FIELD OF THE INVENTION This invention relates to a process for preparing polyethylene terephthalate/iono er compositions which exhibit high impact strength and to articles made therefrom. The process involves melt blending polyethylene terephthalate with an iono er of ethylene, an unsaturated carboxylic acid selected from the group consisting of acrylic acid and methacrylic acid wherein the carboxylic acid groups are neutralized with zinc ions, and an alkyl acrylate, at a shear rate of 3500 to 7000 reciprocal seconds; and thermoforming the blend into an article.

BACKGROUND OF THE INVENTION Polyethylene terephthalate (PET) is widely used as an extrusion and injection—molding resin for the fabrication of various articles for household or industrial use, including appliance parts, containers, and auto parts. Because many of such articles must withstand considerable temperature changes and/or physical abuse, it is customary to blend polyethylene terephthalate with other polymers to improve its impact resistance as shown by notched Izod impact values. There are advantages, however, in keeping PET as the matrix material in PET/polymer blends and those are to retain tensile strength, flexural modulus, elongation percent, weather resistance and heat deflection temperature.

U.S. Pat. No. 3,435,093 discloses blends of polyethylene terephthalate and alpha—olefin/alpha—beta unsaturated carboxylic acid copolymers wherein the carboxylic acid groups are 0—100% neutralized by metal

cations such as sodium, potassium, calcium, magnesium, zinc and lead. Moreover, the polyethylene terephthalate is present in an amount of between 55 to 95 weight percent of the blend. Izod impact values of blends indicated in the Examples of U.S. Pat. No. 3,435,093 range from 27.8 J/m to 59.8 J/m at 23°C.

U.S. Pat. No. 4,680,344 discloses blends containing a linear polyester and at least 60 weight percent of alpha-olefin/alpha—beta—ethylenically unsaturated carboxylic acid ionomer neutralized with zinc, calcium, or magnesium. No third comonomer is present. Izod impact values of blends indicated in the Examples of U.S. Pat. No. 4,680,344 range from 26.7 J/m to 1308 J/m at 23°C. U.S. Pat. No. 4,172,859 discloses multiphase thermoplastic molding compositions containing 60—99 weight percent of polyester matrix resin, and 1—40 weight percent of ionomer having a particle size in the range of 0.1—3.0 microns. The compositions are prepared using a multi—screw extruder to generate high shear.

U.S. Pat. No. 4,172,859, however, gives no indication of which shearing parameters are critical and no direction as to which of many shearing block designs are likely to be successful to accomplish a shear rate of at least 3500 reciprocal seconds which the present inventors have determined to be critical.

PCT Application No. WO 92/03505 discloses a semi—crystalline thermoplastic molding composition containing 60 to 90 weight percent of a polyester resin and 10 to 40 weight percent of an ionomer consisting of ethylene, an alkyl acrylate and an unsaturated carboxylic acid. The ionomer has from 20% to 80% of the carboxylic acid groups neutralized with zinc, cobalt, nickel, aluminum or copper (II) .

U.S. Pat. No. 4,753,980 discloses toughened thermoplastic polyester compositions containing 60 to 97 weight percent of a polyester and 3 to 40 weight percent of an ethylene copolymer such as ethylene/methacrylate/glycidyl methacrylate.

U.S. Pat. No. 4,303,573 discloses high velocity impact thermoplastic polyester compositions containing polyethylene terephthalate, 2 to 20 weight percent of an ionomeric terpolymer which is the zinc salt of a terpolymer of ethylene, methacrylic acid, and isobutylacrylate, and 2 to 20 weight percent of a second terpolymer of ethylene, propylene, and 1,4—hexadiene which has succinate groups pendant from the copolymer chain. In contrast, the present inventors have unexpectedly discovered a process for preparing superior impact resistant thermoplastic polyester molding compositions as determined by notched Izod impact values which are double the impact values found in the previously mentioned patents. The process involves melt blending polyethylene terephthalate with an ionomer of ethylene, an unsaturated carboxylic acid selected from the group consisting of acrylic acid and methacrylic acid wherein the carboxylic acid groups are neutralized with zinc ions, and an alkyl acrylate, at a critical shear rate of 3500 to 7000 reciprocal seconds; and forming the blend into an article. High impact strength is obtained even though the inherent viscosity of the polyethylene terephthalate polyester component is significantly reduced due to the high shearing action/ The high shearing process of this invention which is used to improve the impact strength of a polyester thermoplastic composition is contrary to the teachings of U.S. Pat. No. 4,780,506. Such patent teaches, in column 2, lines 6 to 13 that high shear blending of

PET/polycarbonate blends with impact modifiers leads to unpredictable results and transesterification which can be minimized by the use of inhibitors and/or by lowering the shear level.

SUMMARY OF THE INVENTION It is therefore an object of the present invention to improve the impact properties of polyethylene terephthalate/ionomer blends. Another object of the invention is to provide a process for preparing polyethylene terephthalate/ionomer blends under conditions of high shear.

A further object of the invention is to provide polyethylene terephthalate/ionomer blends which exhibit excellent mechanical properties such as impact resistance, stress crack resistance and heat resistance, and which display excellent melt flowability at the time of molding thereof.

These and other objects are accomplished herein by a process for preparing a polyethylene terephthalate/ionomer blend which exhibits high impact strength comprising: (I) melt blending

(A) 70.0 to 90.0 weight percent of a polyester which comprises

(1) a dicarboxylic acid component comprising repeat units from at least 95 mole percent terephthalic acid; and

(2) a diol component comprising repeat units from at least 95 mole percent ethylene glycol, based on 100 mole percent dicarboxylic acid and 100 mole percent diol, said polyester having an inherent viscosity of 0.4 to 1.2 dl/g; and

(B) 30.0 to 10.0 weight percent of an ionomer comprising repeat units from 80 to 95 weight percent of

ethylene and 5 to 20 weight percent of an unsaturated carboxylic acid selected from the group consisting of acrylic acid and methacrylic acid, and the carboxylic acid groups being neutralized to the extent of 40 to 95 percent with zinc ions; wherein the combined weights of (A) and (B) total 100 percent and the blending is conducted in an extruder capable of providing a shear rate of 3500 sec ^-1 to 7000 sec ^-1 ; and (II) forming the blend into an article.

DESCRIPTION OF THE INVENTION The polyester, component (A) , of the present invention is a polyethylene terephthalate (PET) resin. The polyethylene terephthalate resin contains repeat units from at least 95 mole percent terephthalic acid and at least 95 mole percent ethylene glycol, based on 100 mole percent dicarboxylic acid and 100 mole percent diol.

The dicarboxylic acid component of the polyester may optionally be modified with up to 5 mole percent of one or more different dicarboxylic acids other than terephthalic acid or suitable synthetic equivalents such as dimethyl terephthalate. Such additional dicarboxylic acids include aromatic dicarboxylic acids preferably having 8 to 14 carbon atoms, aliphatic dicarboxylic acids preferably having 4 to 12 carbon atoms, or cycloaliphatic dicarboxylic acids preferably having 8 to 12 carbon atoms. Examples of dicarboxylic acids to be included with terephthalic acid are: phthalic acid, isophthalic acid, naphthalene—2,6—dicarboxylic acid, cyclohexanedicarboxylic acid, cyclohexanediacetic acid, dipheny1—4,4'-dicarboxylic acid, succinic acid, glutaric acid, adipic acid, azelaic acid, sebacic acid, and the like. Polyesters may be prepared from two or more of the above dicarboxylic acids.

It should be understood that use of the corresponding acid anhydrides, esters, and acid chlorides of these acids is included in the term "dicarboxylic acid". In addition, the polyester, component (A) , may optionally be modified with up to 5 mole percent, of one or more different diols other than ethylene glycol. Such additional diols include cycloaliphatic diols preferably having 6 to 20 carbon atoms or aliphatic diols preferably having 3 to 20 carbon atoms. Examples of such diols to be included with ethylene glycol are: diethylene glycol, triethylene glycol, 1 ,4—cyclohexanedi ethanol, propane—1,3— iol , butane—1,4—diol, pentane—1,5—diol, hexane—1,6—diol, 3— ethylpentanediol—(2,4) , 2—methylpentanediol—(1,4) , 2,2,4-trimethylpentane—diol—(1,3),

2—ethylhexanediol—(1,3), 2 ,2—dieth lpropane—diol—(1,3) , hexanedio1-(1,3), l,4-di-(hydroxyethoxy)—benzene, 2 ,2—bis—(4—hydroxycyclohexy1)—propane, 2,4-dihydroxy-l,1,3,3—tetramethyl-cyclobutane, 2,2—bis—(3—hydroxyethoxypheny1)—propane, and 2,2—bis—(4—hydroxypropoxyphenyl)—propane. Polyesters may be prepared from two or more of the above diols. The polyethylene terephthalate resin may also contain small amounts of trifunctional or tetrafunctional comonomers such as trimellitic anhydride, trimethylolpropane, pyromellitic dianhydride, pentaerythritol, and other polyester forming polyacids or polyols generally known in the art. Polyesters comprising substantially only dimethyl terephthalate and ethylene glycol are preferred in the case where the blends of the present invention are used in making thermofor ed crystallized PET articles.

Polyethylene terephthalate based polyesters of the present invention can be prepared by conventional

polycondensation procedures well—known in the art. Such processes include direct condensation of the dicarboxylic acid(s) with the diol(s) or by ester interchange using a dialkyl dicarboxylate. For example, a dialkyl terephthalate such as dimethyl terephthalate is ester interchanged with the diol(s) at elevated temperatures in the presence of a catalyst. The polyesters may also be subjected to solid state polymerization methods. The polyester, component (A) , useful in the practice of this invention is the condensation product of terephthalic acid, usually employed as the dimethyl ester, and ethylene glycol, hereinafter referred to as polyethylene terephthalate or PET. The PET has a melting point (Tm) of 255°C. +5°C. and a glass transition temperature (Tg) of 80°C. +5 ^β C. The PET may exhibit a relatively broad molecular weight range as determined by inherent viscosities of from 0.4 to 1.2. However, inherent viscosities of from 0.5 to 0.9 are preferred.

A preferred polyester for use in this invention is a crystallized polyethylene terephthalate having an inherent viscosity of 0.70 which is commercially available as KODAPAK PET 7352 (trademark) from Eastman Kodak Company.

Component (B) of the present invention is an ionomer. Ionomers suitable for use in the present invention consist of copolymers and terpolymers of ethylene, an unsaturated carboxylic acid selected from the group consisting of acrylic acid and methacrylic acid and, optionally, an alkyl acrylate having from l to 8 carbon atoms in the alkyl group. The carboxyl group—containing copolymers and terpolymers usually are converted at least in part to the salt form or, are neutralized to a certain degree. Such neutralization is

obtained by adding to the carboxyl group-containing polymeric material a calculated amount of a zinc salt, for example, zinc acetate, and heating the mixture to a temperature below 140°C. , while thoroughly mixing the materials together. The resulting partly or completely neutralized carboxylic group—containing polymeric material is known generically as an ionomer.

The present inventors have determined through experimentation that cations other than zinc such as aluminum, potassium, sodium and magnesium do not result in improved impact strength for articles incorporating such carboxyl group—containing copolymers and terpolymers. The ionomer has from 40 to 80 percent of the carboxylic acid groups neutralized with zinc. Preferably, the ionomer has from 50 to 75 percent of the carboxylic acid groups neutralized with zinc and most preferably 70 percent. Some of such ionomeric materials are available commercially, for example "SURLYN" (trademark) ionomer resins of the E.I. DuPont de Nemours and Company. Particularly preferred ionomers are SURLYN 9020 which is a random terpolymer of ethylene/methacrylic acid/isobutyl acrylate 70% neutralized with zinc, and SURLYN 9721 which is a ethylene/methacrylic acid copolymer 70% neutralized with zinc.

The ethylene content of the copolymer or terpolymer is at least 50 weight percent, based on the ethylene/acid copolymer or terpolymer. The unsaturated carboxylic acid content of the ionomer should fall in the range of from 2 to 20 weight percent, the preferred range being from 5 to 15 weight percent and the most preferred range being from 8 to 12 weight percent, based on the ionomer to give the best combination of low temperature impact resistance and high temperature resistance. The alkyl acrylate content of the

terpolymer is from 2 to 15 weight percent. Preferably the alkyl acrylate is n—butyl acrylate or isobutyl acrylate. Most preferably, the alkyl acrylate is isobutyl acrylate. Ionomer copolymers of this invention preferably contain repeat units from 80 to 95 weight percent of ethylene and 5 to 20 weight percent of acrylic acid or methacrylic acid. Ionomer terpolymers of this invention preferably contain repeat units from 70 to 90 weight percent of ethylene, 5 to 15 weight percent of acrylic acid or methacrylic acid, and 5 to 15 weight percent of an alkyl acrylate or methacrylate having 1 to 8 carbon atoms in the alkyl group.

Ethylene/methacrylic acid copolymers partially neutralized with zinc but which do not contain an alkyl acrylate, for example, isobutyl acrylate, are not as effective as ethylene/methacrylic acid copolymers partially neutralized with zinc which contain isobutyl acrylate. The present inventors have determined that the presence of an alkyl acrylate tends to reduce the modulus of the ionomer. Isobutyl acrylate, for example, reduces the modulus of the ionomer which in turn gives a more favorable ratio of PET modulus to ionomer modulus. The ratio of PET modulus to ionomer modulus should be greater than 10:1, and preferably greater than 20:1. Thus, the absence of an alkyl acrylate necessarily requires higher concentrations of the ionomer in the polyester/ionomer blend in order to obtain high impact strength. The ionomer generally is present in the blends of the present invention in an amount of from 10 to 30 weight percent. Consequently, at least 70 weight percent of the blends is PET. Such critical amounts take into consideration the advantages which exist in keeping PET as the matrix material. The advantages

include retention of tensile strength, flexural modulus, elongation percentage, and heat deflection temperature. Preferably the concentration of ionomer should be from 15 to 25 weight percent and most preferably from 18 to 22 weight percent.

The compositions of the present invention may be made from a single polyester resin and a single ionomer or from a polyester and a mixture of ionomers.

The process for preparing the polyester/ionomer blends of the present invention involve preparing the polyester and zinc ionomer, respectively, by processes as mentioned previously. The polyester and zinc ionomer are dried in an atmosphere of dried air or dried nitrogen, or under reduced pressure. The polyester and ionomer are blended and subsequently melt blended or compounded in an extruder operated in a manner to provide a shear rate of 3500 sec ^-1 to 7000 sec ^-1 in the melt phase. Such shear rate is essential to provide the blends of this invention with high impact strength. Preferred extruders are twin screw extruders set up to provide a shear rate of 3500 sec ^-1 to 7000 sec ^-1 . The ionomer(s) are dispersed throughout the polyester as discrete particles, which particles have a number average particle size of less than or equal to 1 micron. The zinc ionomer dispersed phase in PET obtained by this type of blending has particle diameters of 0.1 to 0.3 microns.

Torque can be used as a measurement of the amount of shear being applied to a blend. The highest impact properties are achieved with the blends of the present invention at the maximum torque attainable. The maximum torque attainable by the present inventors is 102 J/ϊti which translates into 6000 sec ^-1 . The present invention, however, is not limited by a torque value of 102 J/m. In fact, higher torque values are expected to

result in even greater notched and unnotched impact strength.

The necessary shearing force can be obtained, for example, in an extruder such as a Werner and Pfleiderer ZSK—28mm or ZSK—30mm corotating, intermeshing twin screw extruder, at a melt temperature of 260°C. It is important to note that the Werner and Pfleiderer ZSK—28mm corotating, intermeshing twin screw extruder has at least two different screw designs, a "hard" screw design and a "medium" screw design. The "hard" screw design is a screw configuration which has 215 mm of kneading block length, eight elements which slide on, near the center and end of the screw for mixing and homogenizing the material. Two of the elements are left—handed elements capable of providing a higher shear field. A left—handed screw bushing element is included to back up the flow in the machine to create higher shear. The total length of the "hard" screw is 800 mm. Within the "hard" screw design, there are infinite settings that would provide the necessary shear. The maximum shear rate obtainable with the "hard" screw design on the Werner and Pfleiderer ZSK-28mm extruder is 5500 sec ^-1 . Thus, the "hard" screw is appropriately named since it is "hard" on the polymer. The "medium" screw design has a mixing screw which is the same length as the "hard" screw. The "medium" screw has 45 mm of kneading block length, four elements which slide on, near the center and end of the screw for mixing and homogenizing the material. The maximum shear rate obtainable with the "medium" screw design is less than 3500 sec ^-1 . The present inventors have determined that the impact strength of blends prepared with the "medium" screw design on the Werner and Pfleiderer ZSK-28mm extruder have significantly lower Izod impact values than blends prepared with the "hard" screw

design. Moreover, the present inventors have determined that blends prepared on single screw extruders have even lower Izod impact values than blends prepared with a Werner and Pfleiderer ZSK—28mm extruder having a "medium" screw design.

The twin screw configuration required to attain the high impact compositions of the present invention requires that 25 percent of the screw length contain kneading blocks. These kneading blocks are distributed in groups of 2 to 4, for example, and each group is generally ended with a left—handed kneading block to insure that the kneading block groups are being maintained at full capacity to maximize their mixing capability. However, other configurations that have at least the minimum length of kneading blocks and left-handed kneading blocks will provide the desired results. Such configurations provide maximum shear rates, good extensional flow and backmixing.

Melt temperatures must be at least as high as the melting point of the polyester component or sufficiently above the glass transition temperature for an amorphous polyethylene terephthalate polyester, which typically is in the range of 260-310°C. Preferably, the melt blending or compounding temperature is maintained as low as possible within said range. The composition is molded preferably at 260°C. to 280°C. under low temperature mold conditions such as 23°C. to provide an amorphous molded specimen. High impact strength is obtained even though the I.V. of the polyethylene terephthalate polyester component has been significantly reduced due to the high shearing action. After completion of the melt compounding, the extrudate is withdrawn in strand form, and recovered according to the usual way such as cutting.

Under melt processing conditions the PET undergoes molecular weight degradation in the presence of contaminants such as water, thus, it is preferable that the polyester be incorporated in anhydrous form into the blends of the present invention. The blends should also be protected from moisture prior to melt processing.

Many other ingredients can be added to the compositions of the present invention to enhance the performance properties of the blends. For example, surface lubricants, deneεting agents, stabilizers, antioxidants, ultraviolet light absorbing agents, mold release agents, metal deactivators, colorants such as titanium dioxide and carbon black, nucleating agents such as polyethylene and polypropylene, phosphate stabilizers, fillers, and the like, can be included herein. All of these additives and the use thereof are well known in the art and do not require extensive discussions. Therefore, only a limited number will be referred to, it being understood that any of these compounds can be used so long as they do not hinder the present invention from accomplishing its objects. The blends of the present invention serve as excellent starting materials for the production of moldings of all types. Specific applications include medical parts, appliance parts, automotive parts, tool housings, recreational and utility parts. The molding compositions of the present invention are especially useful in applications that require toughness in hard to fill injection molded parts. Additionally, the blends can be used to prepare extruded sheets for thermoforming applications.

The materials and testing procedures used for the results shown herein are as follows: Break Elongation: ASTM—D638 Density (gradient tube method) : ASTM-D1505

Flexural Modulus and Flexural Strength: ASTM-790 Heat Deflection Temperature: ASTM-D648 Melt Flow Index: ASTM-D1238

Tensile Strength and Yield Strength: ASTM-T638 Izod Impact Strength: ASTM—D256. The Izod Impact Strength Test was repeated three to five times for each material. The letters CB, PB and NB listed under impact strength have the following meanings:

CB - complete break, brittle failure PB - partial break

NB — no break, ductile failure. Inherent viscosity (I.V.) was measured at 23°C. using 0.50 grams of polymer per 100 ml of a solvent consisting of 60% by weight phenol and 40% by weight tetrachloroethane.

Ionomer A is a 80/10/10 weight percent terpolymer consisting of ethylene, isobutyl acrylate and methacrylic acid, respectively, containing 2.63 weight percent zinc. The degree of neutralization of the acid is 69%. Flexural Modulus at 23°C. is 14,000 psi (100 MPa) . Melt Index at 190°C. (grams per 10 minutes) is 1.0. Polyester/Ionomer ratio is 10:1. Ionomer A is commercially available under the trademark SURLYN 9020 from E.I. DuPont de Nemours and Company. Ionomer B is a 80/10/10 weight percent terpolymer consisting of ethylene, isobutyl acrylate and methacrylic acid, respectively, with 70% of the carboxyl groups neutralized with sodium. Melt Index at 190°C. (grams per 10 minutes) is 1.0. Polyester/Ionomer ratio is 10:1. Ionomer B is commercially available under the trade name SURLYN 8020 from E.I. DuPont de Nemours and Company.

Ionomer C is a 90/10 weight percent copolymer consisting of ethylene and methacrylic acid, respectively, with 70% of the carboxyl groups

neutralized with zinc. Melt Index at 190°C. (grams per 10 minutes) is 1.0. Polyester/Ionomer ratio is 10:1.Ionomer C is commercially available under the trade name SURLYN 9721 from E.I. DuPont de Nemours and Company.

Ionomer D is a 90/10 weight percent copolymer consisting of ethylene and methacrylic acid, respectively, containing 0.93 weight percent sodium. The degree of neutralization of the acid is 70%, flexural modulus at 23°C. is 14,000 psi (100 MPa) , and melt index is 1.0 g/10 min @ 190°C. Melt Index at 190°C. (grams per 10 minutes) is 1.0. Polyester/Ionomer ratio is 10:1. Ionomer D is commercially available under the trade name SURLYN 8527 from E.I. DuPont de Nemours and Company.

In the following examples, all the blends of neutralized acid copolymer and terpolymer with polyethylene terephthalate that were prepared on a Werner and Pfleiderer ZSK-28mm twin-screw extruder with "hard" screw design utilized the following conditions:

SET TEMPERATURE (°C.) Zone Zone Zone Zone Zone 1 2 3 4 5

120 237 260 265 267

MELT TEMPERATURE (°C.)

#1 £2 £3 52 288 273

DIE TEMPERATURE (°C.) TORQUE RPM

263 825 232

The resulting pelletized materials were injection molded on a BOY—22S Injection Molding Machine or on a Toyo T90G injection molding machine using the following condition:

Open Cycle Time 4 seconds Injection and Hold Time 14 seconds Cooling Time 12 seconds Injection Time 4 seconds Total Cycle Time 34 seconds

Zone 1 240°C.

Zone 2 260°C. Mold Temperature 23°C. Nozzle temperature 260°C. Screw Speed 125 rpm Injection Pressure 600 psig (4238 KPa) Hold Pressure 600 psig (4238 KPa)

The invention will be further illustrated by a consideration of the following examples, which are intended to be exemplary of the invention. All parts and percentages in the examples are on a weight basis unless otherwise stated.

EXAMPLE 1 A homopoly er of crystallized polyethylene terephthalate having an I.V. of 0.70 was dried at 150°C. for 16 hours in desiccant air with a dew point <—29°C. The PET was placed in the hopper, under dry N ₂ , of a

Werner and Pfleiderer ZSK—28mm corotating, intermeshing twin screw extruder having the "hard" screw design. The PET was melt processed at 260°C. under high shear conditions, stranded and pelletized. The I.V. of the PET was 0.61.

The pelletized PET was dried at 100°C. for 8 hours in desiccant air with a dew point <—29°C. and injection molded on a Boy 22S injection molding machine using a melt temperature of 260°C. and a mold temperature of 23°C. to provide an amorphous test specimen. The I.V. of the PET after molding was 0.55. The impact properties of the PET is summarized in Table I.

EXAMPLE 2 The PET of Example 1 was dried at 150°C. for 16 hours in desiccant air with a dew point <—29°C. Ionomer A was dried at 60°C. for 16 hours in desiccant air with a dew point <—29°C. The PET and Ionomer A were pellet blended in a polyethylene bag such that the _, concentration of Ionomer A was 10 weight percent. The PET/Ionomer A blend was placed in the hopper, under dry N ₂ , of a Werner and Pfleiderer ZSK-28mm corotating, intermeshing twin screw extruder having the "hard" screw design. The blend was melt processed at 260°C. under high shear conditions, stranded and pelletized.

The pelletized blend was dried at 100°C. for 8 hours in desiccant air with a dew point <—29°C. and injection molded on a Boy 22S injection molding machine using a melt temperature of 260°C. and a mold temperature of 23°C. to provide amorphous test specimens. The impact properties of the blend are summarized in Table I.

EXAMPLES 3-5 The procedure of Example 2 was followed except that the concentration of Ionomer A in the PET blend was changed to provide Ionomer A concentrations of 15, 20 and 30 weight percent, respectively. The effect of the zinc ionomer concentrations in PET are summarized in Table I.

EXAMPLES 6-9 The procedure of Example 2 was followed except that Ionomer A was substituted with Ionomer B. The concentration of Ionomer B in the PET blends was 10, 15, 20 and 30 weight percent, respectively. The results are summarized in Table III.

EXAMPLE 10 A blend containing 80 weight percent of the PET of Example 1 and 20 weight percent of Ionomer A was prepared as in Example 2 except that a Werner Pfleiderer ZSK—30mm corotating, intermeshing, twin—screw extruder was used. The extruder has a screw length of 1061 mm and 266 mm of this length is comprised of kneading blocks and left-handed blocks to provide high shear.

The pelletized blend was injection molded on a Toyo T90G molding machine at 265°C. The test results are summarized in Table I.

The results in Table I indicate that essentially identical high impact strength was obtained with this blend as was obtained by the same blend in Example 2 which was prepared on the Werner Pfleiderer ZSK—28mm extruder and molded on the Boy 22S injection molding machine.

TABLE I

Impact Strength of Zinc and Sodium Terpolymer Ionomers

IONOMER IONOMER IZOD IMPACT STRENGTH (J m)

A B Notched Unnotched Notched Unnotched EXAMPLE fwt%) (wt%) f23°C.) (23°C.) f-40°C.) t-40°C. )

EX. 1 0 0

Ex. 10

Ex. 3 15

E . 4 20

Ex. 30

Ex. 6

E . 7 10

Ex. 8 15

Ex. 9 20

Ex. 10 30

The data in Table I indicates that significant increases in notched impact strength are achieved with the PET/Ionomer A blends in spite of the drastic reduction in inherent viscosity experienced by the PET portion of such blends. The decrease in inherent viscosity would have been expected to have resulted in a severe loss of impact strength rather than an increase. For example, the notched Izod impact strength at 23°C. of the PET/Ionomer A blend of Example 4 is 1145 J/m and

no break failure mode compared to 32 J/m and complete break failure mode for the PET control of Example 1. The notched Izod impact strength at —40°C. of the PET blend of Example 4 is 75 J/m compared to 30 J/m for the PET control.

The data also indicates that PET/Ionomer A blends wherein the acid component is neutralized with zinc exhibit significant increases in notched impact strength at 23°C. and —40°C. as compared to PET/Ionomer B blends wherein the acid component is neutralized with some other ion such as sodium. Moreover, the mode of impact failure for the blends utilizing zinc was ductile as opposed to brittle for the blends utilizing sodium. The data in Table I further indicates that a preferred combination of low temperature impact resistance and high temperature impact resistance was achieved where the PET/Ionomer A blends contained from 5 to 15 weight percent ionomer. In contrast, Ionomer B which is the sodium neutralized ionomer, even at the 30 weight percent level only slightly increased notched Izod impact strength from the PET control. For example, at 23°C. notched Izod impact strength for the PET control is 28 J/m with complete break failure mode, and for the PET/Ionomer B blend at 30 weight percent ionomer level is 101 J/m with complete break failure mode.

EXAMPLE 11 The procedure set forth in Example 2 was followed except that the pellet blend composition fed into the Werner and Pfleiderer ZSK—28mm corotating, intermeshing twin screw extruder having the "hard" screw design consisted of 40 weight percent PET and 60 weight percent Ionomer A. The 40/60 PET/Ionomer A concentrate blend in pellet form was blended with sufficient PET pellets to provide a final concentration of 20 weight percent

Ionomer A in the PET. The test results are summarized in Table II. Test results from Example 1 and Example 4 are included for comparison purposes.

TABLE II

Effect of PET/Ionomer Concentrate

IONOMER IZOD IMPACT STRENGTH (J/m)

A Notched Unnotched Notched Unnotched EXAMPLE fwt%ϊ (23°C. (23°C.) t-40°C. ) f-40°C.)

Ex. 1 0 32 2362 30 1949

(5CB) (5NB) (5CB) (3NB,2CB)

Ex. 4 20 1145 1850 75 2200 (5NB) (5NB) (5CB) (5NB)

Ex. 11 20 1240 2147 93 2131

(5NB) (5NB) (5CB) (5NB)

The results in Table II clearly show that essentially identical improvements in impact strength are obtained by preparing a concentrate of the zinc ionomer in PET and then adding the additional PET needed to provide the desired ionomer concentration prior to molding.

EXAMPLE 12 The PET of Example 1 was dried at 150°C. for 16 hours in desiccant air with a dew point <—29°C.

Ionomer A was dried at 60°C. for 16 hours in desiccant air with a dew point <-29°C. The PET and Ionomer A were pellet blended in a polyethylene bag such that the concentration of Ionomer A was 15 weight percent. The PET/Ionomer A blend was placed in the hopper, under dry N ₂ , of a MPM single screw extruder equipped with a mixing screw. The blend was melt processed at 260°C, stranded and pelletized.

The pelletized blend was dried at 100°C. for 8 hours in desiccant air with a dew point <—29°C. and injection molded on a Boy 22S injection molding machine using a melt temperature of 260°C. and a mold temperature of 23°C. to provide amorphous test specimens. The impact properties of the blend are summarized in Table III.

EXAMPLE 13 The PET of Example 1 was dried at 150°C. for 16 hours in desiccant air with a dew point <—29°C. Ionomer A was dried at 60°C. for 16 hours in desiccant air with a dew point <—29°C. The PET and Ionomer A were pellet blended in a polyethylene bag such that the concentration of Ionomer A was 15 weight percent. The PET/Ionomer A blend was placed in the hopper, under dry N ₂ , of a Brabender single screw extruder equipped with a mixing screw. The blend was melt processed at 260°C. , stranded and pelletized. The pelletized blend was dried at 100°C. for 8 hours in desiccant air with a dew point <—29°C. and injection molded on a Boy 22S injection molding machine using a melt temperature of 260°C. and a mold temperature of 23°C. to provide amorphous test specimens. The impact properties of the blend are summarized in Table III.

EXAMPLE 14 The PET of Example 1 was dried at 150°C. for 16 hours in desiccant air with a dew point <—29°C.

Ionomer A was dried at 60°C. for 16 hours in desiccant air with a dew point <—29°C. The PET and Ionomer A were pellet blended in a polyethylene bag such that the concentration of Ionomer A was 15 weight percent. The PET/Ionomer A blend was placed in the hopper, under dry

N ₂ , of a Sterling single screw extruder equipped with a mixing screw. The blend was melt processed at 260°C, stranded and pelletized.

The pelletized blend was dried at 100°C. for 8 hours in desiccant air with a dew point <—29°C. and injection molded on a Boy 22S injection molding machine using a melt temperature of 260°C. and a mold temperature of 23°C. to provide amorphous test specimens. The impact properties of the blend are summarized in Table III.

EXAMPLE 15 The PET of Example 1 was dried at 150°C. for 16 hours in desiccant air with a dew point <—29°C. Ionomer A was dried at 60°C. for 16 hours in desiccant air with a dew point <-29°C. The PET and Ionomer A were pellet blended in a polyethylene bag such that the concentration of Ionomer A was 15 weight percent. The PET/Ionomer A blend was placed in the hopper, under dry N ₂ , of a Werner and Pfleiderer ZSK-28mm corotating, intermeshing twin screw extruder having the "medium" screw design. The blend was melt processed at 260°C, stranded and pelletized.

The pelletized blend was dried at 100°C. for 8 hours in desiccant air with a dew point <—29°C. and injection molded on a Boy 22S injection molding machine using a melt temperature of 260°C. and a mold temperature of 23°C. to provide amorphous test specimens. The impact properties of the blend are summarized in Table III.

TABLE III Effect of Different Extruders

IZOD IMPACT STRENGTH (J/Bl)

The results in Table III clearly show that single screw extruders do not provide the necessary shear to prepare blends with high notched impact strength as compared to twin screw extruders. (The data from Example 3 is included for comparison purposes.) It is important to note that while the "medium" screw design gives less shearing action than the "hard" screw design, the

"medium" screw design gives more shearing action than a

single screw extruder. However, the results also indicate that twin screw extruders do not necessarily provide the proper amount of shear unless the "hard" screw design is employed.

EXAMPLE 16-20 The PET of Example 1 was dried at 150°C. for 16 hours in desiccant air with a dew point <—29°C. Ionomer A was dried at 60°C. for 16 hours in desiccant air with a dew point <-29°C. The PET and Ionomer A were pellet blended in a polyethylene bag such that the concentration of Ionomer A was 20 weight percent. Samples containing only PET were used as control examples. The blend and control sample were run on the same extruder and injection molded on the Boy 22—S injection molding machine. The blend and control sample were annealed and crystallized at 150°C. in a forced air oven for a period of zero, two, four, six, and eight minutes, respectively. The test results are summarized in Table IV. Unannealed bars of the PET control and the blend are included in Table IV for comparison purposes.

TABLE IV

ANNEALING

TIME AT IZOD IMPACT STRENGTH (J/m)

150°C. Notched Unnotched Notched Unnotched

EXAMPLE (min.) r23°C1 t23 °C . Ϊ (-40 °C . ) t-40 ° C .

The results in Table IV clearly show that the blends of the present invention, even in a highly crystalline form, have better impact strength than crystalline PET. As the annealing time at 150°C. is increased from 2 to 8 minutes the crystallinity of the polymers increases as determined by density gradient tube measurements. After the maximum crystallization time of 8 minutes is obtained, the PET/ionomer blend still retains a notched Izod impact strength at 23°C. of 91 J/m compared to 23 J/m for the similarly treated PET

control. This is a 300% increase in impact strength. Low temperature notched Izod impact strength determined at —40°C. also shows improvement over the annealed control. For example, the PET/ionomer blend still retains a notched Izod impact strength at 23°C. of 82 J/m compared to 33 J/ϊn for the similarly treated PET control. It is important to note that even after 8 minutes of annealing, the notched Izod impact strength is higher than the uncrystallized PET control which is 30 J/m at -40°C.

In addition, the unnotched Izod impact strength of the blend at -40°C. after annealing for 8 minutes also shows significantly improved ductile strength, 5NB, as compared to the PET control which had a value of 4CB,1NB indicating mostly brittle failure.

EXAMPLE 21 The PET of Example 1 was dried at 150°C. for 16 hours in desiccant air with a dew point <—29°C. Ionomer A was dried at 60°C. for 16 hours in desiccant air with a dew point <-29°C. The PET and Ionomer A were pellet blended in a polyethylene bag such that the concentration of Ionomer A was 20 weight percent. The PET/Ionomer A blend was placed in the hopper, under dry N ₂ , of a Werner and Pfleiderer ZSK-30mm corotating, intermeshing twin screw extruder having a screw length of 1061 mm wherein 266 mm of the screw length contains kneading blocks and left—handed blocks to provide high shear. The blend was melt processed at 260°C. under high shear conditions, stranded and pelletized.

The pelletized blend was dried at 100°C. for 8 hours in desiccant air with a dew point <-29°C. and injection molded on a Toyo T90G injection molding machine using a melt temperature of 265°C. and a mold

temperature of 23°C. to provide amorphous test specimens.

The same high impact strength was obtained with this blend as was obtained from the identical blend of Example 4 which was made on the Werner and Pfleiderer ZSK—28mm corotating, intermeshing twin screw extruder and molded on the Boy 22S injection molding machine.

EXAMPLES 22-30 The Werner and Pfleiderer ZSK-28mm twin screw extruder has a torque meter. The effect of torque over a range of 67.8 Joules to 101.7 Joules (600in—lb to 900in—lb) was evaluated. The torque was adjusted by changing the throughput rate of the polymer. A higher throughput rate of polymer resulted in higher torque. Torque was also adjusted by changing the extruder RPM.

Example 22 contained the PET of Example 1. Example 22 was not passed through an extruder but was injection molded, thus no torque was applied. Examples 23 to 30 were passed through an extruder and injection molded. Example 23 contained the PET of Example 1 and 101.7 Joules of torque was applied. Example 24 was a PET/Ionomer A blend such that the concentration of Ionomer A was 15 weight percent, and 101.7 Joules of torque was applied. Example 25 contained the PET of Example 1 and 90.4 Joules of torque was applied. Example 26 was a PET/Ionomer A blend such that the concentration of Ionomer A was 15 weight percent, and 90.4 Joules of torque was applied. Example 27 contained the PET of Example l and 79.1 Joules of torque was applied. Example 28 was a PET/Ionomer A blend such that the concentration of Ionomer A was 15 weight percent, and 79.1 Joules of torque was applied. Example 29 contained the PET of Example 1 and 67.8 Joules of torque was applied. Example 30 was a PET/Ionomer A blend such

that the concentration of Ionomer A was 15 weight percent, and 67.8 Joules of torque was applied.

TABLE V

Effect of Torque on Izod Impact Strength

TEST Ex. 22 x. 3 Ex 24 . 25 Ex. 26

TABLE V (continued) Effect of Torque on Izod Impact Strength

Ex. 27 Ex. 28 x. 29 Ex. 30

The results in Table V indicate that higher torque results in more shear being applied to the sample. The data also indicates that the blends of PET/Ionomer A display significantly more impact resistance than the control samples of PET. Moreover, the highest impact properties are achieved with the blends at the maximum torque. Notched and unnotched impact strength continued

to increase for the blends as the torque was increased from 67.8 Joules to 101.7 Joules.

EXAMPLE 31 The PET of Example 1 was dried at 150°C. for 16 hours in desiccant air with a dew point <—29°C. Ionomer C was dried at 60°C. for 16 hours in desiccant air with a dew point <—29°C. The PET and Ionomer C were pellet blended in a polyethylene bag such that the concentration of Ionomer C was 5 weight percent. The

PET/Ionomer C blend was placed in the hopper, under dry N ₂ , of a Werner and Pfleiderer ZSK—28mm corotating, intermeshing twin screw extruder having the "hard" screw design. The blend was melt processed at 260°C. under high shear conditions, stranded and pelletized.

The pelletized blend was dried at 100°C. for 8 hours in desiccant air with a dew point <—29°C. and injection molded on a Boy 22S injection molding machine using a melt temperature of 260°C. and a mold temperature of 23°C. to provide amorphous test specimens. The impact properties of the blend are summarized in Table VI. The impact properties of Example 1 which is the PET control is provided for comparison purposes.

EXAMPLES 32-34 The procedure of Example 31 was followed except that the concentration of Ionomer C in the PET blend was changed to provide Ionomer C concentrations of 10, 15 and 20 weight percent, respectively. The effect of the zinc ionomer concentrations in PET are summarized in Table VI.

EXAMPLES 35-38 The procedure of Example 2 was followed except that Ionomer C was substituted with Ionomer D. The concentration of Ionomer D in the PET blends was 5, 10, 15 and 20 weight percent, respectively. The results are summarized in Table VI.

TABLE VI

Impact Strength of Zinc and Sodium Copolymer Ionomers IONOMER IONOMER IZOD IMPACT STRENGTH (J/m)

C D Notched Unnotched Notched Unnotched EXAMPLE (wt%^ twt% (23°C.Ϊ t23 °C. t-40 °C . ) f-40°C.)

Ex.

Ex. 31

Ex. 32 10

Ex. 33 15

EX. 34 20

Ex. 35

Ex. 36 10

EX. 37 0 15

EX. 38 0 20

The data in Table VI indicates that significant increases in impact strength at 23°C. and -40°C. are achieved with the PET/Ionomer C blends wherein the acid component is neutralized with zinc exhibit as compared

to PET/Ionomer D blends wherein the acid component is neutralized with some other ion such as sodium. Moreover, the mode of impact failure for the blends utilizing zinc was ductile as opposed to brittle for the blends utilizing sodium.

Many variations will suggest themselves to those skilled in this art in light of the above detailed description. All such obvious modifications are within the full intended scope of the appended claims.