CROSS REFERENCED TO RELATED APPLICATIONS 0001 This application claims priority to U.S. patent application Ser. No. 60/311,784 filed Aug. 10, 2001, and is a continuation-n-part of U.S. patent application Ser. No. 08/532,423 filed Mar. 23, 2000.

TECHNICAL FIELD OF THE INVENTION 0002 This invention relates to flame retardant molding compositions.

BACKGROUND OF THE INVENTION 0003 Epoxy resin is widely used in molding compounds for coating electronic devices such as integrated circuits. For safety reasons, molding compositions containing epoxy resin often include flame retardants. A common flame retardant system is a combination of bromine containing flame retardants and antimony oxide flame retardant synergists. However, these compounds are pollutants of the environment. Some bromine containing flame retardants (especially brominated diphenyl ethers) are toxic and possibly carcinogenic. As to antimony trioxide, it is classified by the International Agency for Research on Cancer as a Class 2B carcinogen (i.e., antimony trioxide is a suspect carcinogen based mainly on animal studies). In addition, this compound is often used at a relatively high level (2-4%) and is also slightly water-soluble, leading to further environmental concerns. This concern is highlighted by the fact that integrated circuit manufacturers currently discard up to one half of the total amount of molding compositions used as waste product in landfills. 0004 Phosphorus-containing compounds have been proposed as flame retardants. Although they are less hazardous, molding compositions containing these compounds generally possess undesirable properties such as a high moisture absorption rate. Thus, there exists a need to develop new flame retardant molding compositions that do not contain brominated flame retardants, phosphorus-containing compounds, or antimony oxide flame retardant synergists.

SUMMARY OF THE INVENTION 0005 In general, the invention relates to flame retardant molding compositions that are substantially free of halogen, phosphorus, and antimony. In addition to having good flame retarding properties, these compositions form good cull cure in short time periods and absorb low amounts of moisture, and can be used to coat electronic or electrical devices such as semiconductors, diodes, and integrated circuits. Such coated devices demonstrate good electrical reliability at high temperature. 0006 In one aspect, the invention features a flame retardant molding composition that is substantially free of halogen, phosphorus, and antimony. The molding composition includes an epoxy resin, a first transition metal oxide containing a refractory metal, and a second transition metal oxide containing an oxyanion of a Group VIA element. 0007 In a further aspect, the invention relates to a molding composition which is substantially free of halogen, phosphorous and antimony and which includes an epoxy resin, a first transition metal oxide containing a refractory metal, and a second transition metal oxide containing an oxyanion of a Group VIA element and a cation of a Group IIA element, such as calcium molybdate. 0008 In another aspect, the invention features a flame retardant molding composition substantially free of halogen, phosphorus, and antimony, including an epoxy resin, a phenol novalac hardener containing a biphenyl or naphthyl moiety, and a transition metal oxide of a Group VIA element. 0009 In another aspect, the invention features a flame retardant molding composition that is substantially free of halogen, phosphorus, and antimony, including an epoxy resin containing a biphenyl or naphthyl moiety, a phenol novalac hardener, and a transition metal oxide of a Group VIA element. 0010 The invention also features a method of preparing a flame retardant polymer composition that is free of halogen, phosphorus, and antimony. The method includes heating a molding composition to a temperature sufficient to cure the molding composition (e.g., about 150 C. to about 200 C., or about 165 C. to about 195 C.). The molding composition cures in about 1 minute to about 2 minutes, and contains an epoxy resin, a first transition metal oxide containing a refractory metal, and a second transition metal oxide containing an oxyanion of a Group VIA element. In a further embodiment, the second transition metal oxide may include a cation of a Group IIA element. The invention also features polymer compositions formed by this method. 0011 The invention also features a method of coating an electrical or electronic device such as an integrated circuit. The method includes heating a molding composition to a temperature sufficient to cure the molding composition (e.g., about 150 C. to about 200 C., or about 165 C. to about 195 C.). The polymer composition thus formed coats the surface of the device. The molding composition contains an epoxy resin, a first transition metal oxide containing a refractory metal, and a second transition metal oxide containing an oxyanion of a Group VIA element. In a further embodiment, the second transition metal oxide may include a cation of a Group IIA element. The invention also features coated devices prepared by this method. 0012 As used herein, a composition that is substantially free of a material means that the amount of the material is negligible in the composition, i.e., less than about 0.001 wt % of the total weight of the composition. 0013 As used herein, a refractory metal is a metal having a melting point of around 2,000 C. or above. Some examples of a refractory metal are zirconium, niobium, molybdenum, ruthenium, iridium, hafnium, tantalum, tungsten, osmium, vanadium, chromium, rhenium, and rhodium. 0014 As used herein, an oxyanion is a polyatomic anion containing oxygen, e.g., molybdate and chromate. 0015 As used herein, a compound is water-insoluble when it has a solubility of less than 0.05 g in 100 ml of water at 25 C. 0016 As used herein, a molding composition is cured when it forms a good cull cure (i.e., strong and not brittle). 0017 Other features and advantages of the invention will be apparent from the description of the preferred embodiments thereof, and from the claims.

DETAILED DESCRIPTION OF THE INVENTION 0018 A preferred molding composition includes an epoxy resin, a hardener, and two transition metal oxides, and optionally a third transition metal oxide. 0019 There is no restriction on the type of epoxy resin that can be used in the molding compositions so long as it contains two or more reactive oxirane groups. Some suitable epoxy resins are epoxy cresol novalac resin, biphenyl epoxy resin, hydroquinone epoxy resin, phenolic novalac epoxy resin, and stilbene epoxy resin: Epoxy cresol novalac resin is preferred. The molding compositions can include more than one epoxy resin, for example, a combination of epoxy cresol novalac resin and biphenyl epoxy resin. The preferred weight percent of the epoxy resin ranges from 4 wt % to about 12 wt %, and more preferably, from about 5.5 wt % to about 8.5 wt %, based on the total weight of the molding composition. 0020 The hardener promotes crosslinking of the molding composition to form a polymer composition. Some suitable hardeners that can be included in the molding compositions are phenol novalac hardener, cresol novalac hardener, dicyclopentadiene phenol hardener, and limonene type hardener. Phenol novalac hardener is preferred. Similar to the epoxy resin component, more than one type of hardener can be included in the molding compositions. The preferred weight percent of the hardener ranges from 1 wt % to about 10 wt %, and more preferably, from about 1.5 wt % to about 6 wt %, based on the total weight of the molding composition. 0021 As to the two transition metal oxides, the first transition metal oxide contains a refractory metal, e.g., chromium, molybdenum, and tungsten, and the second transition metal oxide contains an oxyanion of a Group VIA element, e.g., molybdate and tungstate. While there is no particular restriction on the cation of the second transition metal oxide, it is preferred to be a Group IIB metal cation, e.g., zinc. Both the first and second transition metal oxides are preferred to be water-insoluble. Particularly preferred first and second transition metal oxides are tungsten trioxide and zinc molybdate, respectively. 0022 Alternatively, the cation of the second transition metal oxide may be a Group IIA element. In particular, the cation may be a divalent cation of an alkali earth metal, such as calcium. In such an embodiment, particularly preferred first and second transition metal oxides are tungsten trioxide and calcium molybdate, respectively. 0023 The transition metal oxides are preferred to be in their free form, i.e., they are not associated with materials such as silica or talc. In particular, the transition metal oxides are not present as a coating on an inorganic substance for moisture absorption. The oxides are also preferred to be finely divided, e.g., having a diameter of about 0.1 m to about 10 m, preferably, about 0.5 m to about 5 m, or more preferably, about 0.5 m to about 2 m. The oxides can be obtained commercially, e.g., tungsten trioxide is available from Aldrich Chemical Company (Milwaukee, Wis.) and zinc molybdate and calcium molybdate are available from the Sherwin-Williams Company (Cleveland, Ohio). The molding composition can include, for example, about 0.25 wt % to about 2 wt %, preferably about 0.5 wt % to about 1 wt %, and more preferably about 0.75 wt % of the first transition metal oxide based on the total weight of the molding composition. As to the second transition metal oxide, the molding composition can include, for example, about 0.75 wt % to about 6 wt %, preferably about 1 wt % to about 4 wt %, and more preferably about 3 wt % based on the total weight of the molding composition. 0024 The combination of two separate transition metal oxides provides a synergistic effect, resulting in effective flame retardant properties beyond that seen with either transition metal oxide alone. 0025 The preferred molding compositions can include a third transition metal oxide of Group VIA element. An example of such a metal oxide is molybdenum trioxide. The weight percent of the third transition metal oxide in the molding compositions can range from about 0.1 wt % to about 1 wt %, and preferably about 0.5 wt % to about 1 wt %, and more preferably about 0.75 wt % based on the total weight of the molding composition. 0026 Another preferred molding composition contains an epoxy resin, a phenol novalac hardener containing a biphenyl or naphthyl moiety, and a transition metal oxide of a Group VIA element. Preferably, the phenol novalac hardener is the only component of the molding composition that contains a biphenyl or naphthyl moiety. More preferably, the composition is substantially free of epoxy resins other than the cresol novalac type. 0027 The preferred weight percent of the epoxy resin ranges from 4 wt % to about 12 wt %, and more preferably, from about 5.5 wt % to about 8.5 wt %, based on the total weight of the molding composition. 0028 There is no particular restriction on the phenol novalac hardener so long as it contains a biphenyl or naphthyl moiety. The phenolic hydroxy groups can be attached to the biphenyl or naphthyl moiety of the hardener. A preferred phenol novalac hardener is commercially available from Meiwa Plastic Industries, Ltd., Japan (catalog no. MEH 7851, SS graded). This type of hardener can also be prepared according to the methods described in EP 915 118 A1. For example, a hardener containing a biphenyl moiety can be prepared by reacting phenol with bismethoxy-methylenebiphenyl. The weight percent of the phenol novalac hardener containing a biphenyl or naphthyl moiety can range from 1 wt % to about 10 wt %, and more preferably, from about 1 wt % to about 8 wt %, based on the total weight of the molding composition. 0029 Examples of a transition metal oxide of a Group VIA element include oxides of chromium, molybdenum, and tungsten, with tungsten trioxide being the preferred oxide. The molding composition can include, for example, about 0.25 wt % to about 2 wt %, preferably about 0.5 wt % to about 1 wt %, and more preferably about 0.75 wt % of the transition metal oxide based on the total weight of the molding composition. 0030 Yet another preferred molding composition contains an epoxy resin containing a biphenyl or naphthyl moiety, a phenol novalac hardener, and a transition metal oxide of a Group VIA element. Preferably, the epoxy resin is the only component of the molding composition that contains a biphenyl or naphthyl moiety. 0031 The epoxy resin containing a biphenyl or naphthyl moiety can be obtained commercially, e.g., from Nippon Kayaku Co., Ltd., Japan (catalog no. NC-3000P). The preparation of this type of epoxy resin is also described in EP 915 118 A1. For example, an epoxy resin containing a biphenyl moiety can be prepared by reacting phenol with bismethoxy-methylenebiphenyl, followed by treatment with a glycidyl compound such as glycidyl tosylate to form the desired epoxy resin. The preferred weight percent of the epoxy resin containing a biphenyl or naphthyl moiety ranges from 4 wt % to about 12 wt %, and more preferably, from about 5.5 wt % to about 8.5 wt %, based on the total weight of the molding composition. 0032 The preferred weight percent of the phenol novalac hardener ranges from 1 wt % to about 10 wt %, and more preferably, from about 1 wt % to about 8 wt %, based on the total weight of the molding composition. 0033 As described above, a transition metal oxide of a Group VIA element can be an oxide of chromium, molybdenum, and tungsten. Tungsten trioxide is preferred. The molding composition can include, for example, about 0.25 wt % to about 2 wt %, preferably about 0.5 wt % to about 1 wt %, and more preferably about 0.75 wt % of the transition metal oxide based on the total weight of the molding composition. 0034 The molding compositions of this invention can include other additives (the wt % is calculated based on the total weight of the molding composition): 0035 a filler such as silica, calcium silicate, and aluminum oxide (the preferred molding composition can contain 50-95 wt %, more preferably, 60-90 wt % of a filler); 0036 a colorant such as carbon black colorant (the preferred molding composition can contain 0.1-2 wt %, more preferably, 0.1-1 wt % of a filler); 0037 a mold release agent such as camauba wax, paraffin wax, polyethylene wax, glycerol monostearate, and metallic stearates (the preferred molding composition can contain 0.1-2 wt %, more preferably, 0.2-1 wt % of a mold release agent); 0038 fumed silica such as aerosil (the preferred molding composition can contain 0.3-5 wt %, more preferably, 0.7-3 wt % of a fumed silica); 0039 a coupling agent such as the silane type coupling agent (the preferred molding composition can contain 0.1-2 wt %, more preferably, 0.3-1 wt % of a coupling agent); 0040 a catalyst such as 1,8-diazabicyclo-(5,4,0)undecene-7-triphenylphosphone and 2-methyimidazole (the preferred molding composition can contain 0.1-10 wt %, more preferably, 0.5-2 wt % of a catalyst); and 0041 ion scavengers such as magnesium aluminum carbonate hydrate, which can be obtained commercially from Kyowa Chemical Industry Co. under the trade name DHT-4A (the preferred molding composition can contain 0.1-2 wt %, more preferably, 0.5-2 wt % of an ion scavenger). 0042 The molding compositions can be prepared by any conventional method. For example, U.S. Pat. No. 5,476,716 teaches a method of finely grounding, dry blending, and then densifying all of the components of a molding composition on a hot differential roll mill, followed by granulation. Also described in the patent are methods of preparing a molding composition on a laboratory or a pilot plant scale. Alternatively, one can mix the components of a molding composition in a step-wise fashion to enhance homogeneous mixing. Specifically, the first step of the method involves mixing and heating the epoxy resin and the hardener until melting occurs (around 150 C.). The transition metal oxide is then added to the resin and hardener to form a mixture, which is then blended with a mixer until thoroughly mixed (for around 10 minutes). The mixture is allowed to cool until hardened before grounding to a fine powder. The powder is then added to the rest of the components of the molding composition and dry blended before milling. For example, a large two-roll mill (one roll heated to about 90 C., and the other cooled with tap water) can be used to produce uniform sheets, which are then grounded to powder after cooling. 0043 The molding compositions can be molded into various articles by any conventional method, e.g., by using molding apparatus such as a transfer press equipped with a multi-cavity mold for coating electronic devices. Suitable molding conditions include a temperature of about 150 C. to about 200 C. (preferably about 175 C. to about 195 C.) and a pressure of about 400 psi to about 1,500 psi. 0044 The preferred molding compositions cure in about 0.5 minute to about 3 minutes, more preferably, about 1 minute to about 2 minutes. To determine the time for curing (i.e., minimum time needed for forming a good cull cure), the molding composition is placed in the mold press at 190 C. and is inspected after a pre-set period of time (e.g., 3 minutes). If a good cure (i.e., strong and not brittle) is formed, the experiment is repeated with a shorter period of press time until the minimum time period is determined. 0045 The preferred molding compositions demonstrate a flammability rating of UL 94V-1, more preferably, a flammability rating of UL 94V-0. The ratings are determined by measuring the total burn time of a bar according to the UL 94 flammability test. A 94V-0 and a 94V-1 rating require the total burn time for a single bar to be less than or equal to 10 s and 30 s, respectively. 0046 Preferably, the inclusion of transition metal oxides in the molding compositions does not increase the rate of moisture absorption, which is determined by a method similar to ASTM D570-95. Briefly, the procedure involves placing a weighed molded disk of 3 diameter and thick in a rack in an upright position. The rack is then placed on a platform in an 85 C., 85% relative humidity chamber for a pre-determined time period, and the disk is weighed afterwards. The % weight gain is determined by multiplying the quotient (weight difference of the bar before and after placing in the chamber/initial weight of the bar) by 100%. 0047 The electrical reliability of the coated devices in a moist environment is determined by placing the coated devices, e.g., integrated circuits, with no bias in an autoclave at 121 C., 15 psi, and 100% relative humidity. After a number of hours, the coated devices are dried and tested with an electrical tester. The number of coated devices showing a failure in any one of several electrical parameters is counted. These parameters, set by the manufacturer of the device, include, for example, the net DC input offset current for zero device output, the current from device negative input with zero output, the current from the device positive input terminal with zero output, the average of the two previous parameters, DC input offset voltage for zero device output, etc. The % failure is calculated by multiplying the quotient (number of failed devices/total number of tested devices) by 100%. The preferred molding compositions have good electrical reliability in a moist environment, i.e., less than 50% failure after 1,000 hours under conditions as described above. 0048 The high temperature storage life (HTSL) test assesses electrical reliability of the coated devices in a dry environment (i.e., room humidity) and atmospheric pressure. In the HTSL test, parametric shifts in voltage output levels are monitored, and the coated devices are stored at 200 C. The voltage output levels reflect increased resistance across the ball-bonds of the devices. The preferred molding compositions delay or eliminate the failure due to parameter shifts in voltage output levels of coated devices. Similar to the autoclave test, the % failure is calculated by multiplying the quotient (number of failed devices/total number of tested devices) by 100%. The preferred molding compositions have good electrical reliability in a dry environment, i.e., less than 50% failure after 1,500 hours under conditions as described above. 0049 The following examples of flame retardant molding compositions were prepared by dry blending all of the components simultaneously and testing the compositions.

EXAMPLE 1 0050 Six molding compositions, i.e., 18A-23A, were prepared according to the formulations as indicated in Table 1 below. Each molding composition contained a combination of epoxy cresol novalac resin and biphenyl resin. With the exception of composition 23A, each composition contained two transition metal oxides. The weight % (wt %) indicated below were calculated based on the total weight of the molding composition. 1TABLE 1Sample18A19A20A21A22A23ASilica Filler83.0982.8982.4981.6980.1983.19(wt %)Biphenyl Resin5.345.345.345.345.345.34(wt %)Epoxy Cresol1.001.001.001.001.001.00Novalac Resin(wt %)Fumed Silica0.400.400.400.400.400.40(wt %)Carbon Black0.300.300.300.300.300.30Colorant (wt %)Limonene Type1.451.451.451.451.451.45Hardener (fromYuka Shell)(wt %)Phenol Novalac2.472.472.472.472.472.47Hardener (wt %)Catalysts (wt %)0.230.230.230.230.230.23Waxes (wt %)0.740.740.740.740.740.74Silanes (wt %)1.051.051.051.051.051.05Elastomers (wt %)1.501.501.501.501.501.50WO3 (wt %)0.750.750.750.750.750.75ZnMoO4 (wt %)0.100.300.701.503.000Ion Scavenger1.581.581.581.581.581.58(wt %)

0051 Some properties of the cured compositions 18A-23A, i.e., rate of moisture absorption, flammability, and electrical reliability, were determined and summarized below in Table 2. The rate of moisture absorption was measured using a molded disk 3 in diameter and in thickness at 85 C. and 85% relative humidity according to the method described above. The flammability properties of the cured compositions were determined by the total burn time of a molded bar according to the UL 94 test. The results obtained from two electrical reliability tests were shown here. Specifically, in the autoclave test, forty ALS245 die on nickel/palladium lead frame small outline integrated circuit (SOIC) packages with no bias were coated with molding compositions 18A-23A. The test was conducted at 121 C., 100% relative humidity, and 15 psi pressure for 2,250 hours according to the method described above. In the HTSL test, forty LS00 die on nickel/palladium lead frame SOIC packages were coated with molding compositions 18A-23A. The test was conducted at 200 C., room humidity, and atmospheric pressure for 1,500 hours according to the method described above. 2TABLE 2Sample18A19A20A21A22A23AFlammability test(UL 94)Total Burn Time60>13925299>97(5 bars) (s)StatusfailedfailedV-1V-1V-0failedMoistureAbsorption (%)16 hours0.0940.0940.0900.0920.0850.09024 hours0.1140.1170.1130.1150.1070.11648 hours0.1680.1710.1660.1690.1630.17196 hours0.2130.2170.2100.2150.2160.223168 hours0.2500.2550.2500.2580.2680.269Autoclave Test(% failure)After 2250 hours2.30.00.02.314.00.0of testingHTSL Test(% failure)After 1500 hours0.00.00.00.00.00.0of testing

EXAMPLE 2 0052 Six molding compositions of this invention were prepared according to the formulations as indicated in Table 3 below. Each composition contained epoxy cresol novalac resin as the only epoxy resin. With the exception of Sample 34A, each composition contained a transition metal oxide containing a refractory metal, and a transition metal oxide containing an oxyanion of a Group VIA element. The weight % (wt %) indicated below were calculated based on the total weight of the molding composition. 3TABLE 3Sample34A35A36A37A38A39ASilica Filler81.2081.1080.9080.5079.7078.20(wt %)Biphenyl Resin0.00.00.00.00.00.0(wt %)Epoxy Cresol8.038.038.038.038.038.03Novalac Resin(wt %)Fumed Silica0.600.600.600.600.600.60(wt %)Carbon Black0.300.300.300.300.300.30Colorant (wt %)Limonene Type0.00.00.00.00.00.0Hardener (YukaShell) (wt %)Phenol Novalac4.584.584.584.584.584.58Hardener (wt %)Catalysts (wt %)0.170.170.170.170.170.17Waxes (wt %)0.740.740.740.740.740.74Silanes (wt %)1.051.051.051.051.051.05Elastomers (wt %)1.001.501.501.501.501.50WO3 (wt %)0.750.750.750.750.750.75ZnMoO4 (wt %)0.000.100.300.701.503.0Ion Scavenger1.581.581.581.581.581.58(wt %)

0053 Some properties of the cured compositions 34A-39A, i.e., rate of moisture absorption, flammability, and electrical reliability, were determined and summarized below in Table 4. These properties were measured according to methods described in Example 1 above. 4TABLE 4Sample34A35A36A37A38A39AFlammability test(UL 94)Total Burn Time>138>724142589(5 bars) (s)StatusfailedfailedV-1V-1V-1V-0MoistureAbsorption (%)16 hours0.0880.0860.0870.0920.0860.09024 hours0.1220.1190.1190.1240.1150.11848 hours0.1680.1650.1630.1670.1570.16296 hours0.2260.2280.2330.2340.2250.227168 hours0.2370.2370.2450.2350.2370.238Autoclave Test(% failure)After 2000 hours0.00.00.00.02.60.0of testingHTSL Test(% failure)After 1700 hours0.00.00.00.00.00.0of testing

EXAMPLE 3 0054 Seven comparative molding compositions were prepared according to the formulations as indicated in Table 5 below. With the exception of Sample 40A, each composition contained one transition metal oxide. The remaining six molding compositions contained only one transition metal oxide. Specifically, each of Samples 33A, 41A, and 42A contained a transition metal oxide containing an oxyanion of a Group VIA element, i.e., ZnMoO4, and each of Samples 43A, 44A, and 45A contained a transition metal oxide containing a refractory metal, i.e., WO3. The weight % (wt %) indicated below were calculated based on the total weight of the molding composition. 5TABLE 5Sample40A33A41A42A43A44A45ASilica Filler83.9483.1982.4480.9483.1982.4480.94(wt %)Biphenyl Resin5.345.345.345.345.345.345.34(wt %)Epoxy Cresol1.001.001.001.001.001.001.00Novalac Resin(wt %)Fumed Silica0.400.400.400.400.400.400.40(wt %)Carbon Black0.300.300.300.300.300.300.30Colorant(wt %)Limonene1.451.451.451.451.451.451.45Type Hardener(Yuka Shell)(wt %)Phenol2.472.472.472.472.472.472.47NovalacHardener(wt %)Catalysts0.230.230.230.230.230.230.23(wt %)Waxes (wt %)0.740.740.740.740.740.740.74Silanes (wt %)1.051.051.051.051.051.051.05Elastomers1.501.501.501.501.501.501.50(wt %)WO3 (wt %)00.751.503.00000ZnMoO400000.751.503.00(wt %)Ion Scavenger1.581.581.581.581.581.581.58(wt %)

0055 The flammability properties of the cured compositions 33A and 40A-45A were determined as described in Example 1 above and summarized in Table 6 below. 6TABLE 6Sample40A33A41A42A43A44A45AFlammabilitytest (UL 94)Total Burn51>100>138>135>152>166>113Time (5 bars)(s)StatusfailedfailedfailedFailedfailedfailedfailed

EXAMPLE 4 0056 Two molding compositions were prepared according to the formulations as indicated in Table 7 below. Each of compositions 17A and 18B contained three transition metal oxides, i.e., WO3, ZnMoO4, and MoO3. The weight % (wt %) indicated below were calculated based on the total weight of the molding composition. 7TABLE 7Sample17A18BSilica Filler (wt %)79.8180.17Biphenyl Resin (wt %)0.05.34Epoxy Cresol Novalac Resin (wt %)6.100.0Fumed Silica (wt %)0.800.60Carbon Black Colorant (wt %)0.300.30Phenol Novalac Hardener (wt %)4.864.68Catalysts (wt %)0.240.22Waxes (wt %)0.740.74Silanes (wt %)1.051.05Elastomers (wt %)0.500.50WO3 (wt %)0.750.75ZnMoO4 (wt %)3.003.00MoO3 (wt %)0.750.75Ion Scavenger (wt %)1.501.50

0057 The flammability properties of the cured compositions 17A and 18B, as well as the time required to form a good cull cure, were determined and summarized in Table 8 below. 8TABLE 8Sample17A18BFlammability test(UL 94)Total burn time (5 bars) (s)14StatusV-0V-0Minimum time to form a good cull cure1.01.0(min)

EXAMPLE 5 0058 Four molding compositions, i.e., 10B, 79A, 81A, and 10B, were prepared according to formulations as indicated in Table 9 below. Each of compositions 79A and 81A contained a phenol novalac hardener containing a biphenyl or naphthyl moiety (available from Meiwa Plastic Industries., Ltd., catalog no. MEH 7851 SS) and a transition metal oxide of a Group VIA element, i.e., WO3. Composition 10B contained an epoxy resin containing a biphenyl or naphthyl moiety (available from Nippon Kayaku Co., Ltd., catalog no. NC-3000P) and a transition metal oxide, i.e., WO3. Composition Sample 80A, on the other hand, only contained a transition metal oxide, i.e., WO3 The weight % (wt %) indicated below were calculated based on the total weight of the molding composition. 9TABLE 9Sample80A10B79A81ASilica Filler (wt %)80.7880.7880.7883.45Biphenyl Resin0.00.00.05.64(wt %)Epoxy Cresol8.850.07.250.0Novalac Resin(wt %)NC-3000P (wt %)0.09.750.00.0Fumed Silica0.400.400.400.60(wt %)Carbon Black0.300.300.300.30Colorant (Wt %)Phenol Novalac4.893.991.441.64Hardener (wt %)MEH 7851 (wt %)0.00.05.052.97Catalysts (wt %)0.160.160.160.16Waxes (wt %)0.740.740.740.86Silanes (wt %)1.051.051.051.05Elastomers (wt %)0.500.500.501.00WO3 (wt %)0.750.750.750.75Ion Scavenger1.581.581.581.58(wt %)

0059 The flammability properties of the cured compositions 10B and 79A-81A were determined and summarized below in Table 10. 10TABLE 10Sample80A10B79A81AFlammability test(UL 94)Total burn time (5 bars)>1142814(s)StatusFailedV-0V-0V-0

EXAMPLE 6 0060 Two molding compositions, 101 and 102, were prepared according to the formulations as indicated in Table 11 below. Each molding composition contained a combination of two transition metal oxides, with Sample 101 including ZnMoO4 as the second transition metal oxide and Sample 102 including CaMoO4 as the second transition metal oxide. The weight % (wt %) indicated below were calculated based on the total weight of the molding composition. 11TABLE 11Sample101102Silica Filler (wt %)79.5579.55Epoxy Cresol Novalac6.116.11Resin (wt %)Carbon Black Colorant0.300.30(wt %)Phenol Novalac Hardener5.305.30(wt %)Catalysts (wt %)0.120.12Waxes (wt %)0.740.74Silanes (wt %)1.051.05Elastomers (wt %)1.501.50WO3 (wt %)0.750.75ZnMoO4 (wt %)3.000CaMoO4 (wt %)03.00Ion Scavenger (wt %)1.581.58

0061 The flammability properties, rate of moisture absorption and electrical reliability of the cured compositions 101 and 102 were determined and summarized below in Table 12. The rate of moisture absorption was measured using a molded disk 3 in diameter and in thickness at 85 C. and 85% relative humidity according to the method described above. The flammability properties of the cured compositions were determined by the total burn time of a molded bar according to the UL 94 test. Also, results from three electrical reliability tests are shown. 0062 Specifically, in the autoclave test, 20 ALS245 die on nickel/palladium lead frame SOIC packages with no bias were coated with molding compositions 101 and 102. The autoclave test was conducted at 121 C., 100% relative humidity, and 15 psi pressure for over 3000 hours according to the method as described above. Additionally, two HTSL tests were conducted, both of which included 20 ALS245 die on nickel/palladium lead frame SOIC packages, which were coated with molding compositions 101 and 102. The first HTSL test was conducted at 200 C., room humidity, and atmosphere pressure for over 3200 hours. The second HTSL test was conducted under similar conditions, with the exception that the temperature was 210 C. 12TABLE 12Sample101102Flammability test (UL 94)Total Burn Time (5 bars) (s)023StatusV-0V-0Moisture Absorption (%)24 hours0.1160.11948 hours0.1800.17996 hours0.2450.230168 hours0.2970.262Autoclave Test (% failure)94 hours2.40385 hours2.401010 hours2.401460 hours2.401655 hours2.48.61960 hours2.48.62073 hours2.411.42691 hours21.440.03064 hours42.965.7HTSL Test 1 200 C.(% failure)132 hours00572 hours001100 hours001584 hours2.102288 hours2.103234 hours8.318.8HTSL Test 1 200 C.(% failure)132 hours00572 hours001100 hours001584 hours2.802288 hours2.803234 hours25.010.0

0063 As is apparent, the molding compositions of Sample 101, including ZnMoO4 in combination with WO3, and Sample 102, including CaMoO4 in combination with WO3, both passed flammability testing. In addition, both molding compositions demonstrated favorable moisture absorption, with the molding composition of Sample 102 which includes CaMoO4 in combination with WO3 having a lower moisture absorption after 48 hours than the molding composition of Sample 101 including ZnMoO4 in combination with WO3. 0064 Also, the electrical reliability testing demonstrates acceptable failure levels for the molding compositions of Sample 101 and 102. In particular, the electrical reliability data above was plotted using a standard Weibull Distribution to obtain a statistically determined time to reach 50% failure95% confidence interval. Such plotting obtained a 50% failure time of 3350430 hours for Sample 101, and a 50% failure time of 2830210 hours for Sample 102. These results demonstrate no statistical difference (at 95% confidence) in the time to reach 50% failure for Samples 101 and 102. 0065 Other embodiments are within the claims.