Phosphors with improved lumen output and lamps made therefrom

Abstract

A phosphor having a continuous protective bi-layer coating of alumina surrounding silica surrounding the phosphor particles is disclosed. The method of making a bi-layer coating on phosphor particles is also disclosed. The first layer surrounding the phosphor is silica. The second layer surrounding the phosphor is alumina. The bi-layer phosphor is useful in fluorescent lamps providing improved maintenance and brightness. The bi-layer phosphor can also be used in high color rendition lamps employing blends of phosphors.

Description

The present invention relates to improved manganese activated zinc silicate phosphors. More specifically, the present invention involved coating a willemite phosphor with a layer of silica and then a coating of alumina. The "bi-layer" phosphor obtained shows improved lumen characteristics and is particularly useful in fluorescent lamps.

BACKGROUND OF THE INVENTION

During certain phosphor synthesis and lamp fabrication steps, the finely divided luminescent materials may be exposed to oxidizing (oxygen-rich) atmospheres at elevated temperatures. An example of this is the so-called `lehring` process used to burn away organic aqueous lamp coating dispersion. It is well known that the brightness of the finished fluorescent lamp may be reduced significantly as a result of the lehring operation (the so-called `lehrloss`). This reduction in brightness may result from a partial oxidation of reactive low valence ions present in the phosphor lattice.

A somewhat more involved example relates to the process in U.S. Pat. No. 4,585,673 wherein the formation of protective coatings (typically alumina coatings) upon the surfaces of finely divided phosphor particles via chemical vapor deposition using an organometallic precursor in a gas-fluidized bed is disclosed. When manganese-doped zinc silicate is alumina-coated via the process described in the '673 patent, and when fluorescent lamps are fabricated from the coated phosphor produced therefrom, these lamps display much better lumen maintenance than do similar lamps fabricated using the virgin (uncoated) zinc silicate phosphor. During the fabrication of such lamps, the phosphor particles are typically dispersed in an aqueous medium. Unfortunately, if the water-based suspension is held-over for several days before use (a typical situation), the beneficial effects associated with the '673 coating are lost.

This `holdover` problem can be overcome, however, by annealing the alumina-coated phosphor in the air at a temperature between about 700.degree. C. and about 850.degree. C. for a period of time ranging from about 15 minutes to about 20 hours as described in U.S. Pat. No. 4,805,400. Unfortunately, while this coated phosphor annealing process solves the holdover problem, it also causes the zinc silicate phosphor to react with the alumina coating. Zinc and manganese diffuse into the alumina coating, probably forming a mixture of zinc and manganese aluminates. The coated phosphor develops a `body color`, and suffers a reduction in visible light emission upon exposure to an ultraviolet light source. Moreover, very similar phenomena are observed, as well, when the virgin (uncoated) phosphor is subjected to the annealing process. The increased body color and reduced brightness which result from annealing both the uncoated and the '673 coated zinc silicate phosphor are believed to partially result from the oxidation of some of the divalent manganese ions located on the surface of the uncoated phosphor particles or within and on the surface of the reactive alumina coating.

Prior to the present invention, there was no known means of preventing these detrimental interactions between the phosphor and the oxygen-rich atmosphere within the annealing furnace. By means of the method described below, these detrimental interactions are virtually eliminated, allowing a phosphor coated by the process disclosed in the '673 patent to be thoroughly annealed without suffering reflectance or brightness losses.

Another aspect of the present invention involves the use of high brightness zinc silicate phosphors as components of the triblend phosphors. As discussed previously, coated zinc silicate phosphors are unstable in the water based suspension systems used to manufacture fluorescent lamps. The zinc silicate phosphors must be annealed to stabilize the coated phosphor. However, the performance of a zinc silicate phosphor suffers both in terms of brightness output and lumen maintenance after annealing. Attempts to improve the base phosphor performance prior to annealing include remilling and refiring (RMF) the phosphor. Such an alumina coated "RMF" phosphor shows improved lumen characteristics when used as a component in the high color rendition triblend layer. However, the remilling and refiring process results in a large loss of starting material, thus increasing cost of the phosphor. The present invention solves these problems in a novel and economical way.

In a related application, an improved compact fluorescent lamp can be manufactured using a coated willemite phosphor as the green-emitting component. Compact fluorescent lamps of the twin-tube and double twin-tube variety have become important for energy conservation in recent years since they have efficiencies which far exceed those of conventional incandescent lamps. While these lamps are very cost-effective with very short payback periods, they, nevertheless, have high initial costs which have limited the scope of applications in which they have been exploited. Therefore, it is desirable to further reduce the cost of these lamps through the use of less expensive non-rare-earth-containing substitutes.

The compact fluorescent lamps currently employ two rare-earth based phosphors. They are Y.sub.2 O.sub.3 :Eu (Sylvania Type 2342) for the red emission and Ce,Tb Mg Aluminate: Ce, Tb (Sylvania Type 2293) for the green emission. No blue-emitting phosphor is required, since the blue components of the mercury discharge are used to achieve the proper color temperature of the emitted `white` light. More recently, LaPO.sub.4 :Ce,Tb, manufactured by Nichia Corporation, is being considered as a replacement for the Type 2293. Because these materials contain expensive rare-earths as the activators, they are some of the most expensive phosphors commercially used.

As mentioned previously, a green-emitting zinc orthosilicate phosphor activated with manganese, also known by the mineral name willemite can be improved by the application of a bi-layer coating prior to annealing. The bi-layer consists of a thin coating of silica applied between the base phosphor and a conformal alumina coating which is exposed to the mercury discharge. The base phosphor is a zinc silicate phosphor doped with manganese and tungsten as described in U.S. patent application (Ser. No. 06/902,265) which is still pending. This phosphor can be manufactured on production scale equipment using a single step firing procedure which provide very high yields (typically 90%). These high yield and efficiencies of scale provide substantial phosphor cost savings which far outweigh the cost of applying the intermediate silica layer.

Finally the use of the willemite phosphor which has been coated with silica and then with alumina can be used as the green-emitting component of high color rendition fluorescent lamp.

SUMMARY OF THE INVENTION

A method for forming a continuous layer of silica on phosphor particles is disclosed. The method comprises vaporizing a silicon containing precursor such as tetramethyloxysilane (TMOS) or tetraethoxyorthosilane (TEOS) into an inert carrier gas and passing this gas containing TMOS or TEOS through a phosphor powder wherein the phosphor particles are enveloped in the TMOS or TEOS at at a temperature of greater than 400.degree. C. A oxidizing gas is passed into the phosphor powder which reacts with the TMOS to form a continuous coating of silica on the phosphor particles. The resulting silica coated phosphor can then be further coated with alumina.

In another aspect of the present invention, fluorescent lamps using the double coated bi-layer phosphor are manufactured. The resulting lamps show improved lumen and maintenance performance.

In yet another aspect of the present invention the double coated phosphor can be used in high color rendition fluorescent lamps. The bi-layer phosphor replaces the green emitting rare earth phosphors that are presently used in these lamps. The bi-layer phosphor results in a less expensive lamp as the expensive green-emitting phosphor are replaced while the lamps suffers almost no loss in lumen output.

OBJECTS OF THE INVENTION

It is an object of the present invention to improve the brightness of a fluorescent lamp which contains a willemite phosphor by protecting the phosphor with an alumina/silica bi-layer.

Yet another object of the present invention is to achieve substantial brightness improvements in both large and small particle size willemite phosphor.

A further object of the present invention is to provide a silica diffusion barrier over the surface of a phosphor to prevent detrimental reactions from occurring during the annealing of the phosphor.

Yet another object of the present invention is to provide a compact fluorescent lamp of the twin-tube variety which uses a two-layer coated willemite phosphor as a component in the phosphor blend.

Another object of the present invention is to provide a two-layer coated willemite phosphor as the green-emitting component of single and multiple layer high color rendition fluorescent lamps.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic representation of an apparatus suitable for coating phosphor particles.

FIG. 2 is a bed temperature versus time graph for a TMOS/O.sub.2 coating run using a manganese activated zinc silicate phosphor.

FIG. 3 shows the weight percent of silica on phosphor powder as a function of coating time.

FIG. 4 shows the relative plaque brightness versus weight percent of silica coating on a zinc silicate phosphor.

FIG. 5 shows a cross-sectional view of a phosphor particle coated with a layer of silica which is coated with a layer of alumina.

FIG. 6 shows an elevational view of double coated fluorescent lamp.

FIG. 7 shows a cross-sectional view of the lamp of FIG. 6.

FIG. 8 shows the color points taken at 3 positions of a double coated lamp using Type 2293 phosphor.

FIG. 9 shows the color points taken at 3 positions of double coated lamp using an "RMF" phosphor.

FIG. 10 shows the color points taken at positions of a double coated lamp using the phosphor of the present invention.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

One aspect of the present invention involves the formation of a continuous and conformal coating of silica on the surfaces of zinc silicate or cool white phosphor particles via chemical vapor deposition (CVD) while the phosphor particles are suspended within an isothermal gas fluidized bed. In a second aspect of the present invention silica coatings are used to prevent reductions in brightness and the development of body color when manganese activated zinc silicate (Zn.sub.2 SiO.sub.4 :Mn) phosphors are heated in air at temperatures above about 600.degree. C. These silica coatings also act as diffusion barriers, preventing the migration of zinc and manganese from the surface of the Zn.sub.2 SiO.sub.4 :Mn phosphor through the silica coating and therefore also through continuous and conformal alumina coatings that may be formed on the surfaces of the silica coated phosphor particles.

A schematic representation of the fluidized bed reactor used to coat the phosphor particles with silica is shown in FIG. 1. In FIG. 1, a feeder line 11 carries the inert bubbler gas through valve 54 into a stainless bubbler 12 which contains a silicon containing precursor such as tetramethoxysilane (TMOS) or tetraethoxyorthosilane (TEOS). In the bubbler 12, the coating precursor, TMOS or TEOS is vaporized into the bubbler gas. The bubbler is heated by heating means such as heating tape (not shown). The bubbler gas containing the TMOS or TEOS can be diluted by carrier gas to provide appropriate concentration of reactants. The bubbler gas containing the vaporized TMOS or TEOS is carried through connector line 13 and is diluted by the carrier gas at valve 55 which is carried through line 111. Lines 13 and 111 join and the resulting line is heated by heating tape 30 or other means. The bubbler and carrier gas with the TMOS passes through a stainless steel plenum 40 which is maintained at a temperature of about 32.degree. C. The carrier gas along with the vaporized TMOS or TEOS then flows through a porous stainless steel gas distributor 14. The gas then flows into a quartz glass reaction tube 15. Within the reaction tube 15 is a vibrating mixer 17. Circumferentially located on the shaft of the vibrating mixer 17 and near the vibrating disc 19 are a series of holes 18 through which the oxidizing gas with or without an inert diluting gas enters the reaction tube 15. Oxygen is introduced to the reaction tube through line 21. No diluting gas means for the oxygen is shown in FIG. 1. The quartz glass reaction tube is surrounded by a furnace 20.

EXAMPLE 1

Aluminum Oxide C (0.1%) was blended with each phosphor as a fluidization aid. The temperature of the fluidized bed reactor was maintained between 450.degree. C. and 460.degree. C. during the coating process. Further, due to the moisture produced within the high temperature fluidized bed as a byproduct of the TMOS oxidation reaction, the fluidized bed remained almost perfectly isothermal from the beginning to the end of each coating process run. A typical bed temperature versus time curve for a TMOS/O.sub.2 coating run is shown in FIG. 2. In a typical run, 400 gm of the phosphor are coated using a 32.degree. C. bubbler temperature, with 0.5 1/min nitrogen gas (the fluidizing gas medium) flowing through the TMOS bubbler, and with 0.6 1/min oxygen gas entering the fluidized powder bed (through the hollow stirrer rod) at a point a few centimeters above the level of the porous distributor plate.

Coating reactions were carried out for times ranging between 2.5 hours and 7.5 hours. Subsequently, the amounts of silica deposited were determined analytically for several of the coated phosphors. The results of these determinations are listed in Table 1. These data are plotted vs coating time in FIG. 3. As shown, the amount of silica deposited via the TMOS/O.sub.2 coating reaction increases linearly with increasing coating time.

TABLE 1______________________________________Results of Chemical Analyses for SilicaCoatings on TMOS/O.sub.2 -Coated Zn.sub.2 SiO.sub.4and Cool White Phosphors*Phosphor Coating Time (hr) Weight Percent SiO.sub.2______________________________________Zn.sub.2 SiO.sub.4 5 1.65Cool White 21/2 0.80Cool White 5 1.77Cool White 71/2 2.56______________________________________ *400 gm phosphors 0.5 l/min bubbler flow rate; 32.degree. C. bubbler temperature; 0.6 l/min O.sub.2 flow rate; inert carrier gas for TMOS: N.sub.2.

The silica-coated and uncoated zinc silicate and cool white phosphors were also examined via high resolution scanning electron microscopy. Photographic images were obtained at 20,000.times. and 50,000.times. magnification. There were no features observed in the photomicrographs obtained with the silica-coated materials that were not observed in the photomicrographs obtained with the corresponding uncoated phosphors. Thus, the silica coatings produced via the TMOS/O.sub.2 reaction appear to be uniform and conformal to the surfaces of the underlying phosphor particles.

The continuity of the silica coatings formed on the ZnSiO.sub.4 :Mn and cool-white phosphors were examined using X-ray photoelectron spectrometry. Typical normalized relative atomic concentration data obtained with an uncoated and a silica-coated ZnSiO.sub.4 :Mn phosphor are compared in Table 2. Typical data obtained with an uncoated and a silica-coated cool white phosphor are similarly compared in Table 3. As shown in Table 2, signal corresponding to zinc and manganese are completely absent from the XPS spectra obtained with the TMOS/O.sub.2 -coated zinc silicate phosphor. Similarly, except for a very small calcium signal, the XPS spectra obtained with the TMOS/O.sub.2 -coated cool white phosphor contain no evidence of the underlying phosphor. Thus, the silica coatings formed upon the surfaces of the phosphor particles appear to be continuous as well as conformal.

TABLE 2______________________________________Relative Atomic Concentrations of SurfaceElements from XPS Analyses of Uncoated andSilica-Coated Zn.sub.2 SiO.sub.4 :MnCoating Zn (3p) Si (2p) Mn (2p)______________________________________None 100 73 2.perspectiveto.2 w/o SiO.sub.2 0 100 0(from TMOS/O.sub.2reaction)______________________________________ TABLE 3______________________________________Relative Atomic Concentrations of SurfaceElements from XPS Analyses of Uncoated andSilica-Coated Cool White PhosphorCoating Ca (2p) F (1s) P (2s)______________________________________None 100 23 63.perspectiveto.2 w/o SiO.sub.2 <1 0 0(from TMOS/O.sub.2reaction)______________________________________

EXAMPLE 2

Three different lots of Zn.sub.2 SiO.sub.4 :Mn phosphor were coated with silica according to the method described in Example 1. Samples of each phosphor, before and after silica-coating, were annealed in air for 4 hours at 750.degree. C. Additional samples were likewise annealed in the air for one hour at 800.degree. C. Portions of each annealed material were pressed into so-called plaques (i.e., they were pressed into molds so that uniformly flat horizontal surfaces were obtained.) A spot brightness meter equipped with a green photo-optic filter along with an unfiltered mercury plasma ultraviolet light source was used to measure a so-called plaque brightness for each sample, expressed relative to that of a sample of each uncoated and unannealed phosphor. The results of these measurements are listed in Table 4. The relative plaque brightnesses measured with phosphor lot #1 are also plotted vs w/o silica added during the TMOS/O.sub.2 -coating process in FIG. 4.

TABLE 4______________________________________Relative Plaque Brightness of TMOS/O.sub.2 -Coated Zn.sub.2 SiO.sub.4:Mn* w/o SiO.sub.2 Anneal Relative PlaquePhosphor Lot Coating Conditions Brightness (%)______________________________________1 0 none 100.0 4 hr/750.degree. C. 95.6 11/2 hr/800.degree. C. 96.01 0.40 none 94.8 4 hr/750.degree. C. 99.3 11/2 hr/800.degree. C. 100.41 0.80 none 86.9 4 hr/750.degree. C. 99.3 11/2 hr/800.degree. C. 100.41 1.20 none 84.2 4 hr/750.degree. C. 99.3 11/2 hr/800.degree. C. 100.42 0 none 100.0 4 hr/750.degree. C. 95.9 11/2 hr/800.degree. C. 95.92 1.20 none 78.1 11/2 hr/800.degree. C. 100.33 0 none 100.0 4 hr/750.degree. C. 95.7 11/2 hr/800.degree. C. 95.93 1.20 none 80.0 11/2 hr/800.degree. C. 101.9______________________________________ *Bubbler Temp. = 32-33.degree. C.; Bubbler Flow Rate = 0.5 l/min; O.sub.2 Flow Rate = 0.6 l/min; Powder Weight = 400 gm; Coating Temp = 450-460.degree. C.

As shown, the brightness of the uncoated phosphor is lowered by at least 4% when annealed in the air at 750.degree. C. or at 800.degree. C. A body color (corresponding to a reduction in reflected visible light) also develops during the annealing of the uncoated phosphor. On the other hand, sizable reductions in brightness are also observed with the unannealed silica-coated phosphors. Moreover, the thicker the silica coating, the lower the measured plaque brightness. For instance, the plaque brightness measured with a zinc silicate phosphor coated with 1.20 w/o silica (via the TMOS/O.sub.2 reaction) is only around 80% of that measured with the uncoated phosphor.

In contrast, plaque brightnesses nearly equal to or exceeding those measured with the uncoated and unannealed phosphor are obtained with silica-coated phosphors that have been annealed in the air at temperatures between 750.degree. C. and 800.degree. C. Such silica-coated and annealed materials are also notable for an absence of the body-color that develops during the annealing of the uncoated phosphor. Therefore, whereas the uncoated phosphor cannot be air-annealed without suffering a 4%-5% reduction in plaque brightness as well as a reduction in reflected visible light, the brightness of a silica-coated phosphor actually increases with increasing annealing temperature to a level exceeding that measured with the uncoated and unannealed phosphor, itself.

EXAMPLE 3

The implication of the data shown in Example 2 is that detrimental interactions which normally occur between the phosphor and the air during the annealing step are prevented when the phosphor is coated with a thin layer of silica. The body-color that develops during the annealing of the uncoated Zn.sub.2 SiO.sub.4 :Mn phosphor suggests that the Mn.sup.2+ ions located on or near to the phosphor particle surface are oxidized during the anneal. However, it is possible that the body color is due to oxidized tungsten at the phosphor particle surface. The absence of this undesirable body-color, the undiminished brightnesses obtained with annealed silica-coated Zn.sub.2 SiO.sub.4 :Mn phosphors, and the observed continuity and conformality of the silica coatings themselves indicate that the phosphor surface is stabilized by the presence of the coating, thereby preventing the surface manganese from interacting with the oxidizing atmosphere within the annealing furnace.

That the phosphor surface is stabilized by the presence of the silica coating may also be shown by examining the coated phosphor using x-ray photoelectron spectrometry. Listed in Table 5 are the normalized relative atomic concentrations of Al, Zn, Si, and Mn obtained with several samples from the measured XPS signal intensities corresponding to the Al(2p), Zn(3p), Si(2p), and Mn(2p) electrons, respectively. Sample 1 is a Zn.sub.2 SiO.sub.4 :Mn phosphor coated with alumina according to the method described in U.S. patent application Ser. No. 07/432,941 using aluminum isopropoxide (AIP) as the organometallic coating precursor. Sample 3 was obtained by coating the same phosphor with silica as described in Example 1 (using TMOS as the coating precursor). Samples 2 and 4 were obtained by air-annealing samples 1 and 3, respectively, for 4 hours at 750.degree. C.

TABLE 5______________________________________Relative Atomic Concentrations for XPS AnalysesAIP/O.sub.2 -Coated, TMOS/O.sub.2 -Coated, and TMOS/O.sub.2AIP/O.sub.2 -Coated Zn.sub.2 SiO.sub.4 :MnSampleNo. Material Al (2p) Zn (3p) Si (2p) Mn (2p)______________________________________1 AIP/O.sub.2 -Coated.sup.a 100 0 0 0 Zn.sub.2 SiO.sub.42 Sample 1, 100 13 0 3 annealed 4 hr 750.degree. C.3 TMOS/O.sub.2 - 0 0 100 0 Coated.sup.b Zn.sub.2 SiO.sub.44 Sample 3, 0 0 100 0 annealed 4 hr 750.degree. C.5 Sample 4, AIP- 100 0 0 0 O.sub.2 -.sup.a,b Coated6 Sample 5, 100 0 0 0 annealed 4 hr 750.degree. C.______________________________________ .sup.a .perspectiveto.2% Al.sub.2 O.sub.3 Coating .sup.b .perspectiveto.2% SiO.sub.2 Coating

As shown in Table 5, none of the cations present in the Zn.sub.2 SiO.sub.4 :Mn phosphor are detected in the XPS spectra obtained with the alumina-coated phosphor. This indicates that the AIP/O.sub.2 coating is continuous and thick enough to filter any Zn(3p), Si(2p), or Mn(2p) electrons that might be emitted under x-ray bombardment. In contrast, relatively large Zn(3p) and Mn(2p) signals are detected after annealing the alumina-coated phosphor for 4 hours at 750.degree. C. (sample 2). These results are interpreted to indicate that these cations are mobile enough to migrate through the alumina coating during the anneal. In sharp contrast are the XPS date obtained with the TMOS/O.sub.2 -coated samples. In this case, silicon is the only cationic species detected either before or after the 4 hour 750.degree. C. anneal, indicating that the zinc and manganese ions present on the surface of the Zn.sub.2 SiO.sub.4 :Mn phosphor do not migrate through the silica coating during the anneal. The fact that the alumina-coated phosphor possesses a distinct body-color after the anneal (thought to be due to oxidized manganese), whereas the annealed silica-coated phosphor does not possess such a body-color can be understood from these data.

Finally, consider the results obtained using the Zn.sub.2 SiO.sub.4 :Mn phosphor that has been silica-coated via the TMOS/O.sub.2 reaction and subsequently air-annealed for 4 hours at 750.degree. C. (sample 4). A quantity of this silica-coated and annealed phosphor was coated with alumina via the AIP/O.sub.2 reaction as described in U.S. patent application Ser. No. 07/432,941. As shown in Table 5 (sample 5), Al is the only cationic species detected via XPS analysis of this material, indicating that the alumina coating is continuous and uniformly thick enough to prevent the detection of any Si(2p) electrons that might be generated under x-ray bombardment. Most significant is the fact that an identical result is obtained after annealing the double-coated phosphor for 4 hours at 750.degree. C. In contrast to the results obtained with sample 2 (in the absence of the silica diffusion barrier), the absence of XPS signals indicating the presence of Zn, Si, or Mn near to the surface of the annealed double-coated material and the complete absence of any detectable body-color indicates that the silica coating prevents the interaction between the phosphor and the alumina coating that would otherwise occur.

This conclusion is reinforced by the relative plaque brightness data listed in Table 6. Shown are the measured brightnesses (relative to that of the uncoated and unannealed phosphor) of AIP/O.sub.2 -coated Zn.sub.2 SiO.sub.4 :Mn before and after a 4 hour 750.degree. C. anneal, both with and without a TMOS/O.sub.2 (silica) diffusion barrier. As indicated in the table, the reduction in brightness observed with the unannealed phosphor in the absence of the diffusion barrier is more than twice that obtained with the double-coated phosphor. More significantly, the reduction in brightness observed with the annealed alumina-coated phosphor in the absence of the diffusion barrier is an order of magnitude greater than that obtained when the alumina coating was applied over the diffusion barrier. Thus, the plaque brightness measured with the double-coated and annealed phosphor was only about 1% below that measured with the virgin phosphor.

TABLE 6______________________________________Relative Plaque Brightnesses of AIP/O.sub.2 -Coated Zn.sub.2 SiO.sub.4:Mnwith and without underlying SiO.sub.2 Diffusion Barrier.sup.1w/o SiO.sub.2.sup.2 Anneal Plaque Brightness Relative.sup.3Coating Conditions to that of Uncoated Phosphor______________________________________0 none 94.2% 4 hr/750.degree. C. 87.7%1.20 none 97.3% 4 hr/750.degree. C. 98.8%______________________________________ .sup.1 Ca. 2 w/o Al.sub.2 O.sub.3 Coating .sup.2 SiO.sub.2coated phosphor was annealed 4 hours at 750.degree. C. prior to coating with alumina .sup.3 Plaque brightness relative to that measured with the uncoated phosphor.

Due to the results obtained from Examples 1-3 it was thought that the results of improved plaque brightness and lumen maintenance could be extended to fluorescent lamps. However, it is known that correlation between handlamp plaque brightness and fluorescent lamp performance for a given phosphor frequently do not exist. This results from a multitude of factors including changes in the phosphor which occur during lamp baking, lamp fabrication and the contact of the phosphor with the mercury discharge. Moreover, the mercury discharge within a fluorescent lamp contains about 15% of its emission at 185 nm. This short-wavelength emission can lead to enhanced brightness and/or damage to the phosphor which can influence the observed initial brightness and maintenance.

More specifically, in the case of the alumina coated and annealed willemite phosphor, the fluorescent lamps which have the highest lumen performance possess a plaque brightness of 92% of the virgin phosphor. However, the corresponding fluorescent lamp performance may only be a few percent lower. Further, while the handlamp photoluminescent performance is improved with a silica coated and annealed phosphor, this material exhibits a significant loss in brightness and catastrophic maintenance loss within a fluorescent lamp. This behavior is probably associated with the reaction of the phosphor with the mercury discharge within the fluorescent lamp.

Shown in FIG. 5 is a cross-sectional view of a phosphor particle coated with a bi-layer. The phosphor grain is coated with a silica layer which prevents diffusion from the phosphor grain to the surface coating of alumina. It is also believed that the silica layer prevents diffusion of the alumina layer to the phosphor grain. Potential uses of the phosphor shown in FIG. 5 are discussed below.

Historically silica and silica-containing phosphors used within fluorescent lamps are known to give rise to appreciable maintenance loss, the uncoated willemite, in fact, being a prime example of this. Therefore, any improvement due to a protective layer over a silica coating can be expected to be strongly dependent on the quality and conformality of that layer as well as the intrinsic resistance of the specific phosphor to degradation in a fluorescent lamp.

EXAMPLE 4

Silica coatings were applied to the surfaces of zinc silicate phosphors by CVD in a fluidized bed described in Example 1. However, the tests were carried out using typically 1500 gms. of phosphor in an 80 mm ID quartz tube which employed a quartz frit as the distributor. Aluminum Oxide C was blended with the phosphor at a concentration of 0.1% by weight of the fluidized bed reactor was maintained between 450.degree. and 460.degree. C. during the coating process. This temperature was monitored by a thermocouple placed within the bed located at the midbed height. In a typical run 2 liters per minute were run through a bubbler containing tetramethoxyorthosilicate (TMOS) liquid maintained at 32.degree. C., and 3 liters per minute of undiluted oxygen entered the bed through a hollow stirrer rod which was located a few centimeters above the level of the porous distributor plate. Coating reactions were carried out between 11/2 hrs. and 5 hrs. to effect the deposition of predetermined amounts of silica coating. Table 7 summarizes the powder properties of the virgin phosphors used in the following examples.

TABLE 7______________________________________Particle Properties of the Virgin Zinc SilicatePhosphors Used in the Examples Cited FSS Coulter Counter Surface Area (Fisher Sieve Particle Size (sonic).sup.2Lot No. (M.sup.2 /gm).sup.1 Size, microns) 50% Q.D..sup.3______________________________________66RMF 0.38 6.7 9.0 0.23TK1-2M 0.44 6.1 7.72 0.26TK2-U 0.52 5.1 6.65 0.27______________________________________ .sup.1 Determined by singlepoint BET measurements, using a Quantachrome Monosorb surface area instrument. .sup.2 Based on volume. .sup.3 Q.D. is defined as (d.sub.75% - d.sub.25%)/(d.sub.75% + d.sub.25%) and is a relative measure of breadth in the particle distribution.

Following coating, the phosphors were annealed in quartz boats for about 4 hours at an annealing temperature of approximately 760.degree. C. At this point phosphors were coated with alumina as described below.

The CVD coatings were carried out on the annealed phosphors described above in a fluidized bed using trimethyl aluminum (TMA) and oxygen as precursors. The equipment and procedures for fluidized bed coating of willemite phosphors are described in detail in U.S. patent application Ser. No. 07/406,884. Briefly, a blend of approximately 1000 to 1300 gms. and 0.1% of Aluminum Oxide C by weight of phosphor was loaded into a quartz fluidized bed column comprising an 80 mm ID quartz tube having a quartz frit fused to the bottom which acted as the distributor plate. A 65 mm stainless steel agitator disc was attached to a vibromixer agitator. Approximately 5 cm from the base, a two-micron stainless steel filter element was welded in line and functioned as a diffuser of the oxygen mixture. The agitator disc itself was located approximately 25 mm above the quartz distributor. A thermocouple, located at the midbed height within the fluidized bed, was used to monitor the temperature of the bed, which was maintained between approximately 420.degree. and 450.degree. C.

The apparatus used for carrying out the coating reactions is shown in FIG. 1 with some slight alterations. In a typical run, which lasted between 3 and 5 hours, 1750 cc per minute of nitrogen passed through a bubbler 12 containing trimethyl aluminum (TMA) liquid which was maintained at 30.degree. C. Another 1250 cc per minute of nitrogen carrier gas was used to dilute the flow through line 111. The combined flow was used to fluidize the phosphor particles in reaction vessel 15. The oxygen as an oxygen/inert gas mixture was introduced through line 21 at 2500 cc per minute of oxygen and 50 cc per minute of nitrogen into the fluidized bed through the two-micron filter element described earlier. Table 8 summarizes specific coating parameters used for the examples cited.

TABLE 8__________________________________________________________________________Coating parameters used to apply Al.sub.2 O.sub.3 and SiO.sub.2 layers tothe zinc silicate phosphors described in this disclosure.__________________________________________________________________________No. Sample No. Run No. Bed Loading Bubbler O.sub.2 Coating Time Run No. Bed Loading__________________________________________________________________________6RMFF #385 No silica CWM120-89 2000 gms #441 TMOS-29L 1500 gms 2 l/min 3 l/min 31/2 hrs. CWM330-89 1230 gmsk1-2M #518 No silica CWM1017-89 1500 gms #479 TMOS-38L 1500 gms 2 l/min 3 l/min 11/2 hrs. CWM614-89 1295 gms #476 TMOS-37L 1500 gms 2 l/min 3 l/min 5 hrs. CWM612-89 1070 gmsk2-U #425 No silica CWM313-89 350 gms+ #443 TMOS-30L 1500 gms 2 l/min 3 l/min 5 hrs. CWM403-89 1120 gms__________________________________________________________________________ SiO.sub.2 Al.sub.2 O.sub.3No. Sample No. Bubbler Carrier O.sub.2 Coating Time Wt % Thickness.sup.# Wt % Thickness.sup.#__________________________________________________________________________6RMFF #385 1750 cc 1250 cc 2500 cc 5 hrs. 2.28% 150 .ANG. #441 1750 cc 1250 cc 2500 cc 3 hrs. 1.26% 144 .ANG. 2.22% 147 .ANG.k1-2M #518 1750 cc 1250 cc 2500 cc 41/3 hrs. 150 .ANG. #479 1750 cc 1250 cc 2500 cc 43/4 hrs. 0.81% 80 .ANG. 2.65% 150 .ANG. #476 1750 cc 1250 cc 2500 cc 31/2 hrs. 1.80% 178 .ANG. 2.62% 150 .ANG.k2-U #425 400 cc 350 cc 500 cc 4 hrs. 115 .ANG. #443 1750 cc 1250 cc 2500 cc 3-5/6 hrs. 1.80% 150 .ANG. 3.12% 150 .ANG.__________________________________________________________________________ + This run was carried out in a 40 mm I.D. quartz column; earlier data ha shown equivalence between the small (40 mm) and large (80 mm) column runs Also, no performance differences have been seen between 100-300 .ANG. Al.sub.2 O.sub.3only coated willemite phosphors. .sup.# Values for the thickness are derived from the weight percent SiO.sub.2 or Al.sub.2 O.sub.3 deposited and an assumed density of 2.3 gms/cc and 3.97 gms/cc, respectively.

Once the phosphor particles were coated, they were transferred to quartz boats and annealed at approximately 760.degree. C. for 4 hours.

Lamp Testing

After the annealing step described above, the phosphors were coated in 20WT12 or 30 WT12 fluorescent lamps using conventional water base suspension systems. The lamps thus coated were processed into finished fluorescent lamps and photometrically evaluated. The data were then converted to corresponding 40WT12 data using established correction factors.

In the tests cited below, the bilayer coated and annealed phosphors were tested against Sylvania Type 2293, which is a Ce,Tb, Mg Aluminate phosphor used as a green-emitting component in high color rendition lamps. In addition, the uncoated virgin phosphor and the singly coated and annealed (i.e., alumina only) phosphors were also evaluated with the test.

Table 9 lists the lifetest results of the phosphors. It is clearly evident that the bilayer coating provides 0-hr. brightness values which are either equivalent to or exceeds that of the virgin phosphor lot. Also, both the single- and bilayer-coated phosphors possess maintenance which significantly exceeds the virgin phosphor. However, the bilayer (Al.sub.2 O.sub.3 /SiO.sub.2) annealed phosphor substantially exceeds the brightness of the Al.sub.2 O.sub.3 -only phosphor by 8% for the `RMF` phosphor (66RMF) to as much as 17% for the singly fired willemite phosphors (TK1-2M and TK2-U lots). Note that the brightness observed for the alumina-only coated and annealed singly fired willemite phosphor lots are so low as to eliminate their consideration for commercial use in conventional high color rendition lamps, whereas the substantially higher brightness of the Al.sub.2 O.sub.3 /SiO.sub.2 coated singly fired phosphor allows its use in triblend lamps. The thickness of the silica coatings are derived from the information shown in Table 8.

TABLE 9______________________________________Lifetest Data for Various Willemite Phosphor______________________________________ % M % M66RMF Lot 0 hr 100 hr (0-100) 500 hr (0-500)______________________________________Virgin 5329 3800 71.3 Discontinued#385 Al.sub.2 O.sub.3 4860 4763 98.0 4685 96.4#441 Al.sub.2 O.sub.3 /SiO.sub.2 5245 5101 97.3 5048 96.2FHX164 Type 2293 4923 4768 96.8 4685 95.2______________________________________ % M % MTK1-2M Lot 0 hr 100 hr (0-100) 500 hr (0-500)______________________________________Virgin 5199 -- -- -- --#578+ Al.sub.2 O.sub.3 4491 4424 98.5 -- --#479 Al.sub.2 O.sub.3 / 5219 5193 99.5 4887 93.6SiO.sub.2 (80 .ANG.)#476 Al.sub.2 O.sub.3 / 5239 5137 98.0 4894 93.4SiO.sub.2 (178 .ANG.)FHX343 Type 2293 4873 4757 97.6 4499 92.3______________________________________ + This sample was run in a separate test against Type 2293 Lot FXH343. In that test FHX343 gave 0 hrs = 4903 l, 100 hrs = 4736 l, (0-100% M = 96.6%). Therefore, the major conclusions put forth are not affected. % M % MTK2-U Lot 0 hr 100 hr (0-100) 500 hr (0-500)______________________________________Virgin 5021 -- -- -- --#425 Al.sub.2 O.sub.3 4499 4332 96.3 4264 84.7#443 Al.sub.2 O.sub.3 /SiO.sub.2 5084 4878 95.9 4842 95.2FHX164 Type 2293 4923 4768 96.8 4685 95.2______________________________________ Note that values given have been corrected from corresponding 20WT12 and 30WT12 to 40WT12 fluorescent lamp data using established correction factors. Test Samples for lots 66RMF and TK2U were evaluated in 20WT12 lamps; Lot Tk12M was evaluated in 30WT12 lamps.

As mentioned earlier, the singly fired phosphor can be manufactured on a production scale with much higher yields than can be obtained using the `RMF` synthesis described by Kasenga in U.S. patent application Ser. No. 07/434,741 which is still pending. The much higher yields and efficiencies of scale favor substantial cost savings.

It is important to note that the addition of the silica interface provides a major improvement in the 100-hr. brightness, as well, over the alumina coatings alone. As the data in Table 9 show, Sample #479 has yielded a 100-hour brightness of 5193 lumens. This corresponds to the highest value ever achieved for a willemite phosphor after 100 hours of burning corrected to equivalent 40WT12 fluorescent lamp performance. This has significant commercial implications since published ratings of `initial brightness` are actually those determined after 100 hours of lamp operation.

When tricomponent blend suspensions are used which consist of Y.sub.2 O.sub.3 :Eu for the red emission, Ba,Mg Aluminate:Eu for the blue emission, and the `RMF` coated and annealed phosphor described by U.S. application Ser. No. 07/434,741 for the green emission, a color variation in the emitted light is observed from the fluorescent lamp across its length manifested by a slightly red coloration on the more thinly coated end of the lamp and slightly green coloration on the more thickly coated end of the fluorescent lamp. This end-to-end color variation is believed to be due, in part, to the disparity in particle size between the red and green phosphors in the suspension used to make the coated lamps. In fact, lamp fabrication using Sample #443, a bilayer coated and annealed TK2-U lot of smaller particle size than the `RMF` phosphor, showed a significant reduction in color nonuniformity compared to a coated and annealed 66RMF phosphor (Sample #441) run in the same test. Thus, the bilayer coating allows particle size reduction while maintaining excellent brightness for high color rendition triblend applications.

Table 10 lists the x-ray photoelectron spectroscopy (XPS) analyses of the surfaces after each of the stages of processing leading to the alumina only and alumina/silica coated and annealed phosphors. The data clearly show that the silica layer either eliminates or substantially reduces the migration of Zn and Mn through the alumina coating. It is also expected, although not experimentally verified, that the silica interfacial layer also prevents the migration of Al ions from the alumina coating into the zinc silicate phosphor. Both factors contribute to the elimination of undesirable light-absorbing body color of the annealed phosphor and the interference from impurities that lead to the loss in generation of efficient luminescence.

TABLE 10______________________________________X-ray Photoelectron Spectroscopy Analyses ofCoated Willemite Phosphors Atomic PercentSample Designation Sl Al O Zn Mn C______________________________________66 RMF Virgin Powder 15.8 -- 54.6 22.3 0.7 6.6CWM 120-89 (TMA -- 43.4 51.6 <0.1 -- 4.9coated)#385 (TMA coated/ -- 42.0 52.5 4.3 0.2 1.1annealed)TMOS 29L (Silica 38.5 -- 58.8 1.1 -- 1.4coated)#439 (Silica coated/ 39.0 -- 58.2 -- -- 2.8annealed)CWM 330-89 (TMA -- 44.1 51.6 -- 4.3coated #439)#441 (CWM 330-89/ -- 45.2 51.3 0.2 -- 3.2annealed)TK-2U Virgin phosphor 16.0 -- 48.6 17.7 0.9 16.0CWM 313-89 (TMA -- 43.7 52.8 -- -- 3.5coated)#425 (TMA coated/ -- 36.9 45.8 10.1 0.4 6.6annealed)TMOS 30L (Silica 36.4 -- 56.1 0.2 -- 7.3coated)#440 (Silica coated/ 37.2 -- 53.9 <0.1 -- 8.8annealed)CWM 403-89 (TMA -- 43.6 52.8 -- -- 3.6coated #440)#443 (CWM 403-89 -- 45.4 50.4 <0.1 -- 4.2annealed)TK1-2M Virgin 13.3 -- 44.1 27.7 0.5 14.3phosphorTMOS 38L (Silica 38.1 -- 60.4 <0.1 -- 1.5coated)#477 (Silica coated/ 37.7 -- 59.8 <0.1 -- 2.4anneal)CWM 614-89 (TMA -- 42.5 53.1 -- -- 4.4coated #477)#479 (CWM 614-89 -- 45.1 51.8 <0.1 -- 3.0Annealed)______________________________________

Lamp tests carried out employing the single-coated alumina `RMF` zinc silicate phosphor produced by the Chemical and Metallurgical Division of GTE Products Corporation, Towanda, Pa., have shown a lower brightness level in tri-phosphor blends when compared with the rare-earth-containing blends run in the same test (See Table 11). A bilayer (Al.sub.2 O.sub.3 /SiO.sub.2) coated willemite described employing a single-step fired willemite base material in 40WT12 fluorescent lamps (Sample #541) has yielded performance of 5280 lumens at 0 hours of lamp operation and 5144 lumens at 100 hours with a maintenance of 97.4% which exceeds that of the Type 2293 by almost 1%. This performance substantially exceeds that obtained with the best alumina-coated "RMF" materials heretofore available. (See Table 11 for single-component life test data).

Thus, it is probable that this substantially less expensive silica/alumina coated single-step fired willemite phosphor is useful as single components and as components of blends in double twin-tube lamps. This phosphor will replace the more expensive Sylvania Type 2293 and the Nichia LaPO.sub.4 :Ce, Tb phosphors presently used. Further uses for the silica/alumina coated phosphors are discussed below. TABLE 11__________________________________________________________________________Lifetest data for lamps containing alumina-coated `RMF`willemite and rare-earth green tri-phosphor blends.*__________________________________________________________________________Single-Coat Lamps: 20WT12, 3500.degree. K LumensGreen Test Sample P.Wt. (gms) X Y 0 Hr 100 Hr (0-100 Hr) % M__________________________________________________________________________Control 2293 1.95 0.410 0.398 1433 1389 96.9G077 (Willemite) 1.85 0.408 0.397 1354 1308 96.6Delta Lumens -79 -81Delta Percent -5.5% -5.8%__________________________________________________________________________Double-Coat Lamps: 20WT12, D35 (Designer, 3500.degree. K) Second Coat LumensGreen Test Sample P.Wt. (gms) X Y 0 Hr 100 Hr (0-100 Hr) % M__________________________________________________________________________Control 2293 0.52 0.412 0.400 1353 1339 99.0G077 (Willemite) 0.55 0.412 0.398 1314 1291 98.2Delta Lumens -39 -48Delta Percent -2.9% -3.6%__________________________________________________________________________ *Lifetest data for single components in 40WT12 fluorescent lamps are: 0 Hr 100 Hr (0-100 Hr) % M__________________________________________________________________________Control 2293 4921 4740 96.3G077 (60 RMF) 5019 4695 93.5Virgin 60RMF (uncoated) 5134 4143 80.7__________________________________________________________________________

Referring to FIGS. 6 and 7, there is shown in FIG. 6 and arc discharge lamp of the fluorescent type. The lamp 10 is comprised of an elongated glass tube 12 of circular cross-section. It has the usual electrodes 14 and 16 at each end supported by lead-in wires, 18, 20 and 22, 24, respectively, which extend through glass presses 26, 28 in mount stems 30, 32 to the contacts in bases 34, 36 affixed to the ends of the lamp.

The sealed tube is filled with an inert gas such as Argon or a mixture of Argon and Krypton at a low pressure, for example 2 torr, and a small quantity of mercury, at least enough to provide a low vapor pressure during operation.

The interior of tube 12 is coated with a first layer of phosphor 38 such as, for example, a calcium halophosphate activated by antimony and manganese.

A phosphor coating suspension was prepared by dispersing the phosphor particles in a water-based system employing polyethylene oxide and hydroxyethyl cellulose as the binders with water as the solvent.

The phosphor suspension was applied in the usual manner of causing the suspension to flow down the inner surface of the bulb and allowing the water to evaporate leaving the binder and phosphor particles adhered to the bulb wall.

The first layer 38 was then dried prior to overcoating with a second phosphor layer 40 comprised of narrow-band red- and blue- emitting phosphors and a broad-band green-emitting phosphor. These two narrow-band phosphors can be, for example, a yttrium oxide activated by trivalent europium and having a peak emission at 611 nm; and barium magnesium aluminate activated by divalent europium and having a peak emission at 455 nm. The broader band phosphor was alumina/ silica-coated zinc silicate activated by manganese and having a peak emission at 528 nm.

The second phosphor layer containing the CVD-coated phosphor is applied from a water-based suspension by allowing the coating to flow down over the first phosphor layer 38 until the phosphor coating drained from the bottom of the bulb indicating the coverage of the phosphor layer 38 was complete. Lamps made by this method usually exhibit the thinnest coating thickness on the top end of the bulb and the heaviest thickness on the bottom end, where the suspension is allowed to drain. The double-coated bulbs were then baked and processed into fluorescent lamps by conventional techniques.

In the case where only a single layer lamp was made, the methods were essentially the same as described herein with the exception that the halophosphate layer was not applied.

Control lamps were fabricated by identical techniques as described above but had a narrow.TM.band green-emitting magnesium aluminate phosphor activated by cerium and terbium in the second phosphor layer with a peak emission at 545 nm. This phosphor is generally used in the tri-phosphor blend but was replaced by the green CVD-coated willemite phosphor in this invention.

Lamps employing a representative alumina-only coated and annealed `RMF` willemite phosphor, manufactured by the Chemical and Metallurgical Division of PMG, Towanda, Pa., were also fabricated and incorporated into the testing.

Lifetest data and particle size information for the single component willemite phosphors used for the tri-phosphor blend lamps are given in Tables 12 and 13, respectively. These values have been corrected to obtain the performance levels that would be observed with 40T12 fluorescent lamps using established correction factors.

TABLE 12______________________________________Lifetest data for single-component fluorescent lampscontaining coated willemite phosphors (corrected to 40WT12) 0 Hr 100 Hr 500 Hr % M (0-500 Hr)______________________________________Type 2293 (control) 4923 4768 4685 95.2#425 (TK2-U)+ 4499 4332 4264 94.7#443 (TK2-U)++ 5084 4878 4842 95.2G103 (61RMF)+ 5085 4804 4677 92.0#441 (66RMF)++ 5245 5101 5048 96.2______________________________________ + Phosphor grains coated with only a single coating of alumina, (sample G103 was prepared at Chem. and Met. Div. PMG, Towanda, PA). ++ Phosphor grains coated with both silica and alumina. TABLE 13______________________________________Particle properties of the virgin zinc silicatephosphors used in the examples cited. Coulter Counter FSS Particle Size (sonic)Lot Number (Fisher Sieve Size, microns) 50% Q.D.______________________________________Singly FiredTK-U 5.1 6.7 0.27`RMF`60 RMF 7.4 9.3 0.2161 RMF 8.0 10.0 0.2166 RMF 6.7 9.0 0.23______________________________________

For the evaluation of the tri-blends containing the alumina/silica-coated willemite phosphor, two different lamp types were used, and compared with controls which did not have this phosphor in the blend. The lamps were tested by photometering for light output in a standard photometric sphere, both initially and at stated times. In the following tables, light outputs are expressed in lumens. Lamp color values were obtained by spectral power distribution (SPD) measurements.

EXAMPLE 5

This example compares the light output and maintenance of alumina/silica-coated and annealed willemite phosphors in the tri-phosphor blends for double layered 96-inch Designer 3000.degree. K. High Output Super Saver fluorescent lamps. This lamp type has a tri-phosphor weight in the second layer of about 15% of the total phosphor in the two layers. The lamps were fabricated to obtain the same x and y color coordinates for both the test and control by adjusting the tri-phosphor blend composition. The lamp test results are listed in Table 14. As the test data show, the lumens and lumen maintenance of the test group are equivalent to the rare-earth-containing Type 2293 control. This is in contrast to previous results obtained with alumina-only (single-coated and annealed) `RMF` phosphor where Designer lamp brightness values were typically 3% to 4% lower than those obtained with the standard tri-phosphor blend (see Table 11 for comparison).

TABLE 14__________________________________________________________________________Lifetest data for double-coat lamps where the triblends containrare-earthgreen-emitting phosphor and two Al.sub.2 O.sub.3 /SiO.sub.2 -coatedwillemite phosphors.__________________________________________________________________________Lamp Type: 96T12/D30/HO/SS (Designer, 3000.degree. K, high output,supersaver)Green Composition-Second LayerComponent Tri-Phosphor (Weight %).sup.+ (1) Tri-Phosphor Color CoordinatesTest Sample Phosphor Red (Y.sub.2 O.sub.3 :Eu) Blue (Ba,MgAlO:Eu) Green Weight (Grams) X Y__________________________________________________________________________Control 2293 (Ce,Tb)MgAlO:Ce,Tb 65.8 2.9 31.3 1.84 0.429 0.410#441 (66RMF) Zn.sub.2 SiO.sub.4 :Mn,W `RMF`* 66.7 2.5 30.8 1.90 0.435 0.405#443 (TK2-U) Zn.sub.2 SiO.sub.4 :Mn, 66.1 2.5 31.4 1.92 0.434 0.403 W Single Fired__________________________________________________________________________ .sup.+ First layer is a holophosphate phosphor Sylvania Type 4300 with color coordinates in finished lamp of X = 0.440, Y = 0.405; first layer weight = 11-12 gms. *Alumina/silica coated Photo Metric Results Lumens Lumens Lumens Lumen Maintenance % 0 Hour 100 Hours 500 Hours 0-500 Hours__________________________________________________________________________Control 2293 8753 8573 8060 92.1#441 8763 8430 8082 92.2#443 8765 8368 8063 92.0__________________________________________________________________________ (1) Red component is Sylvania Type 2345 (Y.sub.2 O.sub.3 :Eu); Blue component is Sylvania Type 246 (Ba,MgAlO:Eu)

EXAMPLE 6

As mentioned earlier, color nonuniformities of fluorescent lamps containing the `RMF` coated and annealed phosphor occur and become more severe as the bulb length increases. The 96-inch lamps have shown the most pronounced variation. Color nonuniformities strongly influence consumer acceptance of the lamp since they are a premium-priced product designed for use in high color rendition applications.

FIGS. 8 through 10 show the color points taken from the more heavily coated end (HE), middle (M), and more lightly coated end (LE) of the fluorescent lamp containing the rare earth green, the bilayer-coated `RMF` phosphor (#441), and the bilayer-coated single-fired willemite phosphor (#443), respectively. Also shown are the two- and three-step Macadam ellipses. The Macadam ellipse is a way of assessing differences in visual color perception. For acceptable lamps it is desirable to have all points residing well within the two-step ellipse. As shown in FIG. 8, the end-to-end color variation is excellent in the case of the rare-earth green Type 2293. However, the `RMF`-based phosphor shows a red and green variation along the length of the lamp, as shown in FIG. 9. In contrast, as shown in FIG. 10, the singly fired phosphor (Sample #443) shows a significant improvement in color uniformity over the `RMF` phosphor, well within acceptable limits. It is thought that the color non-uniformity originates from a particle-size disparity between the red and green phosphors that make up the dispersion from which the lamps are made. Thus, the bilayer coating allows particle size reduction of the green component while maintaining excellent briqhtness for high color rendition applications. (As shown in Table 12, the fluorescent lamp brightness obtained with the alumina/silica-coated small particle phosphor (Sample #443) is much superior to that obtained with the same base phosphor coated only with alumina (Sample #425)).

EXAMPLE 7

This example compares the light output and maintenance of alumina-coated and alumina/silica-coated willemite phosphors in 40WT12 4100.degree. K. single-coat tri-phosphor lamps. This test is designed to exaggerate differences in performances between the green components used in the triblends, since the single layer of high color temperature requires the largest amount of green compared to any other lamp that would be fabricated. That is, any differences in performance will diminish as the triblend layer thickness is reduced (in double layer lamps) and as the lamp color temperature is reduced (since the fraction of green component in the blend goes down as the color temperature is reduced).

As the data in Table 15 show, the maintenance of all materials tested are comparable. Further, the color rendering index, measured after 100 hours of lamp operation, is about 3 units higher for the willemite-containing blends compared to the rare-earth-containing blend, even in this single-coat lamp, achieving CRI values in excess of 85.

TABLE 15__________________________________________________________________________Lifetest data for triblends containing rare-earth green-emitting greenphosphor,alumina-coated willemites, and Al.sub.2 O.sub.3 /SiO.sub.2 -coatedwillemite phosphors.__________________________________________________________________________Lamp Type: 40T12/4100.degree. K/single-coat triphosphor (no halo layer)Green Composition-Second Layer Tri-Phosphor CRIComponent Tri-Phosphor (Weight %)* Weight Color Coordinates 100Test Sample Phosphor Red (Y.sub.2 O.sub.3 :Eu) Blue (Ba,MgAlO:Eu) Green (Grams) X Y Hours__________________________________________________________________________Control 2293 (Ce,Tb)MgAlO:Ce,Tb 48.0 10.3 41.6 3.31 0.376 0.385 82.5G103 Zn.sub.2 SiO.sub.4 :Mn,W RMF 51.4 11.4 37.2 3.85 0.376 0.385 85.6(61RMF)** Al.sub.2 O.sub.3 only#441 Zn.sub.2 SiO.sub.4 :Mn,W RMF 54.8 11.6 33.6 4.07 0.380 0.389 85.3(66RMF)***#443 Zn.sub.2 SiO.sub.4 :Mn,W Single 53.1 12.6 34.3 3.74 0.378 0.388 85.4(TK2-U)*** Fired__________________________________________________________________________ *Red component is Sylvania Type 2345 (Y.sub.2 O.sub.3 :Eu); Blue componen is Sylvania Type 246 (Ba,MgAlO:Eu) **Manufactured at Chemical & Metallurgical Division, PMG, Towanda, PA (Aluminacoated willemite phosphor) ***Alumina/Silicacoated willemite phosphor. Photo Metric Results Lumens Lumens Lumens Lumens Lumen Maintenance % 0 Hour 100 Hours 500 Hours 1000 Hours 0-1000 Hours__________________________________________________________________________Control 2293 3427 3319 3252 3161 92.2G103 (61RMF) 3341 3183 3080 3006 90.0#441 (66RMF) 3421 3291 3215 3176 92.8#443 (TK2-U) 3364 3233 3147 3095 92.0__________________________________________________________________________

With regard to brightness, the bilayer-coated `RMF` phosphor (#441) is clearly superior to the representative singly coated `RMF` phosphor (G103). The brightness performance of the bilayer-coated singly fired zinc silicate (#443), while exceeding that of the alumina-coated `RMF` phosphor (G103), is about 2% below that of the rare-earth Type 2293 control, at 1000 hours of lamp operation. However, by way of comparison a singly fired (smaller particle) willemite phosphor with only the alumina coating (i.e., without the intervening layer of silica) is completely unsuitable for use in triblend applications because its brightness is about 10% below that of the corresponding alumina/silica-coated phosphor (i e., less than 4500 lumens initial brightness in a 40WT12 fluorescent lamp), as shown in Table 12.

The use of the bilayer-coated phosphor provides a substantial cost savings over the rare-earth blends in the examples cited above since the tri-phosphor blend represents the major cost of the lamp and, as Table 14 shows, the green component comprises over 30% of the tri-phosphor blend.

Further, the use of the singly fired bilayer-coated phosphor provides even further cost savings, since the large-particle `RMF` material with narrow particle size distribution (described in U.S. patent application Ser. No. 07/434,741) requires two firings in its synthesis with multiple decantations to remove the "fines" fraction. This necessarily results in low yields (typically 60%). However, the smaller particle material can be made, as described by Chenot, using a single-step firing followed by a washing which provides much higher yields (typically 90%). Also, the smaller particle material can easily be manufactured on production-scale equipment. The much higher yields and the efficiencies of scale all favor substantial cost savings which far outweigh the cost of applying the intermediate silica layer.

It is evident that fluorescent lamps will benefit greatly from the use of the bilayer.TM.coated and annealed singly fired willemite phosphor by permitting a lower lamp price that finds more acceptance in the marketplace.

Finally, while what has been described herein has been fluorescent lamps employing CVD-coated phosphors in single- and double-layered configurations, the scope of this disclosure can include lamps which employ multiple layers of phosphor coatings in the fabrication of the lamp, multiple components of the blend in addition to, or otherwise different from, the tri-phosphor blend formulation described herein so long as they contain the bilayer-coated willemite as one of the components, and the use of non-CVD-coated alumina/silica-coated and annealed willemite phosphor.

While there has been shown and described what are at present considered to be the preferred embodiments of the invention, it will be obvious to those skilled in the art that various changes and modifications can be made without departing from the scope of the invention as defined by the appended claims.

Claims

1. A particle of a finely-divided fluorescent lamp phosphor having a bi-layer coating comprising a coating of alumina surrounding a coating of silica which surrounds the particle.

2. A particle in accordance with claim 1 wherein the phosphor is a zinc silicate phosphor.

3. A particle in accordance with claim 1 wherein the phosphor is a calcium halophosphate phosphor.

4. A particle in accordance with claim 1 wherein the bi-layer coating has a thickness of greater than or equal to approximated 100 Angstroms.

5. Particles of a phosphor, each phosphor particle being surrounded by a nonparticulate bi-layer coating comprising a coating of alumina surrounding a coating of silica, said particles when incorporated into a fluorescent lamp having a higher brightness than a lamp incorporating alumina coated phosphor particles.