. Dkt. No.05400-0073-PCT

[0147] Finally, the size of the colloidal nanocrystals was analyzed using dynamic light scattering. The visual observation of minimal scattering in colloidal samples and no observable scattering background in UV-Vis absorption spectra suggest these materials are colloids consisting of isolated unaggregated particles. Dynamic light scattering provides additional quantitative information about the hydrodynamic radius of the particles, which includes the ligand shell and is an important assessment of the quality of colloidal dispersions. It was found that the hydrodynamic diameter of the GaAs nanocrystal colloids increases with increased reaction temperature, following the size trend observed by XRD and SAXS. (FIGS.3A-3D, third plots.) The particle sizes determined by DLS are systematically larger than those determined by SAXS and Scherrer analysis; however, this is because the hydrodynamic size includes both the ligand shell and nanocrystal core. Importantly, no measurable quantities of large (10-100 nm) particles in the samples were observed, indicating large aggregates are not present.

[0148] In sum, the observation of broad powder X-ray diffraction peaks, isolated particles by TEM, single crystal domains by HRTEM, SAXS structure factors consistent with a colloidal gas of isolated particles, fits to the SAXS data showing few-nanometer particles sizes, and dynamic light scattering data showing few-nanometer hydrodynamic diameters all demonstrate that the colloidal GaAs nanocrystals synthesized in molten salts can be recovered as isolated, individual particles with excellent colloidal stability.

[0149] Evaluation of Molten Salt Synthesis Yield, Scalability, and Alternative Reaction Vessels

[0150] The yield of the GaAs nanocrystal synthesis reaction was determined by calculating the theoretical yield of the reaction considering the limiting reagent was entirely consumed to produce GaAs nanocrystals. For the yield calculations only the resulting colloidal particles after all purification steps were included. Both Lewis acidic and Lewis basic reaction conditions exhibited high yields of 50-60% colloidal GaAs.

[0151] This Example demonstrates that the GaAs nanocrystal synthesis reactions can be scaled up ~12x compared to typical exploratory scale reactions. The obtained GaAs nanocrystals displayed structural and optical properties nearly identical to small scale reactions. Remarkably, the size of the resulting GaAs nanocrystals was nearly identical to the small-scale reactions, indicating this chemistry is scale invariant and thus, enables large scale synthesis of monodisperse colloids of GaAs and other nanocrystals. Importantly, band edgeAtty. Dkt. No.05400-0073-PCT PL was observed from the GaAs nanocrystals which showed a peak position nearly identical to smaller scale reactions and appears to be slightly narrower, indicating a narrower size distribution. The large-scale reactions showed high yields (49%). Importantly, nearly 60 mg of GaAs nanocrystals were produced which could be dissolved at high concentrations necessary for device fabrication.

[0152] Finally, it was demonstrated that GaAs nanocrystals can be synthesized in unsealed reaction vessels, which allow for the addition of reagents during the reaction.

[0153] It was found that the nanocrystals synthesized in open quartz tubes using Schlenk line-like manipulations produced GaAs nanocrystals with nearly identical structural and optical properties as those synthesized in sealed quartz tube. Open tube reactions were used for both Lewis acidic and Lewis basic molten salts and in both cases show XRD and UV-Vis spectra which are nearly identical to sealed tube reactions. Importantly, for Lewis basic molten salt solvents, band edge PL from the open-tube synthesized GaAs nanocrystals was observed, indicating the material is a high-quality semiconductor.

[0154] Synthesis and Structural Characterization of GaP, GaP1-yAsy, and GaAs1-ySbyNanocrystals

[0155] Replacing AsI3with PI3, colloidal gallium phosphide nanocrystals were synthesized with narrow size distributions and good crystallinity as observed by powder X- ray diffraction and Raman spectroscopy. Importantly, the GaP synthesized shows well- defined Raman spectra indicating the material is highly crystalline. This GaP displays optoelectronic properties consistent with high quality, defect free GaP. For example, the GaP synthesized in this Example exhibited well-defined TO and LO Raman modes and a well- defined absorption onset which do not show sub-gap absorption features, both indicative of high crystal quality.

[0156] Mixing AsI3 and PI3 enabled formation of mixed pnictide solid solutions which showed XRD peaks intermediate of GaP and GaAs. As the P mole fraction varied from 0.1 to 0.9, it was found that the XRD peak positions gradually shifted and there was a nearly quantitative agreement between the added and measured P:As ratio. It was further found that the P:As ratio had minimal effect on the resulting particle size. Characteristics of solid solution formation for the mixed pnictide nanocrystals were observed via Raman spectroscopy. Importantly, it was observed that the GaAs-like LO and TO gradually shifted to smaller wavenumbers and become broader as more P is incorporated. In addition, theAtty. Dkt. No.05400-0073-PCT growth of a GaP-like LO and TO peak was observed as more P was incorporated. The trends in peak shifts were similar to that for solid solution semiconductors.

[0157] Relatively larger GaSb nanocrystals were prepared directly in molten salts. Highly crystalline nanocrystals were obtained in both Lewis acidic and Lewis basic conditions. Powder X-ray diffraction, TEM, and Raman spectroscopy indicated the formation of highly crystalline GaSb. These results further validate the hypothesis that redox potential governs synthesizability and stability of III-V nanocrystals in molten salt solvents. Upon mixing AsI3and SbI3, nanocrystals were formed which displayed diffraction peaks intermediate of GaAs and GaSb, indicating solid solution formation. Indeed, Raman spectra show characteristics of solid solution formation, including peak shifts and new peak formation similar to the description above. The samples showed relatively narrow size distributions and good colloidal stability. Together, the results demonstrated that a combination of PI3, AsI3, and SbI3may be used to prepare a variety of gallium pnictide solid solution nanocrystals with high quality.

[0158] Optoelectronic Characterization of GaAs, GaP, and GaP1-yAsyNanocrystals

[0159] Photophysics of GaAs Quantum Dots

[0160] In bulk, GaAs is a direct band gap semiconductor with a band gap of 1.42 eV at 300K and an exciton Bohr radius of 11.6 nm. The colloidal GaAs nanocrystals synthesized in this Example display strongly size dependent optical properties and behave as quantum dots. Specifically, it was observed that the nanocrystal colloids were orange-colored for the smallest crystallites (2.8 nm) and nearly black for the largest crystallites (8 nm). The samples demonstrated a size dependent absorption onset. Crude samples even displayed diffuse, but obvious excitonic features in the absorption spectrum. The inventors believe this to be the first example of size dependent excitonic features for colloidal GaAs nanocrystals and that the high crystal quality afforded by the present synthesis methods is the key to achieving this high crystal quality.

[0161] A remarkable correlation between the presence of well-defined Raman modes in the colloidal GaAs nanocrystals and the appearance of room temperature photoluminescence was found. Samples synthesized below 425°C show diffuse Raman spectra, indicating poorer crystal quality and no room temperature PL was observed. However, samples synthesized at 425°C and above show well-defined Raman modes and display strong band edge PL. ThisAtty. Dkt. No.05400-0073-PCT indicates there is a structural basis for the lack of PL in samples synthesized at low temperatures which is mitigated by the higher temperature synthesis.

[0162] Regarding the nature of PL emission from the GaAs nanocrystals, a small shift between the absorption onset and the PL emission was observed, consistent with a small Stokes shift. This suggests the PL is coming from a band edge state rather than a trap state. Further observation of sharp PLE spectra for different detection wavelengths across the PL spectrum (FIG.2G) indicates that the width of the PL spectrum is dictated by the size distribution of the sample and thus that different sub-populations are being sampled with different detection wavelengths. By contrast, for colloidal quantum dots for which the emissive state is a shallow trap, size information is lost in the PLE spectra, with similar PLE spectra observed for different PL detection wavelengths, and the shape of the PLE spectra approximately matches the absorption spectrum. Here, PLE was used to estimate the size dependent Stokes shift from the GaAs quantum dots by measuring the energy difference between the 1stexitonic peak in the PLE spectrum and the detection wavelength (FIGS.4A- 4B). It was found that the Stokes shifts vary from 60-20 meV as the size increases from 3 to 6 nm, respectively. Together, the steady state optical characterization indicates that the GaAs nanocrystals are behaving like high quality semiconductor quantum dots and emitting from a band edge quantum confined state.

[0163] The GaAs nanocrystals show complex time resolved photoluminescence decay traces suggesting non-radiative decay channels. At room temperature an early-time sub-ns decay channel was observed for all samples and the smallest samples had this component as their greatest proportion. The second component of the PL decay trace occurs on the 100 ns timescale which is consistent with band edge emission from a colloidal semiconductor as described above. The excited state dynamics of GaAs nanocrystals was further probed using femtosecond transient absorption spectroscopy. Samples were pumped with a 400 nm 35 fs pulse, and broadband transient absorption spectra were collected. For samples synthesized from 425-500°C, strong bleach signals were observed at early time, consistent with a high- quality direct band gap semiconductor. Interestingly, it was found that the majority of the photoinduced bleach signal decayed within the first 100 ps with larger GaAs quantum dots having a slower bleach decay. Despite this, a small portion of the bleach remained on the induced absorption background out to 1 ns (Fig. S36), and this small population was likely responsible for the 100 ns timescale observed for the PL emission.Atty. Dkt. No.05400-0073-PCT

[0164] In bare InP quantum dots, long-lived TA bleach features (t > 1 ns) are observed, and yet very weak PL, which has a sub 100 ps decay, is also observed. From this it is inferred that holes are primarily trapped based on the following logic: due to the lower degeneracy of the conduction band minimum compared to the valence band, the TA signal primarily tracks the electron population since any filled electron state will bleach the absorption. Such TA indicates there is a long-lived electron population within the InP. However, the lack of long- lived PL from InP indicates that the hole population has disappeared making radiative recombination impossible. In the case of the GaAs nanocrystals synthesized in this Example, the TA bleach signal decays rapidly, indicating the electrons are being trapped quickly (likely by surface states). Given the conduction band electrons in GaAs are more energetic (i.e. closer to vacuum level) than in InP or CdSe, they are more likely to be trapped by surface moieties.

[0165] The optical gaps of GaAs nanocrystals were calculated using the semiempirical pseudopotential method and the Bethe-Salpeter equation to account for electron-hole correlation. The atomistic pseudopotentials were parametrized to accurately reproduce the bulk GaAs band structure and deformation potentials. The calculated optical gaps of the nanocrystals decrease as GaAs size increases (FIG.2F), consistent with quantum confinement. There is excellent agreement between the calculated optical gaps and the measured PL peaks at all sizes, further indicating that the synthesized GaAs nanocrystals are high quality and emitting from a band edge state. The size-dependent optical gaps of GaAs nanocrystals were also compared to those of CdSe and InP nanocrystals (FIG.2F). For nanocrystal sizes larger than 5 nm, GaAs nanocrystals have smaller gaps than CdSe and InP, as expected from the bulk band gaps, exciton Bohr radii and effective masses of the three materials. For even larger GaAs nanocrystals, the optical gaps extend into the near-IR region and asymptotically approach the bulk gap of 1.42 eV.

[0166] The exciton-phonon couplings via the local deformation of the lattices were also calculated using the semiempirical pseudopotential method. Exciton-phonon couplings were aggregated as the reorganization energy, accounting for the energy associated with the rearrangement of the lattice after the vertical excitation from the ground state to the excited state. The GaAs nanocrystals have smaller reorganization energy and overall weaker exciton- phonon coupling than CdSe and InP nanocrystals of the same sizes, which can be explained by the rigidity and covalency of the Ga-As bonds and small local lattice distortions in these nanocrystals. In addition, the absorption and emission spectra were theoretically calculatedAtty. Dkt. No.05400-0073-PCT using the exciton manifold, transition dipole moments, and exciton-phonon couplings. The resulting Stokes shifts, calculated as the peak shift between the absorption and emission spectra, agree very well with the experimental measurements (FIGS.4A-4B), further proving band edge emission and exciton fine structures of these GaAs nanocrystals.

[0167] Altogether, the optoelectronic characterization and electronic structure calculations described above provide strong evidence that the GaAs nanocrystals are emitting from a band edge state. These results highlight that the present methods using the molten salt solvents have dramatically improved the crystal quality of GaAs nanocrystals such that they are now reliable optoelectronic materials.

[0168] Photophysics of GaP Nanocrystals

[0169] In bulk, GaP has an indirect band gap, and in nanocrystal form, GaP is likely to be essentially an indirect gap semiconductor. Millisecond transient absorption signals from GaP colloids were measured by pumping at 350 nm and probing the photoinduced absorption signal at 700 nm. A long-lived transient in the data that decays over ~20 ms was observed, indicating there is a long-lived excited state population. This signal was not correlated with a bleach signal, consistent with indirect gap semiconductors. The long-lived excited state makes GaP useful for a variety of photocatalytic applications where the rate limiting step may be charge transfer. GaP nanocrystals may also be used in various photonics applications.

[0170] GaP is a well-known material for photonic applications since it has a high refractive index which enables photonic elements with high refractive index contrast which subsequently enables design of photonic elements which better confine light or have larger bandwidth. Most gallium phosphide-based photonics require MOCVD-based growth which limits device area, is expensive, and is limited to substrates which can tolerate high growth temperatures. Solution processing of colloidal nanocrystals allows for materials grown under extreme conditions to be deposited under mild conditions and thus greatly expands the possible substrates which can be used.

[0171] Photophysics of GaP1-yAsyNanocrystals

[0172] In bulk, GaAs is asemiconductor with a band gap of 1.42 eV at 300 K and having a valence band maximum and conduction band minimum at the gamma point of the Brillouin zone. In bulk, GaP is an indirect gap semiconductor with a band gap of 2.24 eV at 300 K and having a valence band maximum at the gamma point and a conduction band minimum at the X-point. Addition of P to GaAs results in an increase in the direct band gapAtty. Dkt. No.05400-0073-PCT (primarily by increasing the energy of the conduction band minimum) up to addition of about 50 atomic percent P. At that point, the energy of the conduction band X-point and Γ-point cross making the X-point the lowest energy and the material shows an indirect band gap. The direct to indirect transition with increased phosphorus in GaP1-yAsynanocrystals is similar to the transition of In1-xGaxP from direct to indirect band structure with increased gallium concentration.

[0173] Dramatic color changes for GaP1-yAsynanocrystals were observed as the phosphorus content mirrors pure GaAs, including a NIR absorption onset and hints of an excitonic transition. Increasing the phosphorous concentration to 20-32%, the absorption onset blueshifts to ~650 nm, and any signature of an excitonic transition disappears. This is consistent with observations for In1-xGaxP, for which increased gallium close to the direct-to indirect transition led to less well defined excitonic transitions. Finally, samples with 64-70% P exhibited a featureless absorption spectrum with an onset around 500 nm, consistent with an indirect band gap material. Altogether, the results show that GaP1-yAsy nanocrystals exhibit similar composition defined optical properties as In1-xGaxP nanocrystals.

[0174] Additional information related to this Example, including data indicated as being not shown, may be found in U.S. Provisional Patent Application No.63 / 562,873, filed March 8, 2024, which is incorporated by reference in its entirety.

[0175] Example 2

[0176] This Example describes the direct synthesis of GaPn nanocrystals according to the dehalosilylation scheme shown in FIG.6. Either KI / GaI3 (35% / 65%) eutectic or iodide eutectic (CsI / NaI / KI=52% / 44% / 4%) were used the molten salt solvent. GaI3was used as the III precursor. Neat tris(trimethylsilyl)phosphine (TMS3P) and tris(trimethylsilyl)arsine (TMS3As) were used as the V precursors. The reaction mixture was annealed between 400 °C and 500 °C. After washing in either DMF or MeCN to remove the salts, GaP and GaAs nanocrystals were recovered which could be colloidally stabilized in organic solvent.

[0177] The nanocrystalline GaP and GaAs zinc blende phase was observed by powder X- Ray Diffractometer (data not shown). The as-synthesized nanocrystals were colloidally stable in nonpolar organic solvent (toluene) upon the addition of long-chain organic ligands (data not shown). The GaP nanocrystal solution showed featureless absorption by UV-Vis spectroscopy which is consistent with its indirect bandgap nature (data not shown). The GaAs nanocrystal solution showed a broad shoulder peak around 620 nm which may be attributedAtty. Dkt. No.05400-0073-PCT to its excitonic absorption (data not shown). Small angle X-ray scattering (SAXS) also confirmed that the nanocrystals are colloidally stable (data not shown). High-resolution TEM images confirmed the single-crystalline nature of the GaAs and GaP nanocrystals (data not shown).

[0178] For both GaP and GaAs nanocrystals, use of high temperature ensures high crystallinity. For GaP nanocrystals, a 300 °C reaction temperature did not result in a zinc blende crystalline phase, but rather a very random diffraction pattern not assignable to any known phases so far (data not shown). Increasing the temperature for GaAs synthesis did not change the peak position and peak shape of the PXRD pattern (data not shown). High- resolution TEM images from 400 °C to 500 °C also did not shown changes with temperature (data not shown). However, Raman spectroscopy may be used to characterize crystallinity and point defects and the Raman features were observed to be very sensitive to temperature increases. As shown in FIG.7A, the 400 °C sample exhibited very broad peaks without any overtone features while 450 °C sample showed very distinct LO and TO resonance modes and 2TO, 2LO and TO + LO modes, demonstrating that the high temperature plays a role in curing the point defects of the nanocrystals. As shown in FIG.7B, for the 500 °C sample, after treating the nanocrystals with HF (generated by in-situ reacting benzoyl fluoride and octylamine), band edge PL was observed, an optical property which is extremely rare for GaAs nanocrystals. Transient absorption data also showed this sample has a bleaching at its band-edge position and its lifetime is about 100 ps (data not shown).

[0179] Temperature also plays a role in controlling the size of nanocrystals in the direct synthesis reaction. For the GaAs nanocrystals, PXRD peaks become sharper as the temperature was increased from 400 °C to 500 °C, meaning the crystal domain increases with increasing reaction temperature (data not shown). The sizes added were calculated based on the Scherrer Equation (^^ൌ ^^^^^^cos^^). The absorption feature also showed a red shift upon temperature increase (data not shown). The wavelength of the excitonic feature was derived from the second derivative of the UV-Vis (data not shown) and a sizing curve was plotted to confirm that the energy gap changes with the particle size (data not shown). Specifically, the bandgap of the material may be tuned from 2.25 eV to 1.69 eV. For the gallium phosphide nanocrystals, the size of the nanocrystals also increased as the temperature increased, although the maximum was about 3.8 nm (data not shown).Atty. Dkt. No.05400-0073-PCT

[0180] Besides binary phase nanocrystals, the dehalosilylation reaction scheme may also be used to synthesize ternary phase nanocrystals. For example, arsenic and phosphorus precursors may be mixed with a molar ratio of 2 to 1 and added to the molten salt solvent to provide the ternary phase nanocrystals. PXRD patterns (data not shown) of as-synthesized products showed a shift from GaAs, indicating a change of lattice constant due to the incorporation of P into the lattice. Temperature increases did not greatly affect the size of as- synthesized nanocrystals or the arsenic ratio (data not shown). The Raman spectrum for the ternary phase nanocrystals agreed well with bulk ternary GaAsP (data not shown). A red shift from the GaAs binary phase and a new peak at the TO mode for GaP were observed in the ternary phase spectrum, which confirmed that a true alloy phase, rather than separated phase products, was synthesized.

[0181] Bulk GaP has a wide bandgap of about 2.24 eV, giving it a transparent absorption window throughout most of visible range. However, its refractive index in this range is high, above 3.2. To investigate the possibility of applying as-synthesized GaP nanocrystals into photonics, a nanocrystal film (NC film) was fabricated by spin-coating followed by refractive index measurement with ellipsometry. The refractive index of the as-prepared film was 2.8, which is higher than that of NC films prepared by existing techniques (less than 2.0). The UV-Vis spectrum of the NC film exhibited a transparent window ranging from 450 nm to 800 nm (data not shown). DOLFIN (direct optical lithography for inorganic nanocrystals) was used to form a high-quality pattern from the NC film, further demonstrating that GaP nanocrystals synthesized from molten salt can be fabricated into functional structures.

[0182] Finally, the reaction scheme described above and shown in FIG.6 was used to synthesize other colloidal nanocrystals, including InP, and In1-xGaxP, 0.25 ≤ x ≤ 0.48. The molten inorganic salts used were KBr / InBr3and KBr / GaBr3. Some of the nanocrystals further included ZnS shells on the In(Ga)P cores. Characterization of the resulting colloidal nanocrystals was carried out as described above.

[0183] This Example further demonstrates the use of molten salt solvents as unique media in which high quality III-V nanocrystals may be directly synthesized from molecular precursors. The as-synthesized GaP nanocrystals may be used in a variety of waveguiding and other photonic applications due to their high refractive index and transparency in visible region. GaAs nanocrystals may be widely used in lasing and transistor applications due to their direct band gap and fast carrier mobility.Atty. Dkt. No.05400-0073-PCT

[0184] Additional information related to this Example, including data indicated as being not shown, may be found in U.S. Provisional Patent Application No.63 / 562,873, filed March 8, 2024, which is incorporated by reference in its entirety.

[0185] Example 3

[0186] This Example describes the high-temperature solution synthesis of colloidal metal nitride nanocrystals by reacting metal halides dissolved in molten inorganic salts with NH3 at an elevated pressure. NH3pressure was found to enable control of the reaction product morphology by stabilizing colloidal nitride nanocrystals against aggregation and sintering. Successful syntheses of colloidal nanocrystals of binary nitride phases, including TiN, VN, GaN, NbN, Mo2N, Ta3N5, W2N, as well as ternary TixV1-xN nanocrystals, were demonstrated. These materials expand the scope of important semiconductors (e.g., GaN), superconductors (e.g., VN, NbN), plasmonic materials (e.g., TiN), and catalysts available in form of solution- processable colloidal nanocrystals.

[0187] Materials and Methods

[0188] Chemicals and materials. Cesium iodide (ultra dry, 99.998%), gallium(III) iodide (ultra dry, 99.999%), niobium(V) chloride (ultra dry, 99.9%), sodium iodide (ultra dry, 99.98%), tantalum(V) chloride (anhydrous, 99.995%), titanium(IV) iodide (ultra dry, 99.99%), potassium iodide (ultra dry, 99.998%) were purchased from Alfa Aesar. Gallium metal (99.99%), molybdenum(V) chloride (anhydrous, 99.6%), tungsten(VI) chloride (99.9%), vanadium(III) chloride (anhydrous, 99%) were purchased from Strem Chemicals. Ethanol (200 proof, anhydrous, ≥ 99.5%), methanol (anhydrous, 99.8%), methyl acetate (anhydrous, ≥ 99.5%), methylcyclohexane (anhydrous, ≥ 99%), N,N-dimethylformamide (DMF, anhydrous 99.9%), oleic acid (technical grade, 90%), oleylamine (technical grade, 70%), and toluene (anhydrous, 99.8%) were purchased from Sigma-Aldrich. Oleic acid was degassed under dynamic vacuum at 110 °C for 3 hours before storage in a nitrogen glovebox. Oleylamine was vacuum distilled over sodium (Sigma-Aldrich, 99.8%) and stored in a nitrogen glovebox. Fused quartz tubes (9 mm O.D., 7 mm I.D.) and rods (6 mm O.D.) for ampoule fabrication were purchased from Technical Glass Products, Inc. Ammonia gas (Airgas, anhydrous grade) was condensed inside a high-pressure vessel for high-pressure syntheses. For ambient-pressure syntheses, ammonia was condensed in a glass pressure reaction vessel (Andrews Glass) using a dry ice / acetone bath right before use.Atty. Dkt. No.05400-0073-PCT

[0189] High-pressure setup. High-pressure vessels from Parr Instrument Company were used for the setup. A Series 4740 (75 mL) vessel was used as the reaction vessel and a Series 4650 (1000 mL) vessel was used as a secondary ammonia reservoir. For safe operation of air- free high-pressure reactions, a custom stainless-steel Schlenk line was built using tubes, valves and fittings purchased from Swagelok. A tee with three needle valves was used to connect two vessels to the Schlenk line. Liquid ammonia was condensed into the ammonia tank using an ice / water bath. The reaction vessel was a detachable unit equipped with a silicon nitride liner and a needle valve, and it could be transferred inside and outside of the nitrogen glovebox for air-free handling of the precursor and products. The system was assembled and properly evacuated through Schlenk line when performing reactions. Two ultra-high temperature heating tapes (Fiberglass Heater Tape 1400°F Max, Omega Engineering) were used to heat the reaction vessel and ammonia tank separately. Caution! The pressure of ammonia will be high enough to break glassware, so needle valves must be manipulated with care to ensure high-pressure ammonia is not admitted to the Schlenk line. Safety devices (safety rupture discs, etc.) are highly recommended to be installed in the system where working pressure is close to the pressure limit of the part.

[0190] Preparation of molten salt precursors. To avoid inclusion of oxygen impurities or moisture, preparation of salt precursors was conducted in a nitrogen glovebox with O2 and H2O level less than 0.1 ppm. A powder of the CsI / KI / NaI eutectic mixture was prepared by grinding together CsI, KI, and NaI (molar ratio of 52:4:44) with a mortar and pestle. Ga2I4 was prepared by loading Ga and GaI3 in a 1:2 molar ratio into an oven-dried quartz ampoule with an indentation to support a sealing plug above the sample, and a 6 mm diameter quartz plug was inserted. The ampoule was attached to a vacuum transfer chuck and then mounted on a vacuum manifold without air exposure. The ampoule was sealed to the plug under vacuum using a O2 / H2 torch. After sealing, the ampoule was heated at 450 °C for 3 days. During this process, the ampoule was shaken at regular intervals. After cooling down, the ampoule was taken inside a nitrogen glovebox. The ampoule was cut open to remove the solid lump of Ga2I4, which was ground to a powder.

[0191] Synthesis of metal nitride nanocrystals under high-pressure ammonia. In a nitrogen glovebox, 0.3 mmol of finely ground metal halide powder (TiI4, VCl3, Ga2I4, NbCl5, MoCl5, TaCl5, WCl6), 1000 mg of CsI / KI / NaI eutectic powder, and a glass stir bar were carefully loaded into the silicon nitride liner in the reaction vessel. For ternary TiVN synthesis, different ratios of TiI4and VCl3were used, according to a total amount of 0.3Atty. Dkt. No.05400-0073-PCT mmol of metal. The reaction vessel was then closed and attached to the high-pressure setup. The reaction vessel and the ammonia tank were heated separately. The reaction vessel was heated to the desired temperature (425-575 °C) under ambient pressure nitrogen. The ammonia tank with liquid ammonia inside was gently heated to around 1.5 times of the target pressure. After reaching the desired temperature in the reaction vessel, the magnetic stirrer was turned on and the reaction mixture was kept at the same temperature for another 5 min to equilibrate. Hot ammonia gas with desired pressure (0.8, 2, 2.8, 3.5, or 6 MPa) was then transferred to the reaction vessel by opening the two needle valves between the reaction vessel and the ammonia tank. Caution! The needle valve towards the Schlenk line must be closed during this process. The reaction was allowed to proceed for 5 min and then stopped by closing the two needle valves between the ammonia reservoir and reaction vessel. To avoid condensation of liquid ammonia during cooling, the high pressure gas in the reaction vessel was carefully vented until the pressure reached around 1 MPa. This residual pressure also prevents the back flow of air into the vessel. After cooling down the pressure vessels to room temperature by removing the heating tape, the reaction vessel was transferred to a nitrogen glovebox for post-synthetic treatment.

[0192] Recovery of colloidal nanocrystals from salt matrix. The high-pressure reaction vessel was carefully opened in a nitrogen glovebox. Even after lowering the pressure after the reaction, the pressure inside the vessel can still be well above atmospheric. Before opening the vessel, the needle valve needs to be opened first to ensure no pressure difference between the vessel and the box atmosphere. To prevent damage to the glovebox drying train catalyst, the glove box should be placed under a rapid purge of nitrogen. The liner containing the salt matrix was taken out from the reaction vessel, and 20 mL of methanol was added and stirred until the salt matrix dissolved. The nanocrystals were then separated by centrifugation and the supernatant was discarded. The nanocrystals were washed again with 20 mL of methanol and collected by centrifugation. Colloidal solutions of these products were obtained by mixing the solid precipitates with 5 mL of toluene (or methylcyclohexane), 50 μL of oleylamine, and 50 μL of oleic acid. To remove excess ligands and impurities, nanoparticles were collected by centrifugation with 5 mL of ethanol and finally redispersed in toluene or methylcyclohexane.

[0193] Purification and size-selective precipitation of colloidal nanocrystals. For GaN synthesized at ammonia pressures above 3.5 MPa, large crystals (>30 nm) were observed. The large crystals can be purified by precipitation with the addition of 0.4 volume ratio ofAtty. Dkt. No.05400-0073-PCT methyl acetate into GaN / methylcyclohexane solution prepared as above. The GaN NCs were then precipitated by centrifugation using ethanol and redispersed in methylcyclohexane for optical measurements. Size-selective precipitation was carried out on colloidal TiN, VN, TixV1-xN, NbN, Mo2N, and Ta3N5nanocrystals. Toluene and methyl acetate were used as solvent and antisolvent, respectively. Different batches of colloidal nanocrystals were collected by centrifugation with solvent-antisolvent volume ratios of 0.4, 1.0, 1.4, 2.0 and 3.0. The precipitated nanocrystals were redispersed in toluene for subsequent characterization.

[0194] Synthesis of nanocrystalline metal nitrides under ambient-pressure ammonia. In a nitrogen glovebox, 0.3 mmol of finely ground metal halide powder (TiI4, VCl3, Ga2I4, NbCl5, MoCl5, TaCl5), 1000 mg of CsI / KI / NaI eutectic powder, and a glass stir bar were carefully loaded to the bottom of an oven-dried quartz ampoule. The ampoule was attached to an adapter with closed gas inlet and outlet valves. The ampoule was taken outside the glovebox and then connected to the Schlenk line and the ammonia condenser. All the lines were properly evacuated to ensure inert atmosphere. The reactions were performed using a custom-made aluminum heating block with 3 / 8-inch diameter, 2.5-inch-deep holes drilled in a circle pattern. The aluminum block was preheated to the desired temperature (450-550 °C), and the ampoule was inserted in the heating block. The reaction mixture was kept for 5 min under stirring to equilibrate, and ambient-pressure (0.1 MPa) ammonia was hot-injected to the reaction. Continuous flow of ammonia was maintained at the same temperature for 5 min, then the reaction was stopped. After cooling down to room temperature, the ampoule was transferred inside a nitrogen glovebox. The salt matrix was taken out after cutting the ampoule open, and nanocrystalline metal nitrides were washed and then dispersed in toluene using the same method described above.

[0195] Characterization of the synthesized NCs was carried out using the techniques described in Example 1, above.

[0196] Results and Discussion

[0197] Alkali metal halides were used as solvents to simultaneously provide high- temperature stability and high solubility of ionic compounds. In a typical synthesis, metal halide precursors were mixed with finely ground CsI / NaI / KI eutectic powder and added to a high-pressure reactor. The mixture was heated to a desired temperature (450-600 °C) under dry nitrogen atmosphere, forming a homogeneous solution with no phase separation. High- pressure NH3was then injected into the reactor and the temperature was maintained for 5Atty. Dkt. No.05400-0073-PCT min. The reaction was then cooled down to room temperature, and the products were retrieved from the salt matrix with methanol which dissolved the alkali halide and ammonium halide salt matrix. With the addition of oleic acid and oleylamine ligands, stable colloidal solutions were obtained by subsequent dispersion of nitride NCs in non-polar organic solvents such as toluene.

[0198] The ammonia pressure, together with the reaction temperature, were found to be key parameters in controlling reaction products. For example, the reaction of VCl3with NH3at 500 °C yielded three distinctive outcomes depending upon the NH3 partial pressure. At 100 kPa, nanocrystalline cubic VN consisted of aggregated and sintered grains. When the NH3partial pressure increased to 2.0 MPa, discrete and monodisperse VN NCs were obtained and formed stable colloidal solutions in non-polar solvents when capped with oleate and oleylamine ligands. Powder X-ray diffraction, transmission electron microscopy, and small- angle X-ray scattering confirmed the formation of discrete and uniform cubic-phase VN NCs with well-defined size and shape that formed stable colloidal solutions in toluene. The reaction yield was ~75%. A further increase of NH3pressure to 5.0 MPa, increased the Scherrer size of the VN crystalline domains, preventing the formation of stable colloidal dispersions. The discussion below focuses on the nitride NCs produced at the intermediate pressures, which showed best synthetic control and monodispersity.

[0199] The effect of ammonia pressure on the morphology of synthesized metal nitrides is surprisingly general – multiple binary (GaN, TiN, VN, NbN, Mo2N, W2N, Ta3N5) and ternary (e.g., Ti1-xVxN) metal nitrides successfully nucleate and grow within the pressure and temperature windows examined, including from 450℃ to 600℃ and from 0.8 to 5.0 MPa NH3. Specific reaction conditions as well as PXRD patterns and TEM images of the various nitride NCs synthesized in molten salts are shown in FIGS.8A-8F. Monodisperse VN NCs with well-defined cuboid shape were synthesized using 2.0 MPa of NH3 at 500 °C (Fig.2C and 2D). Colloidal GaN NCs with mainly zinc blende phase were produced using 2.8 MPa of NH3at 450 °C (FIG.8C). However, zinc blende and wurtzite polytypes could be formed by small variations of the reaction conditions. Higher temperatures were found to be necessary to grow ~4 nm NbN NCs (FIG.8D) and Mo2N NCs (FIG.8E). Slight shifts relative to the XRD peak positions in bulk phases were observed in XRD patterns of cubic phase TMN NCs. The reaction between Ta(V) halides and NH3yielded a orthorhombic Ta3N5phase instead of cubic TaN (FIG.8F). Most of the Ta3N5 NCs exhibited anisotropic growth, appearing as rod structures. W2N NCs can also be produced by the same method using WCl6Atty. Dkt. No.05400-0073-PCT as the precursor, but the retrieved products were less colloidally stable and contained a small amount of impurities, possibly due to the lower phase stability of W2N. All materials formed stable colloidal solutions.

[0200] The synthesis pressures used in this Example did not change chemical potentials of the solid phases being studied. However, the pressure can impact kinetics of NC nucleation and growth, either through the activation volume (a pressure analog of activation energy in the Arrhenius equation) or through pressure-related amounts of the nitrogen species (NH3, NH2-) delivered to the reaction mixture. It is believed that no experimental information is available on how pressure impacts the kinetics of nucleation and growth of colloidal NCs. Thus, a series of control experiments were carried out using nitrogen pressures ranging from ambient to 200 MPa. Surprisingly, the effect of gas pressure was negligible on the size and size distribution of the model NC system studied (PbS quantum dots). This affirms that the activation volumes of NC nucleation and growth are too small to impact the synthesis by pressure alone. However, it is noted that in the nitride synthesis, NH3 is a reagent, and partial NH3pressure is directly proportional to its concentration dissolved in the molten salt. The very fact that NH3 pressure has strong effect on NC size suggests that NH3 or associated nitrogen species (e.g., NH2–or NH4+formed via autoionization of NH3) are participating in the in the rate-limiting step, which can be chemical reaction or diffusion toward NC surface.

[0201] The nitrides synthesized at 100 kPa partial NH3 pressure typically formed aggregates and irregularly shaped crystallites likely formed through the oriented attachment and sintering of smaller NCs, while an increase of NH3 pressure to 2.0 MPa under otherwise identical conditions yields single-domain cubic NCs terminated with (100) facets. This observation suggests a correlation between ammonia pressure, aggregation and oriented attachment of nitride NCs. To aggregate, the NCs should first come in contact, which requires overcoming repulsive barriers responsible for colloidal stability. Since colloidal stability in molten salts is related to surface-templated layering of the ions around each NC, the presence of surface-binding ions is critical for stabilization of colloidal dispersions. In iodide molten salts, Lewis-acidic metal ions present at the NC surface are interacting with Lewis basic moieties including I–, NH3, and NH2–. To better understand surface interactions, DFT studies of binding energies (Eb) of the above species toward (100) surfaces of a VN NC immersed in KI molten salt were conducted. The results showed that VN NCs exhibit only a weak affinity toward I–ions when immersed in molten KI. NH3 binds to VN (100) surface much more strongly, but its binding energy strongly decreases with increasing surfaceAtty. Dkt. No.05400-0073-PCT coverage, and only partial coverage can be achieved, even when excess NH3 is present in the molten salt. Moreover, neutral NH3molecules bonded to non-polar (100) NC facets are unlikely to template the ion layering needed for colloidal stability in a molten salt. In contrast, NH2–ions strongly bind to (100) facets of VN NCs and Ebis practically independent of surface coverage. The charges of surface-bound NH2–ions can provide colloidal stability and prevent aggregation of nitride NCs, if sufficient concentration of these groups present at NC surface.

[0202] The solubility of NH3 in alkali nitride eutectic molten salt used in this Example is ~ 0.16 ^^^^^^ ^^−1^^^^^^−1, and a saturated solution at 100 kPa partial pressure contains only ~ 0.023 ^^^^^^ ^^−1of NH3, which is much smaller than ~1 ^^^^^^ ^^−1initial concentration of reactive metal ions dissolved in the molten salt. Dissolved NH3partially dissociates forming NH2–and NH4+ions, with the equilibrium constant strongly increasing with temperature. The equilibrium concentration of NH2–ions is less than 10-5^^^^^^ ^^−1at 500 °C and 100 kPa NH3pressure. As a result, at an ambient ammonia pressure, nitride NCs nucleate and grow under a strong deficiency of nitrogen precursors. This is far from ideal because such conditions cannot provide a sufficient coverage of surface-bound NH2–ions to enable colloidal stability and prevent aggregation and sintering of nitride NCs. When the ammonia pressure is raised to 2 MPa, the equilibrium concentration of dissolved NH3 increases to 0.46 ^^^^^^ ^^−1, which is comparable to the concentration of metal species, and falls into the range of conditions that favor colloidal stability. When ammonia pressure is too high, the crystal growth is too fast, leading to larger crystals and reduced colloidal stability.

[0203] The synthesis of VN NCs can be described as a metathesis reaction: VCl3+4NH3 → VN + 3NH4Cl. The syntheses other nitride NCs also involves redox processes where metal ions change their oxidation state. NH3 can actively participate in redox reactions, either reducing to N(-III) + H2or oxidizing to N2+H+. In GaN NC synthesis, it was found that the metathesis reaction using GaI3 did not produce crystalline GaN. However, the reactions of NH3 with reduced gallium halides, Ga2I4 or Ga2I3, produced highly crystalline GaN (FIG. 8C). Gas chromatography tests of the reactor head space revealed the presence of H2after the Ga2I4 reaction, indicating that reduced Ga halides are simultaneously Ga sources and reductants that activate NH3during the formation of GaN NCs. Redox reactions were also suggested by the products of the reactions between Ti(IV), Nb(V), Mo(V) halides and NH3 in molten salts, whereas Ta(V) showed less tendency to be reduced, because of stability of Ta(V) oxidation state.Atty. Dkt. No.05400-0073-PCT

[0204] Further regarding the temperature used in the synthesis of the metal nitride nanocrystals reported in this Example, colloidal GaN NCs synthesized below 500℃ showed broad emission from mid-gap states. However, starting from about 550℃, the GaN NCs showed predominantly band edge PL. Although this transition is not correlated with the TEM and X-ray diffraction data, it is correlated with the Raman spectra, suggesting that GaN NCs synthesized below 500℃ incorporate defects, likely the nitrogen vacancies. The observed effect of synthesis temperature on band edge PL is similar to the GaAs QDs synthesized as described in Examples 1 and 2, above. Additional spectroscopic studies (transient absorption, PL lifetime) further supported the conclusion that the GaN emission is from a band edge state. This observation enables blue QD emitters for QLED technologies.

[0205] Regarding the transition metal nitrides synthesized in this Example, the absorption spectra of a colloidal solution of TiN NCs exhibited plasmonic resonances. (See FIG.9C.) This supports the use of the transition metal nitrides as plasmonic materials for various theragnostic applications. Transition metal nitrides are also widely used as coatings in the cutting- and machining-tool industry due to their hardness, thermal stability and resistance to corrosion. These properties, derived from strongly covalent metal-nitrogen bonds, were also evident in the synthesized NCs. Specifically, both GaN and transition metal nitride NCs reported in this Example showed excellent stability against oxidation, even after a prolonged (24 hours) heating at 150℃ in air.

[0206] Additional Structure characterization and photophysics of GaN quantum dots

[0207] This section shows that the colloidal GaN NCs synthesized in molten salts as described herein (1) have high crystal quality on an atomic scale and (2) are emitting from a band edge state. The reaction scheme corresponding to the synthesis of the GaN NCs is Ga2I4 + 6NH3= 2GaN + 4NH4I +H2.

[0208] First, the structure of the GaN NCs is discussed. To investigate the pressure effect, GaN NCs were first synthesized at 450°C using different ammonia pressures. As an ambient- pressure (0.1 MPa) comparison, the reaction mixture containing the same amounts of salts was heated to the same temperatures, followed by a passing a continuous flow of ambient- pressure NH3for 5 minutes through the mixture. Qualitatively different reaction products were observed when applying different pressures. Generally, with an increase of NH3 pressure, GaN transformed from polycrystalline sintered NCs, to discrete NC colloids, andAtty. Dkt. No.05400-0073-PCT then to larger crystals. PXRD patterns of GaN samples revealed a continuous size increase with increased pressure. The Scherrer size of GaN samples synthesized using 0.1 MPa, 0.8 MPa, 2.0 MPa, 2.8 MPa, 3.5 MPa, 6.0 MPa ammonia pressure were 2.6 nm, 4.2 nm, 5.1 nm, 7.3 nm, 12.1 nm, and 31.6 nm, respectively. TEM images showed the large sintered polycrystalline domains synthesized with 0.1 MPa NH3. Discrete GaN NCs were recovered when using 2.0, 2.8 and 3.5 MPa NH3, and a phase transition from wurtzite to zinc blende was also observed by high-resolution TEM among the major products, in accordance with their XRD characterization. Single-crystalline zinc blende GaN NCs with lateral size >40 nm were synthesized when using 6.0 MPa NH3. The GaN NCs synthesized using high-pressure ammonia were colloidally stable in non-polar organic solvents such as methylcyclohexane, and a red shift of absorption spectra was observed for GaN synthesized with higher pressure. GaN NCs synthesized using 2.0-3.5 MPa ammonia pressure had long-time colloidal stability against aggregation.

[0209] To investigate the temperature effect, GaN NCs were synthesized at different temperatures between 425°C and 575°C. A 0.8 MPa ammonia pressure (room temperature ammonia vapor pressure) was used to minimize the pressure difference between reactions. The XRD patterns showed that the size of GaN NCs increased as the temperature increased, evidenced by the sharpening of peaks. The Scherrer size of GaN samples synthesized at 425°C, 450°C, 475°C, 500°C, 525°C, 550°C, 575°C were 3.1 nm, 4.2 nm, 5.1 nm, 6.2 nm, 6.8 nm, 7.5 nm, and 12.6 nm, respectively. Also, the formation of wurtzite GaN phase was more preferred as the temperature increased, which is expected for GaN since wurtzite phase is slightly more thermodynamically stable than zinc blende phase. More distinctive ^101ത0^, ^0002^, and ^101ത1^ peaks were observed as temperature increased, indicating larger wurtzite crystalline domains. TEM images showed that GaN NCs were more prone to aggregate or sinter as the temperature increased. The nanocrystalline GaN sample synthesized at 575°C was difficult to be colloidally recovered.

[0210] Raman spectroscopy was used to investigate the structural quality of the GaN nanocrystals. A commercial CVD-grown wurtzite GaN sample was used as a reference. Samples synthesized at 475°C and below showed poorly defined A1(LO) and E2(high) modes and no overtone peaks. A weak A1(LO) peak was observed for samples synthesized at 500°C. Samples synthesized at 525°C and above have well-defined A1(LO) and E2(high) modes and overtone peaks. It was found that the A1(LO) peak for GaN NCs was shifted to lower wavenumbers compared with bulk GaN, possibly due to the lattice strain or the mixture ofAtty. Dkt. No.05400-0073-PCT zinc blende phase. These results demonstrate that a synthesis temperature >500°C is important to minimize the structural defects and form high-quality GaN NCs.

[0211] Next, the photophysics of GaN NCs is addressed. Bulk GaN is a direct band gap semiconductor. At 300K, wurtzite GaN has a band gap of 3.4 eV and an exciton Bohr radius of 3.2 nm. Zinc blende GaN has a band gap of 3.2 eV and an exciton Bohr radius of 3.3 nm. The colloidal GaN NCs had dimensions (>6 nm) larger than the Bohr radius, therefore weak size dependent optical properties were expected. This is different from other III-V colloidal quantum dots with large Bohr radius (>10 nm) such as InP, InAs and GaAs, which exhibit strong size dependent optical properties.

[0212] Weak band-edge-like emission and strong trap emissions was observed from colloidal GaN samples synthesized at 450°C. To better understand the trap emission from low-temperature GaN NCs, a combination of femtosecond transient absorption measurements and ultrafast PL measurements captured by streak camera were used. Long-lived ground state bleach features (t >1 ns) were observed from transient absorption measurements. The observation of bleach suggests radiative recombination occurs near the conduction band to the valence band or a deep acceptor level, consistent with a high-quality direct band gap semiconductor. Despite this, the band-edge-like emission is weak and short-lived (<2 ns), while the trap emission has higher intensity and longer lifetime. Therefore, it is inferred that such emission may be due to the presence of hole trapping states since transient absorption indicates a long-lived electron population within the GaN NCs.

[0213] Next, the temperature effect on GaN PL properties is investigated. A remarkable correlation was observed between the appearance of Raman modes, the reduction of trap emission, and the appearance UV photoluminescence in the colloidal GaN NCs. Samples synthesized below 500 °C show poorly defined Raman modes and broad trap-emission from. Samples synthesized at 525°C and above show well-defined Raman modes and strong UV band-edge PL, without much trap emission. These observations strongly suggest the synthesis temperature >500°C is important to minimize the structural defects and mitigate the trap states in GaN NCs. To simultaneously achieve high GaN crystal quality and good colloidal dispersion, the optimal GaN NC synthesis conditions were found to be 525°C using 2.0-3.5 MPa ammonia pressure.

[0214] Transient absorption measurements were performed on 2.0 MPa and 3.5 MPa samples to investigate their excited state dynamics. Samples were pumped with a 280 nm 100Atty. Dkt. No.05400-0073-PCT fs pulse laser, and broadband transient absorption spectra (340-700 nm) and bleach spectra (260-400 nm) were collected separately. Long-lived ground state bleach (t >1 ns) and photoinduced absorption features were observed, consistent with high-quality direct gap semiconductors. A clear spectral difference was observed in the bleach region between two samples: the band-edge exciton bleach maximum was 323 nm for the 2.0 MPa sample and 340 nm for the 3.5 MPa sample. Such a difference indicates the narrowing of the band gap in GaN NCs due to size increase and phase transition. The TA kinetics probed at the bleach maximum showed complex decay traces: triexponential fitted time constants were 25.7 ps, 223 ps, 1.84 ns for the 2.0 MPa sample and 17.0 ps, 165 ps, 2.23 ns for the 3.5 MPa sample. Such slow decay traces over nanosecond timescale are consistent with typical direct-gap semiconductor NCs. The bleach features correspond well to the excitonic features measured by linear absorption of the samples.

[0215] Next, the PL properties of GaN NCs synthesized at 525°C were examined. All samples exhibited strong emission in the UV range, and no trap emission was observed. The PL peaks for the 2.0 MPa, 2.8 MPa and 3.5 MPa samples were 335 nm, 348 nm and 350 nm, respectively. Compared with their transient absorption, the Stokes shifts were ~140 meV for the 2.0 MPa sample and ~100 meV for the 3.5 MPa sample. The time resolved photoluminescence decay traces suggested the 3.5 MPa sample showed longer radiative decay time than the 2.0 MPa sample. The dependence of radiative decay time on emission is also consistent with band edge emission. Both samples show fast decay (τ1) components and long decay (τ2) components by biexponential fitting. The fast decay components (τ1) are typically attributed to non-radiative trapping processes in semiconductor NCs. The 10 ns scale radiative lifetime is consistent with band gap recombination from direct gap semiconductor NCs.

[0216] In conclusion, the optoelectronic characterization demonstrates that the GaN NCs are emitting from a band edge state. High temperature and elevated ammonia pressure are the key tuning parameters to achieve high-quality colloidal GaN NCs. The synthesis strategy used in this Example may be used to provide ternary compounds (InGaN, AlGaN). Surface passivation and shell growth may be used to amplify the band-edge emission of GaN NCs.

[0217] The word "illustrative" is used herein to mean serving as an example, instance, or illustration. Any aspect or design described herein as "illustrative" is not necessarily to beAtty. Dkt. No.05400-0073-PCT construed as preferred or advantageous over other aspects or designs. Further, for the purposes of this disclosure and unless otherwise specified, "a" or "an" means "one or more.”

[0218] The foregoing description of illustrative embodiments of the disclosure has been presented for purposes of illustration and of description. It is not intended to be exhaustive or to limit the disclosure to the precise form disclosed, and modifications and variations are possible in light of the above teachings or may be acquired from practice of the disclosure. The embodiments were chosen and described in order to explain the principles of the disclosure and as practical applications of the disclosure to enable one skilled in the art to utilize the disclosure in various embodiments and with various modifications as suited to the particular use contemplated. It is intended that the scope of the disclosure be defined by the claims appended hereto and their equivalents.

[0219] If not already included, all numeric values of parameters in the present disclosure are proceeded by the term “about” which means approximately. This encompasses those variations inherent to the measurement of the relevant parameter as understood by those of ordinary skill in the art. This also encompasses the exact value of the disclosed numeric value and values that round to the disclosed numeric value.

[0220] In recognition of the inherent nature of chemical synthesis, throughout the present disclosure, terms and phrases such as “free,” “does not comprise,” etc. encompass, but do not require a perfect absence of the referenced entity.

[0221] The term “type” as used herein refers to chemical formula such that a single type means the same chemical formula and different type means different chemical formula. Similarly, use of “more” as in “one or more” refers to use of different types of the relevant entity.

[0222] Terms such as “comprising” and the like may be replaced with terms such as “consisting” and the like.

Claims

Atty. Dkt. No.05400-0073-PCT WHAT IS CLAIMED IS:

1. A method of synthesizing nanocrystals, the method comprising combining a metal molecular precursor comprising a group III element or a transition metal element, and a group V molecular precursor comprising a group V element, in a molten inorganic salt solvent comprising a molten inorganic salt and under conditions to form a metal-V compound in the form of nanocrystals, wherein the metal molecular precursor, the group V molecular precursor, and the molten inorganic salt solvent form a reaction mixture that does not comprise an organic solvent.

2. The method of claim 1, wherein the metal molecular precursor is a group III molecular precursor comprising the group III element and the metal-V compound is a III-V compound.

3. The method of claim 2, wherein the group III molecular precursor acts as a reducing agent to reduce the group V element of the group V molecular precursor under the conditions to form the III-V compound.

4. The method of claim 2, wherein the group III molecular precursor is a molten inorganic salt.

5. The method of claim 2, wherein the group III molecular precursor has formula MmXn, wherein M is the group III element, X is a halogen, m is in a range of from 1 to 2, and n is in a range of from 1 to 5, and the group V molecular precursor is a pnictogen halide or a pnictogen hydride.

6. The method of claim 5, wherein the group III molecular precursor has formula M[MX4], wherein M is the group III element and X is the halogen.

7. The method of claim 5, wherein the group III molecular precursor has formula M[M2X6], wherein M is the group III element and X is the halogen.

8. The method of claim 5, wherein the group III molecular precursor has formula MX, wherein M is the group III element and X is the halogen.Atty. Dkt. No.05400-0073-PCT 9. The method of claim 1, wherein the molten inorganic salt solvent comprises a molten inorganic salt having formula AM’X’4, wherein A is an alkali metal, M’ is a group III element, and X’ is a halogen; an alkali metal halide; a eutectic mixture of alkali metal halides; or a combination thereof.

10. The method of claim 2, wherein the group III molecular precursor has formula M[MX4], wherein M is the group III element and X is a halogen; the group V molecular precursor is a pnictogen halide or a pnictogen hydride; and the molten inorganic salt solvent comprises a molten inorganic salt having formula AM’X’4, wherein A is an alkali metal, M’ is a group III element, and X’ is a halogen; a eutectic mixture of alkali metal halides; or both.

11. The method of claim 10, wherein M is selected from Ga, In, Al, and combinations thereof, and X is selected from Cl, Br, and I.

12. The method of claim 10, wherein the group V molecular precursor has formula PnH3, PnX”3 or Pn2X2”, wherein Pn is the group V element and X” is a halogen.

13. The method of claim 12, wherein Pn is selected from N, P, As, Sb, and combinations thereof, and X” is selected from Cl, Br, and I.

14. The method of claim 10, wherein the group III molecular precursor is Ga[GaI4]; the pnictogen halide is PI3, AsI3, SbI3, or a combination thereof; the pnictogen hydride is NH3; and the molten inorganic salt solvent comprises KGaI4, KGaCl4, a eutectic mixture of alkali metal halides, or a combination thereof.

15. The method of claim 2, wherein the group III molecular precursor has formula MX3, wherein M is the group III element and X is a halogen and further wherein the group V precursor is a silylated pnictogen.

16. The method of claim 1, wherein the conditions include a reaction temperature of greater than 425 ºC.

17. The method of claim 1, wherein the metal molecular precursor is a transition metal molecular precursor comprising the transition metal element and the metal-V compound is a transition metal-V compound.Atty. Dkt. No.05400-0073-PCT 18. The method of claim 17, wherein the transition metal molecular precursor has formula MXn, wherein M is selected from Ti, V, Nb, Mo, Ta, and W; X is a halogen; and n is from 3 to 6.

19. The method of claim 18, wherein the group V molecular precursor is a pnictogen hydride.

20. The method of claim 19, wherein the pnictogen hydride is NH3.

21. The method of claim 1, wherein the nanocrystals of the metal-V compound are single-crystalline nanocrystals.

22. The method of claim 2, wherein the nanocrystals of the III-V compound exhibit band edge photoluminescence upon illumination with light under room temperature.

23. A colloid comprising single-crystalline nanocrystals composed of a III-V compound, the single-crystalline nanocrystals exhibiting band edge photoluminescence upon illumination with light under room temperature.

24. The colloid of claim 23, wherein the III-V compound has Ga, and optionally, one or more of Al and In, as a III element; and N, P, As, Sb, or a combination thereof as a V element.

25. The colloid of claim 24, wherein the III-V compound is GaAs or GaN.

Images