Background art Physical blends of different polymer resins are well known in the art. Several attempts have been made to obtain materials endowed with improved properties by blending polypropylenes and polyamides. However, these polymer resins are incompatible and a compatibilizing agent is requested. Thus, US Patent 5,342,886 discloses a polymer alloy comprising a polyamide, a polyolefin and a compatibilizing amount of a graft copolymer.

The latter is prepared by introducing onto a polyolefin back- bone a reactive group, most preferably maleic anhydride (MAH).

Subsequently, a separately prepared monoamino-substituted polyamide oligomer is reacted with the MAH groups to obtain a compatibilizing graft copolymer. US 5,605,961 discloses a thermoplastic composition comprising a polypropylene, a poly- urethane and a compatibilizing agent consisting of a polyprop- ylene grafted with 0.5 % MAH reacted with a polyamide (PA6,36) resin.

An alternative route to blending is the polymerisation of a monomer in the presence of an already prepared polymer made from a different monomer. However, unless both the newly pre- pared and existing polymers are miscible, compatibilization will be needed to avoid phase separations in the blend.

Compatibilization functions by improving the morphology and interfacial properties within a blend. The compatibilizing agent may be an interfacial agent, an emulsifier or a specific agent, such as a graft or block copolymer added separately or

prepared in situ during the blending process. The compati- bilizer will self-locate to the interfaces between the main components of a blend and thus reduce the interfacial ten- sions. As a result the dispersion of the phases will be im- proved, and the adhesion between the interfaces enhanced.

A separately prepared compatibilizing agent has the advantage of having a controlled molecular structure. Unfortunately, for many commercially relevant blends this procedure is not app- licable. The method of preparing the compatibilizing agent in situ during the blending process, often called reactive blend- ing or reactive extrusion, is cost-effective since the com- patibilizer is formed directly at the interfaces within the composition. Moreover, in situ methods eliminate the problems related to dispersing a separately prepared agent into a poly- mer matrix, followed by its slow diffusion to the interfaces and the possibilities of micelle formations. This makes the in situ method very attractive. On the other hand, since the reaction takes place primarily at the interfaces, it will be difficult to control the actually prepared amount of com- patibilizer and its molecular structure.

The in situ procedure above has been applied by Kotlar et al. in WO 96/06872 to obtain compatibilized blends of PP and PBT, the disclosure of which is included herein by reference.

As noted above, an important pair of polymers that might re- sult in attractive polymer blends are PP and PA. Unfortu- nately, they are immiscible and simply blending them will lead to a material having poor mechanical properties. Moreover, it is virtually impossible to obtain a pure graft or block co- polymer of PP and PA. In the art it is known to functionalize a PP with maleic anhydride (MAH) and prepare a graft copolymer through blending because MAH will react readily with the ter- minal amine group of PA. However, once an anhydride moiety on the PP chain has reacted with a PA chain, the nabouring anhyd- ride moiety may for steric reasons become inaccessible for the next PA molecule. This explains the experienced low amount of

copolymer formation of MAH grafted PP with PA (in general less than 5 % by weight).

It has now surprisingly been found that an improved blend may be obtained by grafting onto the PP backbone a reactive group of isocyanate which can function as a starting point for an in situ polymerisation.

Summary of invention Thus, the present invention provides a compatibilized polymer composition of polyolefin and polyamide, comprising a grafted copolymer consisting of a polyolefin homopolymer grafted with 3-isopropenyl-a, a-dimethyl-benzene isocyanate (TMI) and bonded to the grafted TMI groups poly (e-caprolactam) chains.

The present invention also provices a method for the prepara- tion of the compatibilized composition defined above, compris- ing the following steps: (i) grafting a polyolefin homopolymer with an isocyanate com- pound, (ii) polymerising from the grafted isocyanate groups e-capro- lactam monomers in the presence of an activator and a catalyst into poly (e-caprolactam) chains.

Brief description of drawings Figure 1 shows the influence of styrene when TMI is grafted onto PP chains at various concentrations of initiator DHBP.

Figures 2 and 3 show torque and temperature during in situ polymerisations of e-caprolactam in the presence of PP and/or PP-g-TMI in a batch mixer.

Figure 4 shows DSC thermograms for PA6, PP/PA6 blends and PP-g-PA6 copolymers.

Best modes for carrying out the invention The first step in obtaining copolymers of PP having side chains of polycaprolactam involves grafting of reactive iso- cyanate groups onto the PP backbone. Suitable polypropylene resins comprise all polypropylene homopolymers and copolymers that can be grafted with reactive isocyanate groups. Parti-

cularly preferred polypropylenes are polypropylene homopoly- mers, and copolymers of propylene and ethylene containing up to 30 % of ethylene. These polypropylenes should preferably have a molecular weight of 150,000 to 500,000, and a melt flow index (MFI) of 0.2 to 500 g/10 minutes, preferably 0.2 to 200 g/10 minutes, determined according to the method of ASTM D 1238 using 230°C and 2.16 kg load. The polypropylene resins may be used in the form of granules or as a powder.

To initiate the grafting process any free radical generating peroxide compound known in the art having a suitable decom- position temperature may be used. In particular the following peroxide initiators have been found to give acceptable result- s: 2,5-bis (tert-butylperoxy)-2,5-dimetylhexane ("DHBP"), 2,5-bis (tert-butylperoxy)-2,5-dimethyl-3-hexyne ("Trigonox 145") and bis (tert-butylperoxyisopropyl) benzene ("Perkadox 14"). A preferred level of peroxide in the polymer blend is in the range between 0.05 to 5 %, more preferably from 0.05 to 2 %, by weight based on the polypropylene resin.

The functionalized reactive isocyanate group to be grafted onto the PP polymer backbone must have sufficient thermal stability to endure normal polyolefin extrusion conditions (exposure to air and humidity at high temperature). Preferred compounds to be grafted onto the PP backbone are isocyanates selected from the group comprising (2-alkyl vinyl benzene) alkyl isocyanates. A preferred compound is 3-isopropenyl-a, a- dimethylbenzene isocyanate (TMI) having the formula: A preferred level of TMI in the polymer blend is in the range from 1 to 10 %, preferably about 3 %, by weight based on the

polypropylene resin.

This functionalizing monomer can be fixed onto the PP chains by free radical grafting. In what follows, PP modified with TMI will be denoted PP-g-TMI. The free radical grafting yield of TMI onto polypropylene is usually low, unless large amounts of a free radical initiator is charged. Moreover, the mole- cular weight of PP suffers from a significant reduction by (3-scission. The mechanisms involved are now well understood and have been documented. Basically there are two solutions to overcome the problem. One is to enhance the free radical reac- tivity of the double bond of the grafting monomer toward PP tertiary macroradicals, and the other is to improve the reac- tivity of the macroradicals towards the double bond of the monomer. Addition of electro-donating species such as styrene allows to increase reactivity by a charge transfer mechanism through the forming of an adducted charge transfer complex (CTC). By adding styrene the amount of grafted isocyanate groups increases and simultaneously the PP-chain degradation is reduced. This is illustrated in figure 1 which shows the grafting yield of TMI onto PP at various concentrations of initiator (DHBP) at a temperature of 180 °C and an amount of TMI = 3.0 phr without the presence of styrene (D) and in the presence of 1 mole styrene per mole of TMI ( [St] i/ [TMI] i = 1.0 mole/mole, or 0.52 g/g) (|). Suitable styrenic compounds are those having the formula: in which R2 is H, OH, CH3 or allyl. Preferably R2 is H, making styrene the preferred styrenic compound. Other compounds than those mentioned above having conjugated unsaturated double bonds, such as quinone, may give similar results. A preferred level of styrene in the polymer blend is in the range between 0.01 to 10 %, more preferably from 0,5 to 5 %, by weight based on the polypropylene resin.

To obtain the modified polypropylene (PP-g-TMI) the PP as a powder or granules may suitably be premixed with adequate

amounts of the free radical initiator, the isocyanate and styrene. Preferably, these components are mixed mechanically at ambient temperature until the liquid isocyanate and styrene have been absorbed completely into the PP granules. The obtained mixture should then be subjected to a more intensive blending at elevated temperatures. During this blending the free radical initiator must decompose and initiate the graft- ing reactions. It is not necessary to melt the PP granules.

However, in a continuous process an extruder may be used and the PP subjected to reactive extrusion.

The obtained degree of grafting can be determined by IR ana- lyses on thin films of the grafted material. The IR spectro- grams show a peak at 2255 cm-1 corresponding to the isocyanate group of TMI and a peak at 2772 cm-1 characteristic of PP. The areas under these peaks may be used to calculate the grafting efficiency.

Polyamide 6 (PA6) is obtained through the polymerisation of £-caprolactam monomers, which are commercially available as a solid having a melting temperature about 71 °C at normal pres- sure. The polymerisation is performed in the presence of a catalyst. Suitable catalysts may be selected from the group comprising alkali metal or earth alkali metal caprolactam com- pounds. A preferred catalyst is sodium caprolactamate, com- mercially available as solid flakes. This catalyst is used in an amount preferably corresponding to 0.5 to 3.5 moles sodium per kg £-caprolactam, more preferably about 1.4 moles sodium per kg e-caprolactam.

In addition to the catalyst an activator should also be used.

Suitable activators may be selected from the group comprising alkyl diisocynates. A preferred activator is hexamethylene diisocyanate containing blocks of caprolactam. This activator is used in an amount corresponding to 0.5 to 4.0 moles of the contained isocyanate per kg e-caprolactam. A preferred level is 2.0 moles of the contained isocyanate per kg e-caprolactam.

This activator is commercially available in a powdery form.

In this invention polyamide 6 (PA6) is preferably prepared by an activated anionic polymerisation of e-caprolactam in the presence of the catalyst and activator indicated above.

Despite still some debate over the reaction mechanism, it is believed that the reaction scheme below should represent a reasonable description of the reaction mechanism, when side reactions are disregarded.

Initiation Propagation The polymerisation starts with the formation of an acyl capro- lactam as a result of the reaction between an isocyanate acti- vator, such as R-N=C=O, and e-caprolactam (1). This acyl cap- rolactam then reacts readily with a catalyst such as sodium caprolactamate (Na-caprolactamate) creating a new reactive

sodium salt (2). This reaction process then initiates the polymerisation of e-caprolactam, and the catalyst (Na-capro- lactamate) is restored (3). Repetition of Reactions (2) and (3) finally leads to a high molecular weight PA6 (4,5). As opposed to the hydrolytic polymerisation of PA6 which requires a reaction period of several hours, the activated anionic polymerisation of e-caprolactam can be accomplished in a few minutes. This time range is comparable to the residence time in a screw extruder.

Any reactive isocyanate group is able to function as a starting site for the polymerisation of e-caprolactam mono- mers. This also applies to isocyanate groups grafted onto a polymer backbone. The reaction scheme is similar to that out- lined above, including the use of a catalyst and an activator.

According to reactions (1) and (2) above, the isocyanate activator and the e-caprolactam will form acyl caprolactam which acts as a starting site for the polymerisation of e-cap- rolactam. The initiation step involving an isocyanate-bearing polymer (macro-activator precursor) may be illustrated with equation (6): Thus, by grafting isocyanate groups onto a polypropylene homo- polymer, these groups will function as starting sites for the homopolymerisation of e-caprolactam. In this way it is poss- ible to obtain a copolymer having a polypropylene backbone and polyamide chains bonded thereto. In the following this copoly- mer will be termed grafted copolymer, abbreviated PP-g-PA6.

The expected molecular architecture of the copolymer formed can be schematically represented as follows:

The polymerisation of e-caprolactam in the presence of the grafted PP-g-TMI can theoretically follow two different reac- tion paths: the e-caprolactam may polymerise from the isocya- nate groups grafted onto the polypropylene backbone, or e-cap- rolactam may polymerise into polyamide chains which then may react with the isocyanate groups grafted onto the polypropyl- ene backbone. Thus, in the e-caprolactam/PP-g-TMI polymerising system, it could well be that e-caprolactam is first converted to PA6 which then reacts with PP-g-TMI leading to the for- mation of the desired copolymer. In order to clarify this possibility, PP-g-TMI was added to a mechanical blend of PP/PA6 and blended in a batch mixer at reaction conditions.

The obtained material was investigated in a scanning electron microscope. No morphology changes were observed. This confirms that the PP-g-TMI does not react with end groups of the poly- amide molecules. Consequently, in the formation of the grafted copolymer, PP-g-TMI will initiate an anionic polymerisation of e-caprolactam.

The polymerisation of e-caprolactam will obviously both result in polyamide chains on the polypropylene backbone, and the formation of separate PA6 as such. The reaction rates of the activated anionic polymerisation, the polycaprolactam chain length of the PP-g-PA6 copolymer and PA6 homopolymer, as well as the morphology, will obviously be influenced by parametres such as the concentration of the TMI groups available, the level of monomeric e-caprolactam, catalyst and activator, as well as the reaction temperature and other polymerisation con- ditions. The variation of such parametres will be obvious to an artisan.

The blends of the present invention will comprise the PP-g-PA6 copolymer, a certain amount of PA6 homopolymer and at least a minor amount of non-grafted PP from the grafting step. It will be obvious that the amounts of PA6 homopolymer and ungrafted

PP may be chosen freely by selecting the conditions during the preparation. Consequently, it will also be possible to obtain a final composition consisting essentially of PP-g-PA6 copoly- mer. In principle, this copolymer may be refined to obtain a pure copolymer. PP is soluble in hot xylene and PA6 in formic acid. Therefore, these two components may be solubilized com- pletely, while the PP-g-PA6 copolymer will remain as an insol- uble rest. Theoretically, a pure graft copolymer of PP and PA6 can be obtained if PP-g-TMI is used as the only activator. In practice, in this case the conversion of e-caprolactam to PA6 will be rather low. The reason is that e-caprolactam and PP-g- TMI are immiscible and a mixture of them is phase-separated.

It is believed that the copolymer is formed at the interphase between PP-g-TMI and e-caprolactam, and since the nature at the interphase between the grafted polymer and the monomer (a small molecule) is very different from that between two polymers, relatively large amounts of copolymer can still be obtained. This mechanism will be different from using maleic anhydride functionalized PP and PA6 polymer, where the reac- tion will take place between two polymers.

The composition of the invention can be used with good results as a compatibilizing agent in any blends of polyolefins and polyamides, since it will be miscible with both these types of polymer resins. However, the preferred polyamide is PA6. Pre- ferred polyolefins are polyethylene and polypropylene homo- polymer and copolymers, and mixtures thereof. Most preferably the polyolefin is the polypropylene used in the grafting pro- cess. In this way it will be possible to obtain directly a final composition blend.

Articles made from the preferred compositions of the present invention comprising blends of PP, PP-g-PA6 and in situ poly- merised PA6 are endowed with properties from both the poly- propylene and polyamide 6 components of the composition. Com- pared to common polypropylene materials the compositions of the invention will have a better oxygen barriere, improved paintability, improved printability, improved dyability, higher impact strength and higher modulus of elasticity, and a

higher heat deflection temperature. The composition is parti- cularly suited for conversion to blown and cast film, injec- tion molded articles, and calandered sheets.

In the following examples specific features and preferred embodiments of the invention are explained in more details.

Based on presented experimental results also figures 1 to 4 will be explained.

Examples Example 1 Preparation of modified polypropylene (PP-q-TMI): A flask equipped with a stirrer was charged with 40 g of gran- ules of a PP homopolymer having a melt flow index (MFI) of 3.5 g/10 min (230 °C, 2.16 kg load) (commercially available from company Borealis AS, Norway, under the trade name"Borealis PP BC 145M") together with 2.5 g of a technically pure grade of 3-isopropenyl-a, a-dimethyl-benzene isocyanate (TMI) and 2.5 g of styrene (both the TMI and styrene are liquids at room tem- perature), as well as 0.6 g of 2,5-bis (tert-butylperoxy)-2,5- dimetylhexane (DHBP) as a free radical initiator which has a half-life of about 6 seconds at 200 °C. The components were mixed mechanically until the liquids had been absorbed into the PP pellets (after about 15 minutes). The obtained mixture was then transferred to a Rheocord Haake mixer (50 cm3) having two sigma rotors rotating in opposite directions at 64 rev./min (rpm), preheated to 200°C and allowed to react for 10 to 15 minutes. The modified PP granules were then dissolved in boiling xylene and precipitation in acetone at room tempera- ture to obtain a powder of pure, modified PP.

A sample of the obtained PP-g-TMI powder was hot pressed be- tween two stainless steel plates into a film of thickness about 100 pm. This film was then used to determine the degree of TMI grafting by using a Fourier transform infrared (FTIR) method. In the obtained spectra the grafted PP-g-TMI is characterized by the appearance of a peak at 2255 cm~1 corre- sponding to the isocyanate group of TMI. This peak was used to quantify the yield of TMI grafting. The peak at 2772 cm~1,

which is characteristic of PP, was chosen as an internal re- ference. It was obtained a degree of grafting of 1.8 parts by weight of TMI per hundred parts by weight of PP.

Example 2 (Comparative Example) In situ polymerisation of c-caprolactam in the presence of PP The Haake mixer used in Example 1 was preheated to 230 °C and charged with a premixed mixture comprising 50 phr of the PP homopolymer specified in example 1,50 phr of e-caprolactam monomer which is a solid at room temperature (melting point about 71 °C at normal pressure), 3.0 phr of a catalyst con- sisting of Na-caprolactamate containing 1.4 moles sodium per kg e-caprolactam (commercially available in the form of flakes as catalyst"C10") and 3.0 phr of an activator being hexa- methylene diisocyanate having blocks of caprolactam and con- taining 2.0 moles isocyanate per kg e-caprolactam (commer- cially available in a powdery form under the trade name "V5/C-20 Activator"). The e-caprolactam, catalyst and activator were milled together into a powder and premixed with the PP. This mixture was heated in the mixer at 230 °C and processed with a mixing speed of 64 rpm for 10 minutes, then collected and quenched with liquid nitrogen.

The obtained blend contained no copolymers of PP with PA6 and is in the following termed"PP/PA6 mechanical blend".

Details are given after example 3 below.

Example 3 In situ polymerisation and compatibilization of PP/PA6 blends The general polymerisation procedure of Example 2 was repeat- ed, except that at least a part of the PP homopolymer was re- placed with the PP-g-TMI (grafted PP) from Example 1. In all experiments the reaction mixture comprised 50 phr of e-capro- lactam, 3.0 phr of the catalyst and 3.0 phr of the activator, while the amounts of PP homopolymer and the grafted PP-g-TMI were 0 and 50 phr; 20 and 30 phr; 30 and 20 phr by weight, respectively. In the experiment with no PP homopolymer the PP-g-TMI contained 0.4 % of TMI, in the other experiments 1.8 % by weight.

In Examples 2 and 3 during the polymerisations the mixing torques and temperatures within the mixtures were measured.

The recorded values are plotted vs. time in figure 2. The tem- perature profile was virtually the same in all the systems and therefore only one graph is presented. The evolution of torque vs. time is quite similar for all the polymerising systems of examples 2 and 3. The torque first increases very rapidly, reaching a maximum in about 4 to 5 minutes and then decreases drasticly. This indicates that in each polymerising system the polymerisation of e-caprolactam starts immediately after that the mixing chamber has been charged with the reactants. It is likely that the polymerisation reactions are completed within a reaction time of 5 minutes. The ultimate decrease in torque is believed to imply that after the polymerisation is accom- plished, the resulting biphasic PP/PA6 systems undergo a sort of morphological rearrangements to minimize energy. A slightly higher torque is observed for the system containing 20 % of PP and 30 % of PP-g-TMI. In this particular system, more copoly- mer of PP-g-PA6 is probably formed because of the greater <BR> <BR> <BR> amount of PP-g-TMI present (30 % b. w.) and because the PP-g-TMI contains a relatively high amount of grafted TMI (1.8 phr). The produced copolymer causes a viscosity increase of the system through interfacial and/or morphological modifi- cations.

Example 4 In situ polymerisations of e-caprolactam in the presence of PP and varying amounts of PP-Q-TMI The general procedure of example 3 was repeated, except that in all experiments it was used 30 phr of e-caprolactam, 3.0 phr of catalyst and 3.0 phr of activator, while the combined amount of PP homopolymer and grafted PP-g-TMI was 70 phr, of which the amount of introduced PP-g-TMI was 0 %, 2 %, 10 %, 20 % and 40 % by weight, respectively. The PP-g-TMI component contained 1.8 phr grafted TMI.

During the polymerisations the mixing torques and temperatures within the mixtures were measured. The recorded values are plotted vs. time in figure 3. The temperature profile was vir-

tually the same in all the systems and therefore only one graph is presented. This figure indicates that in all experi- ments PP starts melting soon after the mixing chamber is charged with the polymerising system. This corresponds to the first peak of the torque histogram. The torque then goes down quickly to almost zero, which is indicative of the formation of an immiscible liquid system of e-caprolactam and PP. This is then followed by a rapid increase in torque, indicating that the polymerisation of e-caprolactam has been proceeding rapidly. After passing through a maximum, for the reasons stated above the torque finally decreases in a gradual manner.

It is interesting to note that the higher the amount of PP-g-TMI, the greater the maximum and final values of the torque.

Example 5 Example 4 was repeated, except that the amount of e-caprolac- tam comprised 50 phr by weight, and the combined amount of PP and PP-g-TMI comprised 50 phr by weight, of which the amounts of PP-g-TMI used were 0 %, 20 % and 30 % by weight, respec- tively. The obtained blends, as well as the blends obtained in example 4, were investigated in a scanning electron microscope (SEM). The PA6 phase could be identified as interspaced glo- bules in a PP matrix. The sizes of the PA6 globules were measured and recorded. The results of the measurements of average diametres of PA6 particles in PP matrix for two poly- merisation systems are presented in table 1 below.

From Table 1 it is seen that the average diameters of the PA6 globules depend on the amount of PP-g-TMI in the reaction mix- ture, and they decrease with an increasing amount of PP-g-TMI.

As a rule, when the diameter of the dispersed globules is about 1 um or less, the composition is regarded as being suficiently compatibilized.

TABLE 1

Average diameter of PA6 globules (pm) vol% PP-g-TMI Polym. system I w Polym. system II z ; 0 14 100.0 2 5.8 4.4 10 1.5 20 1.0- 30 1.0 40 0. 6- Polym. system I: 30 % of e-caprolactam + 70 % of combined PP and PP-g-TMI <BR> <BR> 2) Polym. system II: 50 % of e-caprolactam + 50 % of combined PP and PP-g-TMI Example 6 (Comparative) To a 40 g sample of the polymeric blend obtained in Example 2 was added 10 g of the PP-g-TMI (containing 1.8 phr of TMI) obtained in example 1, and the mixing was performed under the same conditions as used in Example 2 for 10 minutes. The mor- phology of the final polymer blend had not changed from that observed with the mechanical blend of example 2.

Example 7 The purpose of this example is to verify that the e-capro- lactam polymerises from the TMI group of the PP-g-TMI.

The polymerisation procedure of Example 3 was followed. The weight ratios between PP, PP-g-TMI, e-caprolactam, catalyst and activator were as specified in Table 2. In all experiments the reaction mixture comprised 50 phr of e-caprolactam, 3.0 phr of catalyst and 3.0 phr of activator, while the amounts of PP homopolymer and the grafted PP-g-TMI were 50 and 0 phr; 48 and 2 phr; 30 and 20; 0 og 50 phr by weight, respectively. In the experiments the PP-g-TMI contained 1.8 % by weight. The obtained polymer compositions were solublized in xylene and formic acid, and the insoluble part determined.

The copolymer formation was quantified by solvent extraction.

PP is soluble in hot xylene and PA6 in formic acid. For a mechanical PP/PA6 blend, it is found that the PP and PA6 phases can be solubilized completely in hot xylene and formic acid, respectively. When the compatibilized PP/PA6 blend was subjected to successive treatments in hot xylene and formic acid, there remained an insoluble rest. This rest was not crosslinked because it had the behaviour of a thermoplastic resin. This rest was a graft copolymer of PP and PA6. The amounts of insolubles in various PP/PA6 (50/50) blends are presented in Table 2. Note that these blends differs only in the ratio between PP and PP-g-TMI. Clearly, a higher amount of PP-g-TMI resulted in a greater insoluble part. When the entire PP phase is composed of PP-g-TMI, the copolymer formed amount- ed to 16.1 % by weight, based on the weight of the entire final composition.

TABLE 2 Ex. PP-g-TMI Insoluble phr phr wt% 7. 1 50 0 0 7. 2 48 2 2. 4 7. 3 30 20 12. 1 7. 4 0 50 16. 1 Example 8 The blends obtained in examples 2 and 3, as well as a PA6 homopolymer, were analyzed with differential scanning calori- metry (DSC) and the heating and cooling curves recorded. Heat- ing and cooling rates were always 10 °C/min. The resulting curves are presented in figure 4. They show that during the heating stage (see fig. 4a) both PP/PA6 mechanical blends and the PP-g-PA6 copolymer display two distinct peaks at 161 °C and 212 °C, which correspond to the melting temperatures of PP and PA6, respectively. Upon cooling (see fig. 4b), the PP/PA6 mechanical blend has two distinct peaks at 101 °C and 160 °C, respectively. They correspond to the recrystallization tem- peratures of PP and PA6, respectively. However, the PP-g-PA6

copolymer of example 3 exhibits only one peak at 101 °C. The peak at 160 °C was absent. This is typical of a compatibilized PP/PA6 blend and can be viewed as a proof of PP-g-TMI having acted as an precursor to form the graft copolymer of PP-g-PA6.

Examples 9 to 15 Blends of PP/PA6 were produced as described in example 3 using various amounts of PP-g-TMI. The PP-g-TMI used contained 0.8 phr TMI. The amounts of PP, PP-g-TMI and £-caprolactam used are specified in table 3. In examples 14 and 15 no PP-g-TMI were used and consequently these examples fall outside the scope of the invention.

The obtained preblends were extruded in a single screw extrud- er at 230 °C and a mixing speed of 100 rpm. Specimens from the extruded article were subjected to tensile tests. The test data are presented in table 3.

TABLE 3 Elongation Example PP PP-g-TMI e-CL E-modulus at break No. (wt%) (wt%) (wt%) (MPa) (%) 9 40 30 30 1250 180 10 20 30 50 1430 23 11 0 50 50 1120 265 12 20 20 60 905 200 13 10 30 60 890 220 14 50 50 980 13 comp. 15 100 0 0 1100 200 comp.