IR-LUMINESCENT NANOCOMPOSITE PIGMENT AND SOL-GEL PROCESS FOR MANUFACTURING THEREOF

BACKGROUND OF INVENTION:

Brand protection and document security is a large and ever growing concern to which governments as well as many private entities are paying close attention. Given the increasing technical sophistication of counterfeiters especially in recent years, traditional security features are gradually losing ground. As a result, there is a continued search for new security features to stay ahead of counterfeiters.

In the last few years, photoluminescent upconversion materials and related applications have been identified as a potential technology platform. The photoluminescent material can be incorporated into inks and coatings for security printing of, for example, banknotes (for instance, currency, checks, and negotiable documents). They can also be incorporated into plastics and other packaging materials for brand-protection, as well as food and drug safety applications. They can also be incorporated into substrates, for example paper, to be used in secured and verifiable document applications. Non-security applications such as printable IR-activated solid-state displays are also possible.

The photoluminescent upconversion pigments currently available commercially are of mostly oxysulfide (e.g., dysprosium oxysulfide, yttrium oxysulfide) or oxide (e.g., yttrium oxide) base materials. One example is the Lumilux® line of upconversion pigments supplied by Honeywell. However, these materials are generally characterized as being very heavy (having a specific gravity of about 6.0), with high hardness, high abrasiveness, sensitivity to pH and a number of solvent systems, and being sensitive to grinding. These undesirable characteristics cause a number of problems particularly in printing applications. For instance, the abrasive nature of these material cause severe wear and tear of expensive printing plates (e.g., gravure

cylinder) and print heads (e.g., inkjet). The high density (i.e., 4 times heavier than regular organic pigments) causes the pigment to settle fast and leads to separation of pigment and varnish, and thus render them virtually unusable, particularly in low viscosity ink system such as inkjet ink and gravure ink. The apparent remedies to this problem, for example installing a stirring device in the ink tank, is usually expensive and in some case impractical. These materials are of crystalline particles without any protecting barrier which makes them sensitive to some solvent systems, and limits them both as to formulation and application flexibility. For example, oxysulfide material will disintegrate in low pH aqueous system and start to lose its photoluminescent properties. It is also understood that in many printing applications such as inkjet, small particle size is highly desirable. These materials are generally of large particle size (e.g., 1 to 10 μm), which is unsuitable. The apparent remedy such as milling and grinding for particle size reduction is hampered by the extreme hardness of these materials in that they are harder than most of grinding media. There is also one more significant drawback for such a remedy. These materials loss much of their photoluminescent properties upon grinding, due to the destruction of crystalline structure.

Inorganic photoluminescent pigments can be conveniently divided into two general categories, ceramic and glass-ceramic, based on the morphology. The former is a one phase system, which is mostly crystalline. The latter is a 2-phase system, with an amorphous phase as continuous matrix, and a crystalline phase as isolated domains therein, i.e., the dispersed phase. When the crystalline phase is of nanometer size, this type of material is also called an inorganic nanocomposite. Despite of many inherent shortcomings as pigments, most of current commercial photoluminescent upconversion pigments belong to the ceramic family. See, for example, EP 1314766 Al.

The ceramic pigments are, in general, heavy with specific gravity around six, are abrasive due to high crystal hardness, are chemically unstable at low pH, and are extremely sensitive to grinding. The high density led to quick settling in ink, which is highly undesirable. High abrasion led to severe wear and tear of printing equipment. Due to the appreciable solubility of oxide and oxysulfide at low pH, such pigments loss their photoluminescent properties gradually as crystal dissolve and disintegrate. The pH sensitivity of ceramic pigments limits flexibility of ink formulation, and excludes them from being used in certain printing processes where aqueous acid fluids are involved. Since the photoluminescent properties (i.e., upconversion efficiency, intensity) depend strongly on the crystal integrity, any disruption of order in the crystal lattice leads to degeneration. It is understood that most commercially available ceramic pigment is of 1-10 μm size, which is far too large for certain printing applications such as inkjet. As a result, extensive milling and grinding are required for particle size reduction. Such grinding leads to fracture of crystal and distortion of crystal lattice at the surface, and thus results in the loss of photoluminescent intensity. In short, due to its physical characteristics, the ceramic type of upconversion pigments are not particularly printing friendly (i.e., printable security features).

Glass-ceramic type of photoluminescent materials has been known for more than ten years in optoelectronic industry and in the past ten years, glass-ceramic upconversion material has been of intensive interest both for academia and industry, mainly for its promising potential to be next-generation optic fiber, optical amplifier, phosphors and lasing materials (Fujihara S., Kato T., Kimura T., "Sol-Gel Synthesis and Luminescent Properties of Oxyfluoride LaOF:Eu ³⁺ Thin Film", J. Mater. Sci. Lett., 20:687-689, 2001; Biswas A., Marciel G.S., Friend C.S., Prasad P.N./'Upconversion Properties of Transparent Er ³⁺ -Yb ³⁺ Co-doped LaF3-Siθ2 Glass-Ceramics Prepared by Sol-Gel Method", J. Non-Cryst. Sol. 316:393-397, 2003; Tada M, Fujihara S Kimura T,

Sol-Gel Processing and Characterization of Alkaline Earth and Rare Earth Fluoride Thin Film, J. Mater. Res., 14(4):1610-1616 (1999); US 6,385,384; US 5,955,388).

While the concept of adapting luminescent material for security-related applications has been acknowledged by individuals and companies in security field, see e.g., Becidyan, Luminescent Pigment in Security Applications, Color Research and Application 20(2):124-130 (1995), there has been a dearth of reported work or research efforts in synthesizing or manufacturing this material in such way that the physical performance is made suitable for printing applications as opposed to optoelectronic components (e.g., optic amplifier). Almost all work has been done with ceramic materials. There are no reports in the literature of presenting data concerning any attempt to identify the significant modification and redesign (i.e., in composition, morphology control, and post-processing) of glass-ceramic materials described in the prior art (Biswas 2003, supra; Fujihara, 2001, supra; Tada 1999, supra; Fujihara S., Mochizuki C, Kimura T., "Formation of LaF3 Microcrystals in Sol- Gel Silica", J. Non-Cryst. Sol. 244;267-274, 1999) necessary to impart the physical characteristics that are necessary for obtaining a printing pigment.

U.S. Patent number 6,613,137 describes a pigment comprising glass ceramic composite particles containing crystalline particles embedded in a glass matrix, but the description is very general and no data or experimental details are presented. It is stated that the pigment can be made by saturating glass with a rare earth fluoride and heating to cause crystals to form provided that crystallization does not occur to the extent that transparency is adversely impacted. It is apparent that as a result, not all of the total rare earth fluoride present is crystallized and the molar content of the crystals weill be limited.

The current invention addresses many issues mentioned above, and rectifies the problems mentioned above with a new family of photoluminescent pigment of different physical construction (i.e., a nanocomposite instead of naked crystal) and a

novel manufacturing process for producing the pigment. The design and production process imparts many desirable properties to the pigments, such as low density, low abrasion, low opacity, low sensitivity to solvent, low sensitivity to grinding, ease of particle reduction, as well the flexibility in manipulating these critical performance characteristics.

The differences and advantages to the prior art in terms of pigment composition, precursor formulation and production process will be apparent from the discussion below. Nevertheless, a few key differences between the prior art and the invention can be mentioned here. The most important distinction is the rare-earth fluoride crystal content. In the closest published information in terms of synthesis route and starting materials - Biswas, 2003, Fujihara, 1999, 2001 and Tada 1999, the highest rare earth fluoride content reported were 5.2%mol and 12.0%mol, respectively. The current invention achieves a 20%mol to 60%mol of rare earth fluoride content without losing the integrity of glass matrix. In the paper by Wang ("New Transparent Vitroceramic Codoped with Er ³⁺ and Yb ³⁺ for Efficient Frequency Upconversion", Appl. Phys. Lett. 63(24) :3268-3270, 1993), which is often referred and quoted as the landmark paper on glass-ceramic phosphors, a composition with ll%mol rare earth fluoride (i.e., Yb + Er), 44%mol transitional metal fluoride and 45%mol silica and aluminum oxide was reported. This composition differs from the current invention in at least two critical aspects. First of all, Wang's synthesis route is the conventional hot-melt (dry) method vs. sol-gel (wet) method used in current invention. Another difference is again the rare-earth fluoride content, which is only ll%mol in Wang's recipe, far below the compositional window of current invention. Last but not the least, the incorporation of lead fluoride in the composition will makes it nearly unusable as pigments for many printing applications, due to regulatory issues associated with lead related compounds. The two U.S. patents cited above disclose higher contents but the method of synthesis is of conventional hot-melt process similar to that of Wang.

SUMMARY OF INVENTION:

This invention concerns a family of new photoluminescent pigments, and more specifically a family of infra-red laser responsive glass-ceramic nanocomposite pigments, which are intended for security and display applications. The pigments emit visible light under the excitation of an (invisible) infra-red laser, a phenomenon generally referred as quantum upconversion. The pigments are of particle-in-matrix construction. The embedded phase is a rare earth fluoride crystal, which is responsive for quantum upconversion, and is of nanometer in size to maximize the surface area and the response to IR-laser. The continuous or encapsulating phase is amorphous glass such as silica, to provide a barrier to the rare-earth fluoride crystals. The pigments are synthesized via a sol-gel process, which is specifically designed for producing materials highly suitable for printing applications.

The pigments of the invention are characterized by a crystal concentration of about 20-60 mol%, preferably about 40 to 50 mol%, a low density of about 2.5 to 4 g/cm ³ , preferably about 2.5 to 3.0 g/cm ³ , a hardness which is low enough to avoid printing apparatus abrasion, substantial transparency, low sensitivity to solvent and grinding in that the pigment retains at least about 90 %, preferably at least about 95 %, of the quantum upconversion when exposed to organic liquids in the ink or coating in which they are contained or ground to pigmentary size. They are of pigmentary size, i.e., about 0.1 to 1 micron, and preferably about 0.1 to 0.5 micron.

BRIEF DESCRIPTION OF THE DRAWING

Figure 1 illustrates a manufacturing process to make the pigment of this invention.

DESCRIPTION OF THE INVENTION

This invention is regarding a new family of glass-encapsulated security taggants with a rare-earth fluoride crystalline phase of about 20-60 mol%, and a sol-gel manufacturing process for producing thereof. They can be called glass-ceramic nanocomposite pigments based on morphology and can also be called IR-VIS photoluminescent pigments based on their optical characteristics. They are also sometimes referred as laser-activated or laser-responsive security taggants based on the working mechanism. This family of pigments has a higher rare-earth fluoride crystal loading than in the prior art. The manufacturing process imparts the pigments of current invention with desirable characteristics (i.e., solvent inertness, pH resistance, low abrasion, low settling and low opacity) for printing applications.

The primary application for this family of pigment is security printing such as banknotes, checks, postal stamps, event tickets, passport other verifiable documents. Other applications includes but not limited to printable solid-state displays and substrate taggants. Given that the printing is primary way applying such taggants, the "printability" of the taggants is of paramount importance.

Taggants (pigments) can be directly printed or painted onto substrates in forms of inks or coatings, or directly incorporated into substrates via methods such as co- extrusion and co-precipitation. The presence of taggants can be verified via two modes, IR laser activation - visual recognition — or IR laser activation - machine recognition. The former uses semi-overt authentication. An inquirer can verify the existence of security taggants by observing light emission (i.e., red, green, blue) from the substrate, which is under the illumination of an invisible IR laser (e.g., using an inexpensive low power 975 nm laser pen). The second mode is of covert nature. Instead of describing the emission as red, green or blue light, an electronic reader

(photo-detector) detects the characteristics emission spectra, which often consist of multiple peaks. They are the "fingerprints" of the security pigments. Two "green lights", which appear to be similar to a casual observer's naked eyes, can register very different emission spectra on electronic reader. In short, the naked eye see the overall appearance of combined visible emissions peaks, whereas electronic readers see the discrete emissions peaks in both visible and invisible region. This dual capability for semi-overt and covert authentication provides added security, and is particular suitable for applications such as currency and banknotes.

The emission spectra and overall visual color is a strong function of the dopant (active ions in the crystalline phase which imparts specific photoluminescent properties as it is the actives center responsible for photon absorption and re- emission), dopant concentrations, and dopant ratios. Given the wide range of dopants and dopant combinations possible, many unique spectra fingerprints can be generated by a compositional change. This allows for the creation of a family rather than a few security taggants based on essentially the less same technical (synthesis) platform. For instance, examples 1 and 2 below shows how a small variation in dopant composition can produce two different IR-luminescent pigments, one green light emitting and one blue light emitting. This flexibility also allows for making tailored security taggants for specific customers, in a promptly and inexpensive manner, so that the users can stay ahead of counterfeiters. Since precursor dopants are used in very small quantity in the pigment, and final calcination (firing) also effectively erases the precursor history, forensic analysis and reverse engineering are very difficult even for people skilled in art. The large bank of unique fingerprints and the flexibility in tailoring products, in combination with high barrier for reverse engineering (i.e., long-life cycle) make this technology very attractive for security applications.

It is understood that there are three major performance criteria for security taggants (pigments). They are

1. Uniqueness (i.e., spectra complexity, hard to reverse engineering)

2. Fitness for Printing (i.e., sedimentation, abrasion, inertness, etc)

3. Detectability and Ease of Authentication (i.e., emission intensity, detection threshold)

The importance of uniqueness, in this case unique emission spectra, is self-evident. It is the very concept of anti-counterfeiting technology. It is achieved in this invention through the manipulation of dopants type, dopant concentration and dopant ratio.

Fitness for printing refers to a collection of physical requirements related to printing and its importance is self-evident (e.g., banknote and check printing). Based on physical requirement for printing, it was concluded by the inventors of current invention that it was necessary to have light emitting crystals embedded in a carrying matrix which is light-weight, porous and yet non-penetrating, chemically inert, soft-textured and yet with decent mechanical integrity. The light weight matrix is necessary to reduce overall pigment density, and thus less pigment sedimentation in ink. Since crystal density can not be adjusted easily, density reduction is achieved by making the carrying matrix porous so that the density of the pigment itself is on the order of about 2.5 to 4.0 g/cm ² . It is also understood that carrying matrix needs to provide a barrier function to the embedded crystal to protect its photoluminescent properties (from degeneration by oxygen, moisture, ionic species, solvents). As a result, the carrying matrix should be non-penetrating to those harmful species. Balancing the competing objectives of low density and good barrier function, the engineering requirement for carrying matrix becomes that the ultrastructure of matrix should substantially be high-void uniformly microporous (i.e., high void fraction with small uniform pore and channels that are impenetrable

by the harmful species). The high void requirement means that the matrix must have sufficient voids to load about 20 to 60 mol% dopant (based on the total pigment) and the small pores and channels should be close to the size of the dopant, i.e., up to about 75 ran (in order to accommodate the preferred about 20-50 ran- nanocrystals). In general, the pore or channel size should not exceed about 150% of the dopant particle size, which is less than 500 nm, and preferably less than about 100 nm. In the current invention, glass carrying matrix was prepared (synthesized) to achieve both objectives via DCCA (drying control chemical additives) technology, and the appropriate selection of glass precursor, diluent solvent, inhibitor and catalyst concentrations and ratios. For general background on DCCA technology, one may consult L.L. Hench, Sol-Gel Silica: Properties, processing and Technology Transfer, Noyes Publication, 1998. Based on the requirements described herein, one skilled in this art can fabricate the pigments based on formation information available in the literature. The fabricator must also recognize that the carrying matrix should not be abrasive to in printing plate (e.g., in gravure printing) or printhead (e.g., in inkjet). The abrasiveness of a pigment is function of inherent hardness and fracture toughness of the material, pigment geometry and pigment surface morphology (i.e., smooth or with sharp edge). The description below addresses the abrasion issue by choosing silica as the matrix material, and lowering its fracture toughness by increasing matrix porosity. When experiencing high friction with the printing plate, this allows the carrying matrix to fracture and dissipate the energy rather than making scratches on the plate surface. The low fracture toughness of the glass matrix has another added benefit during grinding and milling for pigment particle size reduction. A lower fracture toughness compared to the embedded nanocrystals ensures that the cleavage or breakage site is predominantly in the amorphous (glass) domain rather than crystalline domain since materials always fails at their weakest point. A soft-textured glass matrix acts as a scarifying agent for embedded crystals, to ensure the crystals are intact and so are the photoluminescent properties. This is the major advantage over the ceramic type

of security pigments, which are extremely sensitive to grinding. With primary- crystals having a size of about 20nm to 50nm, the size of the pigments of current invention can be as low as about 100 nm without losing significant upconversion intensity, whereas the ceramic type of photoluminescent products lose much of their upconversion intensity upon grinding. This advantage allows the pigments of current invention for wider arrays of printing applications, particularly where small particle size is desired.

As for the third criteria, the term "detectability" refers to how easy can a viewer see the visible upconversion, or how easy can an electronic reader pick up the emitted light. Detectability is the apparent function of four factors, inherent emission intensity of a pigment, the pigment concentration in ink film, the power of activating laser and the sensitivity of eyes or readers. From pigment performance stand point of view, the only thing that can be improved to impact detectability is inherent emission intensity (on per unit weight base). There are three possible ways to achieve high emission intensity per unit mass:

1) Increase the glass matrix's transparency, and thus less input power loss (i.e., more power in means power out), as well as less scattering of visible emission at specular angle

2) Increase crystal loading, and thus more receiving surface for laser per unit volume

3) Increase crystal upconversion efficiency, and thus more emitted light for same amount of laser input

AU three routes can be exploited in the current invention. In many instances, good matrix transparency (i.e., matrix homogeneity) is a competing objective to high crystal loading, since the crystal precursor often interferes with the glass network formation. In the current invention, we found that this dilemma can be successfully

addressed by the selection of glass and crystal precursor formulation, dopant concentration and dopant ratio, gelling and drying algorithm, and incorporation of suitable drying control chemical additives (DCCA). As a result of such efforts, we were able to achieve a higher crystal loading than prior art, and with good properties for printing applications. Typical conditions are presented in the examples below but other suitable combinations can be readily determined by those skilled in this art from this description combined with routine optimization following that description.

The Manufacturing Process

The IR-VIS upconversion pigments are an inorganic nanocomposite, more specifically a rare-earth doped glass-ceramic material. The rare earth can be Y, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb or Lu, or a combination of them. The active ions, e.g., doped lanthanum fluoride crystals, constitute the inner or dispersed phase, whereas the amorphous silica (glass) is the outer continuous phase. The doped LaF3nanocrystals, typically about 20 to 50 nm in size, are responsible for quantum upconversion, whereas the glass matrix provides environmental insulation to LaF3 crystals to prevent them from various chemical attacks. The soft and microporous glass matrix also allows for easy particle reduction and provides buoyancy to the heavy crystals. Low crystal loading will yield low emission intensity and low density, whereas high crystal loading does the opposite. In the current invention, the rare earth crystal loading is about 20% to 60%mol. The density of the pigment is about 2.5 g/cm ³ to 4.0 g/cm ³ . They are designed to suit various applications, based on the different needs in pigment coverage, visibility and settling requirements.

The inventive pigments are manufactured via a sol-gel process. It consists of five key steps:

1. Precursor formulation and mixing

2. Controlled gelation

3. Controlled drying

4. Controlled calcination

5. Pigmentation and Post-processing

1. Precursor Formulation and Mixing: Glass-matrix precursor solution and rare- earth crystal precursor solutions were prepared separately and then combined for homogenization. In broad terms, the glass matrix precursor solution contains a silica precursor, hydrolyzing agent, hydrolysis inhibitor and drying control chemical additives (DCCA) which can provide desired characteristics to the matrix formed. A preferred preparation formulation is a combination of TEOS (tetraethyl orthosilicate), ethanol, water, TFA (trifluoroacetic acid) and DMF (dimethylformamide) but other combinations can be employed. In the preferred formulation, the TEOS can be hydrolyzed and then condensed into the glass backbone, water is a hydrolyzing agent, ethanol is a hydrolysis inhibitor and gel volume extender, and the TFA is both hydrolysis and condensation catalyst. The DMF influences the condensation rate, primary particle size of silica, and pore size distribution of matrix. The balance between these agents determines whether a homogenous, transparent, high porosity and yet well-insulated matrix can be formed. Preferred combinations are set forth in the Table below and recipes for green upconversion pigment and blue upconversion pigments are also provided in the examples but the invention is not limited to these.

The rare-earth crystal precursor solution is prepared by dissolving one or more rare earth salts, such as acetate salts, in a mixture of a fluorine source and water. The dissolution is carried out under chilled condition to remove the heat from the hydration of TFA and dissolution heat of the salts. TFA is a preferred fluoride source for rare earth ions, and will form a complex with the rare earth ions such as erbium, ytterbium and lanthanum ions. The water assists in dissolving acetate salts.

After the preparation of both precursor solutions, the rare-earth crystal precursor solution is slowly added into glass-precursor solution under stirred and chilled

conditions. Overly fast addition and insufficient agitation can lead to premature condensation of glass-precursor, which lead to heterogeneity in the system.

Table 1

The pigments of the present invention can be formulated into inks and coating by combining the pigment with a carrier suitable to the intended application in an effective colorant amount. In general, that amount is about 0.1 to 50 % of the composition. Any ink vehicle, and preferably a security ink vehicle, as well as any coating vehicle can be used as the carrier. The formulation of inks and coatings is well known in the art and to the extent necessary, reference may be made to Chapter I of the Printing Ink Manual (Fourth Edition), incorporated herein by reference, which reviews the fundamentals of ink formulations.

In order to further illustrate the invention, various non-limiting examples are given below. In these examples, all parts and percentages are by weight and temperatures in degrees Centigrade, unless otherwise indicated.

Example 1: Precursor Formulation of a Green Upcon version Pigment

Pigment Chemical Composition: 1. OErF ₃ -10YbF ₃ -37LaF ₃ -52Siθ2 (by mole)

This pigment can be made by the procedure described in Example 3.

Example 2: Precursor Formulation Super-High Crystal Loading Blue Upconversion Pigment

Pigment Chemical Composition: 0.36TmF ₃ -7.92YbF ₃ -0.18YF ₃ -39.54LaF ₃ -52SiO ₂ (by mole)

This pigment can be made by the procedure described in Example 3.

Example 3: Typical Operating Procedure for Precursor Mixing

The following is a typical procedure for precursor mixing using 0.3TmF3-6.6YbF3-

0.15YF3-32.95LaF3-60SiO2 as an example and the process shown in Figure 1.

Into vessel #2 were pumped 108 parts of water, followed by 276 parts of ethanol and 35.04 parts DMF. Then 2.292 parts Tm(Ac)S, 0.906 parts Y(Aφ, and 27 parts distilled water were mixed in a glass vessel until dissolved. The solution is stored in the "dopant concentrate" tank (see Figure 1). Next, 232 parts La(Aφ and 50.95 parts Yb(Aφ were put into a powder charger. Eighty-one parts of water were pumped into reaction vessel #1 and agitated at 300 RPM. Chilled water was used to keep the vessel wall temperature at around 5 ⁰ C.

Into vessel #1 were pumped 324 parts of TFA over 4 hours, and the vessel temperature kept below 25°C. 41 parts of TFA were pumped into vessel #2 at a slow speed (over time span of 1 hour). The vessel was cooled by chilled water and with vessel wall temperature set at around 5 ⁰ C. Then the La(Ac)3+Yb(Aφ powder was charged slowly into vessel #1 in four one hour portions, under 300 RPM agitation. After all solids dissolved in vessel #1, the dopant concentrate solution was introduced into vessel #1. Next, 249.6 parts of TEOS were pumped into vessel #2 under agitation, and stirred for one hour for homogenization upon completion of pumping.

The solution from vessel #1 was pumped into vessel # 2 slowly over a time span of 4 hours under agitation (300 RPM) and cooling. The mixing process is highly exothermic and care is taken to control vessel temperature to be under 25 ⁰ C. Above that temperature, pre-mature gelation or TEOS precipitation can occur, leading to

failure of production. The vessel might need to be cut-open for cleaning if gelation occurs. Upon completion of pumping, the solution was stirred for additional 0.5 hour for homogenization.

2. Controlled Gelation: After mixing, the homogenized mixture is then pumped into perorated gelling and drying vessels. The venting cross-section to evaporation surface ratio, heating surface to volume ratio, and characteristic (shortest) heat transfer dimension (Dmin) were appropriately controlled in light of the facts that the failure to keep Dmin constant and at below a critical value can lead to severe thermal stress and consequently heterogeneity of sample. The Dmin is dependent on gelation temperature. The higher the gelation temperature, the lower is the Dmin value. At a gelation temperature between 3O ⁰ C to 70 ⁰ C, the preferred threshold is about 0.5 cm or below. After pumping the sample into the gelling and drying vessels, the sample is then put into a low temperature (about 30 to 70 ⁰ C) baking oven for gelation. The gelation stage can last about 24 hours. During gelation, the silica backbone percolates through the solution volume and a transparent, monolithic and nearly crack-free gel forms. At this point, thermal stress is low, and thus no syneresis (shrinkage) occurs. If syneresis occurs, the rare earth solution can be expelled from the gel, which is highly undesirable, since it results in compositional change in the pigment. The gelling temperature is selected in such way that it balances need for fast gelation and the desire for low thermal stress.

3. Controlled Aging and Drying: After 24 hours of gelling, the samples are then moved into drying stage. In drying stage, the solvents (ethanol, acetic acid, excess TFA, DMF, water) evaporate, the gel shrinks, and silica network densifies. This stage can lasts about two weeks. The temperature is raised in steps between 5O ⁰ C to 200 C. with the thermal treatment profile designed in such way that it matches the network contraction rate with evaporation rate, so that no solution expulsion occurs. The

currently preferred thermal profile consists of three step heating. The end product at the drying stage should be a dense, glassy and transparent dry gel. At this point, glass formation is completed and an amorphous rare earth trifluoroacetate microphase is forming. The domains are very small and hardly visible under high power TEM. XRD (X-ray diffraction) shows a flat baseline and thus confirms no crystalline structure exists.

4. Controlled Calcination: After drying, the samples are then sent into a high temperature ceramic furnace for final calcination under a programmed profile. The upconversion intensity is highly dependent on the way that the material is calcined, since it dictates how and when the organic components are burned off, and how and when rare earth fluoride crystals precipitate in the glass matrix, and how and when the glass matrix further densify. The thermal profile has a tremendous impact on degree of crystallinity, and the texture of glass matrix. A typical algorithm is l°C/min slow heating to 65O ⁰ C - hold at 650°C for 2 hours - continuing slow heating at rC/min to 900°C - furnace cool at 20°C/min. The first heat ramp from room temperature to 650°C is to slowly burn-off the organic components inside the dry gel. The heating rate is preferably below 2°C/min to minimize thermal stress due to gas expansion (e.g., CO2, CO, HF from organic species degradation). The isothermal holding at 65O ⁰ C for two hours has two purposes, letting residual gas escape and encouraging crystal growth of rare-earth fluoride. It is understood that silica undergoes a quick densification (pore closing) at above 700°C. As a result, at above 700°C, the pores in the glass matrix will become too small or too disconnected to allow for gas escaping. The trapping of organic species, which are known for excellent IR absorption, will impede the luminescent performance of final pigment. As a result, the holding temperature is set to be just below 700 ⁰ C threshold to keep pores open for gas, and yet high enough to initiate the rare-fluoride crystal growth. Following the holding step, the temperature continue to rise at slow rate to a ceiling

temperature at 900°C. During this step, the pores in glass matrix are closed and channels become disconnected. In another word, the barrier function (against moisture, oxygen and other environmental factors) of glass matrix is established during this step. The slower the heating rate, the denser the glass will become. A balance needs to be struck between good barrier function and low density. In the current invention, and we have found about 0.5 to 2.0°C/min is a good operating window. During this stage, the rare-earth fluoride also continues to grow, evidenced by the ever increasing crystalline peak intensity by XRD. As temperature pass 900 ⁰ C, the oxyfluoride crystalline peak start to show up on XRD spectra and the rare earth fluoride crystalline domain and silica glass domain start to merge. This transition is also marked by the sharp drop-off of photoluminescent intensity. To retain the high emission intensity, the ceiling temperature should not go beyond 900°C in this example. After reaching 900°C, the material is allowed to undergo furnace cooling at about 20°C/min. At the end of calcination, a glossy, dense and translucent to opaque pigment is formed. It can be grinded to the desired particle size depending on the desired application. XRD and TEM can be used to confirm the formation rare-earth crystal and size of the crystals can be calculated. In the current invention, both TEM and XRD can be used to confirm the average crystal size is about 10 to 50 nm.

5. Pigmentation:In this step, calcined virgin material is ground (or pigmentized) to the right particle size, depending on the desired application. The particle size is generally about 0.1 to 10 microns, and preferably about 0.1 to 0.5 microns. The particulate pigment is then combined with a suitable varnish, such as SunBlue-HL 1049-054 (available from Sun Chemical Corp.).

A typical process for pigmentation, designed for inkjet application, is given below in example 4.

Example 4: Milling Process for Preparing Inkjetable Upconversion Pigment

The monolithic upconversion glass-ceramic (cm starting size) is milled in a ball mill and sieved to collect the under 106 μm fraction. The powder is mixed with water and dispersant to make a slurry using a homogenizer keeping the solid content around 10-15%wt, with the dispersant content being 1-5% to the total weight. The dispersant should provide adequate stabilization to the pigments at milling stage, without causing any rheological problem at the final ink formulation. The slurry is charged into a re-circulating mill containing 0.5 mm zirconia media and milled for 6- 10 hours until the desired particle size is reached. The progression in particle size reduction can be monitored using an off-line dynamic laser light-scattering (DLLS) measurement device.

The ground slurry is sieved to eliminate broken grinding media and then vacuum filtrated before being diluted with water to about 5%wt consistency under constant stirring before spray drying at 120°C to 170°C. The collected pigment is letdown into an ink vehicle for a sedimentation test, and for both dynamic light scattering measurement and TEM inspection, to make sure over 99%wt pigment is below 1 μm, and polydipsersity is below 2.5. Any pigment that fails the test is re-slurried and re- milled.