FORM 2 THE PATENTS ACT, 1970 (39 of 1970) The Patent Rules, 2003 PROVISIONAL SPECIFICATION (See section 10 and rule 13) TITLE OF THE INVENTION "A PROCESS FOR PREPARING ATOMOXETINE HYDROCHLORIDE AND (±) ATOMOXETINE OXALATE SALTS" We, CADILA HEALTHCARE LIMITED, a company incorporated under the Companies Act, 1956, of Zydus Tower, Satellite Cross Road, Ahmedabad-380015, Gujarat, India. The following specification describes the invention: Field of Invention The present invention relates to the process for preparing Atomoxetine hydrochloride. More particularly, the invention relates to crystalline form of N-methyl-3-phenyl-3-(o-tolyloxy) propylamine oxalate (here in after referred as "(+/-)atomoxetine oxalate"), which is an useful intermediate for the synthesis of atomoxetine hydrochloride. Atomoxetine HC1 is marketed as STRATTERA®, which is a selective norepinephrine reuptake inhibitor. Designation. Atomoxetine HC1 is chemically known as (-)-Ar-Methyl-3-phenyl-3-(o-tolyloxy)-propylamine hydrochloride and represented by below mentioned formula. H3C NH.HCI CH3 BACKGROUND AND DESCRIPTION OF PRIOR ART STRATTERA® capsules are intended for oral administration only. Each capsule contains atomoxetine HC1 equivalent to 10, 18, 25, 40, or 60 mg of atomoxetine. The capsules also contain pregelatinized starch and dimethicone. The capsule shells contain gelatin, sodium lauryl sulfate, and other inactive ingredients. The capsule shells also contain one or more of the following: FD&C Blue No. 2, synthetic yellow iron oxide, titanium dioxide. The capsules are imprinted with edible black ink. . The molecular formula is C n H 2i NOHC1, which corresponds to a molecular weight of 291.82. The chemical structure is: 10 H3C O Formula 1 U.S. Patent No. 4,314,081 describes 3-Aryloxy-3-phenyl polyamines, which possess central nervous system activity. Atomoxetine is a member of the above class of compounds, and is a useful drug for the treatment of depression. Atomoxetine was claimed in U. S. Patent No. 4,314,081 and the patent describes a process for the preparation of atomoxetine and related compounds in two different ways as depicted below as Scheme A and Scheme B, respectively. Scheme A Resolution OH H,CV H3C NH CH3 Atomoxetine Scheme B ^CH3 N Diborane CH, "" ^CH3 N 3 11 CH3 Resolution Atomoxetine The process illustrated in Scheme A involves the preparation of the atomoxetine using 3-phenyl chloropropyl amine (Formula 5) as a starting material. The process involves bromination of said starting compound (Formula 5) by using N-bromosuccinimide. Further the bromo derivative is condensed with o-cresol to result in a compound of Formula 7, which is then subjected to amination using methylamine. Though the process looks very simple, it involves the following disadvantages: ^O 3 (a) N-bromosuccinimide being a corrosive and sensitive chemical, its usage demands special care; (b) The work up of the compound formula 7 involves high vacuum (0.03 torr) distillation at 135-145°C, which is a tedious and cumbersome process to carry out at the plant level. (c) The reaction conditions involved in some of the steps are harsh, for example the amination reaction is conducted at 140°C. at pressures of 10 kg/cm2 for 12 hours in autoclave. All the above points make the process not viable for practicing on a commercial scale. On the other hand, Scheme B describes the preparation of atomoxetine using [beta]-dimethylaminopropiophenone produced by a Mannich reaction; which is reduced to the hydroxy derivative having Formula 9 using diborane; further the hydroxy compound (Formula 9) is converted to the corresponding chloro derivative of Formula 10 using dry HCI gas and thionyl chloride and is followed by condensation with o-cresol. The said reaction is carried out in methanol at reflux for a duration of five days to achieve the compound of formula 11 and is followed by demethylation using cyanogen bromide to end up with atomoxetine. As can be clearly understood the process is associated with the following problems: In scheme B, chloro compound (formula 10) treated with o-creosol condense to obtain compound (formula 11) then demethylation to obtain compound (formula 4) the several unite operation in as above process. One most disadvantage is that the using o-cresol commercial grade, o-cresol which contain phenol, p-cresol as impurity so in final API more purification are required. The use of costly reagents such as diborane makes the process uneconomical; The passage of dry HCI gas followed by thionyl chloride addition is very cumbersome and is not advisable in the plant. This is a time-consuming process, involving a reaction which requires five days for its completion; and iv) use of cyanogen bromide, which is highly toxic, is not desirable. Further, M. Srebnik et al., Journal of Organic Chemistry, Vol. 53, pages 2916-2920 (1988); E. Corey et al., Tetrahedron Letters, Vol. 30, pages 5207-5210 (1989); U.S. Patent No. 4,868,344; Y. Gao et al., Journal of Organic Chemistry, Vol. 53, pages 4081-4084 (1988); J. Deeter et al., Tetrahedron Letters, Vol. 31, pages 7101-7104 (1990); and U.S. Patent No. 4,950,791 disclose stereospecific methods for the preparation of 3-aryloxy-3-phenylpropylamines; the enantiomers of 3-hydroxy-3-phenylpropylamines are prepared by the stereospecific reduction of the corresponding ketones. The thus obtained (S)-3-hydroxy-3-phenyl propylamines are subjected to condensation with aryl alcohols using the Mitsunobo reaction as shown in below Scheme C, Scheme C OH Diisopinocamphenyl I II Diisopinocamphenyl 1 ^^^\^^\ chloroborane r^^r^^^^CI DEAD Triphnyl phosphine U 13 ^14 :JO H2NMe 1 ' ' CH3 NH < Atomoxetine The first critical step is an asymmetric reduction of the ketone to its corresponding alcohol. The second critical step involves the condensation of the obtained enantiomeric alcohol with the corresponding aryl alcohol. The process suffers from the following disadvantages: The reagent used for the asymmetric reduction of the ketone is highly expensive; The reagent diethyl azodicarboxylate ("DEAD") is expensive; The DEAD reagent is known to be highly carcinogenic, thus creating problems in handling; and the reaction involves the use of triphenylphosphine and DEAD and the resulting byproducts formed in the reaction, phoshineoxide and a hydrazine derivative, are very difficult to remove. Therefore, commercial applicability of the said process is limited owing to the above noted disadvantages. WO 00/58262 relates to a stereo- specific process for the preparation of atomoxetine using nucleophilic aromatic displacement of an aromatic ring having a functional group, which can be converted to a methyl group. As can be seen, the process is very lengthy and involves many steps and is thus not commercially desirable. OH Scheme D CH3^r^ CHn °Y> J^J