Metal catalysts comprising nitrogen-containing aromatic ligands and use thereof

The present invention relates to the field of living radical polymerization and of metal catalysts with nitrogen- containing ligands for use therein.

Radical polymerization processes have witnessed a great progress in the past 50 years along with the development of organometallic catalysts. In general, transition-metals catalyse these reactions at low temperatures, working at high rates and tolerating protic compounds such as water. Ruthenium compounds are amongst the most studied versatile catalysts employed in radical polymerizations. Their efficiency has been attributed to the variety of oxidation states that ruthenium can assume during catalytic processes and the relatively low redox potential between its main oxidation states. This explains the variety of proficient ruthenium-catalysts carrying various ligands that have been employed in polymerization reactions. The first ruthenium-containing complexes used were of type [RuCl ₂ (PR ₃ ) _n ] and ever since, a continuous development aims at controlling their activity, stability and chemoselectivity (see for instance Kamigaito et al, 2001, Chem. Rev. 101, 3689-3745) . This is normally done by finding a suitable equilibrium between the electronic and steric environment around the metal. In this way, a great variety of ligands has been studied, among which phosphines, Schiff bases and half-metallocene complexes stood out as the most efficient ones. Nitrogen-containing ligands containing pyridine groups constitute the most prolific class of ligands utilised in copper-assisted radical polymerization (see Kamigaito et al, 2001). The diverse types of functionalities tested exemplify the effort to increase the solubility of the catalysts in an attempt to enhance the rates and molecular weight distributions of the polymers. For instance, the alkyl chains attached to the pyridine rings have been fine-tuned: by testing shorter, longer, linear, branched and altered chains, the rates have been improved significantly.

In an effort to upgrade the control of the growing polymer chain, the influence of the solvent in emulsion systems was studied. In copper-catalysed processes, CuX2(bpyR2) catalysts were employed where the bpyR ₂ ligand has long chains. In this way, the catalyst is retained in the organic phase, and polymers with good polydispersities were obtained. In addition, the combination of these ligands with non-ionic surfactants resulted in polymers with controlled chain growths and narrow molecular weight distributions. Collomb-Dunand-Sauthier et al, 1996 (J.

Electroanalytical Chemistry, 410, 21-29) disclose the electrochemical behaviour of three Ru(II) complexes with substituted 2, 2' -bipyridine ligands in a non-coordinating solvent. However, nothing is said about the use of these compounds as a catalyst for polymerization.

The present invention discloses the use of transition metal catalysts containing nitrogen-containing aromatic ligands for living radical polymerization. These catalysts enable the synthesis of polymers with high molecular weights and low polydispersities. Contrary to (copper) catalysts in emulsions, the catalysts as described herein do not require the presence of a surfactant. In addition, the polymerization reaction does not require an initiator.

In a first aspect, the present invention discloses a method for the preparation of a polymer from vinyl group- containing monomers comprising the step of polymerizing the vinyl group-containing monomers in the presence of a metal catalyst for living radical polymerization having a moiety M ²⁺ (Y) ₄ (X) ₂ wherein M is a transition metal selected from the group of Ru, Mn, Fe, Co, Zr, Rh and Os, each Y independently is a nitrogen-containing aromatic mono-ring ligand, at least one ligand Y being substituted on at least one position with a carbon-containing substituent and, optionally, either two or four ligands Y being covalently bonded to each other in pairs,, and X is chosen from H ₂ O, CH ₃ CN, (CH ₃ ) ₂ SO, NH ₃ , and a mono-, di- or tri (C ₁ -C ₄ ) -alkyl amine.

The transition metal M of the catalyst preferably is selected from the group of Ru, Mn, Co, Zr, Ru, Rh and Os, more preferably is Ru.

In the metal catalyst, the nitrogen atom of each nitrogen-containing aromatic mono-ring ligand Y is coordinated to the metal.

In one embodiment, two or four aromatic mono-rings are covalently bonded to each other in pairs. In this embodiment, the catalyst may comprise two pairs of covalently bonded rings or one pair of covalently bonded rings and two "free" rings.

In one embodiment, the nitrogen-containing aromatic ligand Y is a pyridine, at least one pyridine being substituted on at least one position with a carbon-containing substituent.

In a preferred embodiment, two pyridine rings are covalently bonded to each other, wherein the catalyst comprises one or two pairs of covalently bonded pyridine rings.

A covalent bond between two pyridine rings may involve one carbon atom of each pyridine ring, as in the structure C-B _n -C wherein the Cs are part of different pyridine rings, n may be 0 or 1, and when n is 1, B is C-D-C, wherein D is selected from C, Si, N or S. An example of two pyridine rings covalently bonded to each other is a dipyridine. Such a covalent bond preferably involves the carbon atom at the 2-position of each ring, such as a 2 , 2 ' -dipyridine, which is a preferred structure.

The covalent bond between two pyridine rings may also involve two carbon atoms of each pyridine ring, preferably the carbon atoms at the 2- and 3-position of each ring, for instance as in a phenantroline .

At least one nitrogen-containing aromatic ligand Y ring is substituted with at least one carbon-containing substituent. The substituent may be present at any carbon that does not participate in a covalent bond with an adjacent ring, with a maximum of two substituents per ring. It is also possible to have two substituents in one ring and none m an adjacent ring. Preferably, the substituent is present at the 4 position of a pyridine ring, more preferably at the 4 position of all pyridine rings . The carbon-containing substituent preferably is an aliphatic, heteroaliphatic or aliphatic-aromatic moiety that may be branched or straight, saturated or unsaturated and/or may contain a heteroatom like 0, S, Si and/or N, as an end group or within the chain or aromatic ring, and/or F as a substituent of

one or more hydrogen atoms of a hydrocarbon moiety. More preferably, the carbon-containing substituent comprises 1 to 32 carbon atoms. Even more preferably, the substituent is an alkyl group of 1 to 32 carbon atoms, most preferably of 10 to 20 carbon atoms, especially preferably an alkyl group having a straight chain.

Examples of suitable substituent configurations are shown below, as exemplified on dipyridine (L-I). The linear chain configuration as shown below is an example and may suitably be extended to a length as desired.

A preferred structure for two substituted covalently bonded ligands Y is 4, 4' -dialkyl-2, 2' -dipyridyl . The alkyl group preferably is a linear chain of 1 to 32 carbon atoms, more preferably of 10-20 carbon atoms.

The ligand X is chosen from H ₂ O, CH ₃ CN, (CH ₃ J ₂ SO, NH ₃ , and a mono-, di- or tri (Ci-C ₄ ) alkyl amine, such as diisopropylamine or triethylamine. Preferably, X is chosen from

H ₂ O, CH ₃ CN and (CH ₃ J ₂ SO, more preferably the ligand X is H ₂ O. The active catalyst with the ligand X may also be generated "in situ" by using a catalyst containing a halogen (F,

Cl, Br, I, especially Cl) as ligand, and replace it by the X ligand during the polymerization process. The reaction is immediate and the resulting polymer is not affected.

The counter ion A of the metal catalyst of the invention may be any suitable ion, for instance NO ₃ ^" , SO ₄ ^2" ,

CF ₃ SO ₃ ^" , CF ₃ COO ^" , BF ₄ ^" , SbF ₅ ^" , PF ₆ ^" . A preferred catalyst is a Ru catalyst comprising two dipyridyl ligands substituted at the 4 positions with a saturated alkyl group of 1 to 32 carbon atoms, preferably from

10 to 20 carbon atoms, and two H ₂ O molecules, the counter ion being NO ₃ ^" , such as [Ru (H ₂ O) ₂ (4, 4' -diheptadecyl-2, 2' -

dipyπdyl) ₂ ] (NO ₃ ) ₂ and [Ru (H ₂ O) ₂ (4 , 4 ' -ditπdecyl-2 , 2' - dipyπdyl) ₂ ] (NO ₃ J ₂ -

It is known that a way to improve the performance of a catalyst m controlled transfer radical polymerizations is by employing surfactants. Occasionally, this can lead to higher yields, but it mostly results in polymers of higher molecular weights and broader polydispersities .

Surprisingly, a catalyst with an apolar substituent at the pyridine ring advantageously may undergo self-aggregation and does not require the presence of a surfactant. Thus, such a catalyst acts both as a catalyst and as a surfactant. Aggregation of the catalyst may be influenced by various conditions, like solvent composition, solvent polarity, temperature, pressure, catalyst structure, in particular the nature of the substituent group. The detection of aggregates is done routinely with a dynamic light scattering setup.

The nature of the substituent group, in particular the chain length of a linear chain, advantageously may be varied to obtain the appropriate solubility and/or aggregation characteristics of the catalyst m the solvent of choice. A longer linear chain length, e.g. Ci ₇ or longer, enables the catalyst to form self-assembled aggregates m a solvent, e.g. toluene, styrene and/or hexane. An example of such a catalyst is [Ru(H ₂ O) ₂ (4, 4'-diheptadecyl-2, 2' -dipyπdyl) ₂ ] (NO ₃ ) ₂ . Catalysts with a shorter chain length, e.g. a linear Ci ₃ -tail or shorter, for example [Ru (H ₂ O) ₂ (4 , 4 ' -ditridecyl-2, 2 ' -dipyπdyl) ₂ ] (NO ₃ ) ₂ , do not self-assemble m e.g. toluene or styrene, but may do so in e.g. hexane.

Aggregated catalysts advantageously cause the formation of microemulsions . A favourable aspect of using a catalyst that self-aggregates is that surfactants will not be needed to form the desired microemulsions, thereby employing less chemicals and reducing chemical waste.

In microemulsions, polymers tend to grow m a controlled way until the polymer becomes too large to be confined to the emulsion droplet containing the catalyst. Typically, in microemulsions polymers of very high molecular weights can be obtained, because they promote the miscibility between catalyst and polymer. However, this miscibility often

may not be efficient and polymers may possess a very high polydispersity making them useless in various applications. Surprisingly, catalysts of the present invention both in emulsions and in homogeneous solutions enable the preparation of large polymers with a relatively low polydispersity.

Thus, the catalysts can work in homogeneous solutions as well as in microemulsions .

The catalyst as described herein advantageously allows the preparation of polymers that have a high molecular weight (up to 1.5xlO ⁶ Dalton) and a low polydispersity (between 1.2 and 1.9) and that are atactic.

The vinyl group containing monomers may consist of a single monomer, resulting in the formation of a homopolymer, or may contain a mixture of at least two different monomers, to form a copolymer.

Suitable monomer compound containing a vinyl group to be used in the method of the invention are acrylate, methacrylate and/or styrene monomers. Single monomers or mixtures of two or more monomers may be used. Examples of suitable acrylate, methacrylate and styrene momoners are given below .

Acrylates :

Methacrylates :

Styrenes :

The polymerization reaction, which is due to radical transfer, may be performed in the presence or absence of a radical initiator. Although the polymerization also functions when an initiator (for example CHI ₃ , CCl ₃ Br, or Cl ₃ CCOOMe) is added, its presence may slow down the polymerization reaction. Therefore, the polymerization using the catalyst as described herein works more efficiently in the absence of an initiator, which is beneficial since initiators are in general toxic compounds. This aspect contradicts the copper-catalyzed process known in the art, where initiators are always necessary. In particular, complexes of the [Ru (bpyR2) 2 (H∑O) 2] ²⁺ type as presented here are excellent catalysts for radical transfer polymerization not needing the aid of initiators. Thermal initiation, although slower, is sufficient to conduct the reaction.

The polymerization reaction may be carried out in a pure liquid form of the polymerizable monomer (bulk solution or neat substrate), for instance a styrene liquid, or in a solution of the polymerizable monomer in a suitable solvent, like toluene. Typically, polymer molecular weights will be higher when reactions are performed in bulk solution with respect to solutions where monomer is diluted in solvent.

The reaction works in the presence of O ₂ as well as under a inert gas (N ₂ ) atmosphere.

The polymerization may be performed in a range of temperatures between 20°C and 15O ⁰ C, but is preferably performed at a temperature of 40°C to 130 ⁰ C. Typically, the temperature may be tuned to obtain either a high molecular weight polymer or a high conversion. In living radical polymerizations, working at a higher temperature results in an increment of the concentration of reactive species, which accounts for a higher conversion. An additional effect of the temperature in

controlled radical polymerizations is that the average molecular weight of a polymer decreases as the temperature increases.

A similar relation exists between catalyst concentration, conversion and molecular weight of the polymer. The higher the catalyst concentration, the higher the conversion and the lower the molecular weight. Typically, the catalyst concentration may vary between 10 ^~7 and 1(T ¹ M, but will preferably range between 10 ^~5 and 10 ^~3 M.

The monomer concentration m solvent may have values as low as 10 ^~3 M, but polymerization is preferably carried out m neat substrate.

The polymerization reaction may proceed from 1 up to 72 hours, preferably between 2 and 48 hours.

After all monomers in a solution are polymerized, the catalyst is still alive. This means that when fresh monomers are added the polymerization will continue.

The catalyst as described herein advantageously allows the preparation of polymers with molecular weights (M _w ) reaching about 1.5xlO ⁶ Dalton. These high molecular weights are overwhelmingly higher than typical molecular weights found in common living radical polymerization, which range between 50,000 and 80,000. Molecular weights ranging from about 50,000 up to 1.5xlO ⁶ Dalton may be reached depending on the concentration of the catalyst and the polymerization temperature. Molecular weights of resulting polymers typically may be analyzed by Size Exclusion Chromatography coupled to MultiAngle Laser Light Scattering (SEC-MALLS) .

The polydispersity of the polymers varies between 1.2 and 1.9 for molecular weights between 40,000 and 1.5xlO ⁶ Dalton. Since the catalyst promotes polymerization in the absence of any solvent, specific shapes that are difficult to process may be made by direct polymerization in the mold. After polymerization of all monomers and thus hardening of the obtained shape, the catalyst can be easily washed out and reused.

In addition, the catalyst may be used to polymerize monomers m nano-porous materials, in which one phase has to be hardened. Such nano-porous, polymeric material can be used for

various applications in bio-compatible materials, molecular filters and nano-structured templating materials.

In a second aspect, the present invention provides a novel metal catalyst for living radical polymerization having a moiety Ru ²⁺ (Y) ₄ (X) ₂ wherein each Y independently is a nitrogen- containing aromatic mono-ring ligand, at least one ligand Y being substituted on at least one position with a carbon- containing substituent and, optionally, either two or four ligands Y being covalently bonded to each other in pairs, , and X is chosen from H ₂ O, CH ₃ CN, (CH ₃ J ₂ SO, NH ₃ , and a mono-, di- or tri (C ₁ -C ₄ ) -alkyl amine, with the proviso that the metal catalyst is not Ru ²⁺ (L ¹ ) ₂ (OH ₂ ) ₂ , Ru ²⁺ (L ² ) ₂ (OH ₂ ) 2 and Ru ²⁺ (L ³ ) 2 (OH ₂ ) 2, wherein L ¹ , L ² and L ³ have the structures as follows:

Further options for Y and X and for the counter ion are as detailed for the catalysts as described in the first aspect of this invention. In a third aspect, the present invention provides a method for preparation of the catalyst as described herein.

The catalyst conveniently may be prepared by a combination of known methods.

The ligand Y may be reacted with a metal halide to obtain a metal-ligand Y complex. The metal complex then is

digested with silver ions and a compound for occupation of the X position is added, to obtain the metal catalyst. The metal catalyst may be recovered from the reaction mixture by any method know to the person skilled m the art. A ligand substituted with alkyl chains in the longer range may be synthesized from a precursor ligand substituted with short alkyl groups (methyl, ethyl) by treating this precursor with a lithiating agent followed by reaction with the appropriate halogenated precursor to obtain substituents with the desired size. For instance, see Ellison and Iwamoto (1983), Tetrahedron Lett. 24, 31 for the synthesis of symmetric dipyridmes and G. Chelucci and R. P. Thummel (2002), Chem. Rev. 102, 3129-70 for the synthesis of asymmetric dipyridmes. Alternatively, a ligand substituted with an -OH group may be reacted with a suitable halogenated precursor m the presence of a base, for instance see A. Bellusci et al. (2005), Inorganic Chemistry 44(6), 1818-25 and J. Skarzewski et al (1979) Heterocycles 12, 1403-6.

In particular, a suitable ligand with small alkyl groups is 4 , 4 ' -dimethyl-2, 2' -bipyπdyne, which may be reacted with lithium dnsopropylamide at -70°C, followed by addition of the appropriate halogenated substituent precursor, e.g. 1- bromohexadecane .

The metal halide for reaction with the ligand Y may be MQ _n (n = 1, 2 or 3) (Q = Cl, Br or I), preferably Q = Cl, more preferably the metal halide is MCl ₃ . The solvent for the reaction is a polar solvent (such as water, methanol, ethanol, ethyleneglycol, DMF, DMSO) . The reaction may occur at a temperature of 60 to 140 ⁰ C for a period of 8 to 48 hours. The obtained ligand Y-metal complex is treated m analogy to a procedure reported by Chi-Mmg Che and Wai-Ho Leung (1987), J. Chem. Soc, Chem. Commun., 1376-7. Briefly, it is digested with a silver salt, wherein the counter ion of the silver cation suitably defines the counter ion for the catalyst. Preferably, a stoichiometric amount of silver salt is used. Then, a suitable compound is added that defines the ligand present in position X of the catalyst of the invention.

The obtained catalyst may be recovered by filtering the mixture through a plug of celite and evaporating volatiles on a

rotavapor. The resulting solid may be dried under vacuo for 24 hours .

Brief description of the Figures Figure 1. Conversion vs. catalyst concentration in the polymerization of styrene at T = 7O ⁰ C with 2a (A) and 2b (■) .

Figure 2. Polystyrene molecular weight vs. log

(catalyst ^' concentration) in the polymerization of styrene at T = 50 ⁰ C with 2a (A) and 2b (■) . Figure 3. Polystyrene molecular weight vs. log

(catalyst concentration) in the polymerization of styrene at T = 7O ⁰ C with 2a (A) and 2b (■) .

Figure 4. Conversion vs. temperature in the polymerization of styrene at [cat] = 10 ^"4 M with 2a (A) and 2b (■) .

Figure 5. Polystyrene molecular weight vs. temperature in the polymerization of styrene .at [cat] = 10 ^~4 M with 2a (A) and 2b (■) .

Experimental

Polymer analyses. Polystyrenes were analysed using SEC- MALLS with IR detection to perform absolute Mw determination using ASTRA software (version 4.50, Wyatt Technology Corp., Santa Barbara, CA) . The light scattering detector was a DAWN DSP-F (Wyatt Technology Corp., Santa Barbara, CA), refractive index detector Shimadzu RID-IOA (Shimadzu, 's Hertogenbosch, The Netherlands) . The column set for SEC consisted of three 300x7.5mm Mixed-B PL gel columns in series (Polymer Laboratories, Church Stretton, Shropshire, UK) . Unstabilised tetrahydrofuran from Biosolve (Valkenswaard, The Netherlands) was used as the mobile phase.

General. All synthetic experiments with metal complexes and dipyridine ligands were performed under an atmosphere of dry nitrogen using standard Schlenk techniques. All non-deuterated solvents, reagent grade or better were dried according to standard previous procedures under an atmosphere of nitrogen. Deuterated solvents were used as received. Commercially available reagents were used as received, except styrene, which

was dried by three consecutive freeze-pump-thaw cycles and stored under N ₂ atmosphere at 4 ⁰ C.

¹ H and ¹³ C( ¹ H) NMR spectra were recorded at 300.13 and 75.48 MHz, respectively, using a Varian Unity 300 MHz spectrometer. Unless otherwise stated, the NMR measurements were performed in CDCl ₃ . ¹ H and ¹³ C( ¹ Hl NMR are reported in parts per million high frequency from an external reference of tetramethylsilane.

Example 1

Synthesis of ruthenium complexes 4,4' -ditridecyl-2,2' -dipyridine (a) and 4,4'- diheptadecyl-2 ,2' -dipyridine (b) . We decided to use 4,4'- (dialkyl) -2, 2' -dipyπdyl ligands for the synthesis of ruthenium- based catalysts where alkyl stands for H-Ci ₃ H _2S and ω-C ₁₇ H ₃₅ . The synthesis of these ligands (a and b) proceeds smoothly m good yields at low temperatures (-78 ⁰ C) using lithium diisopropylamide as lithiating agent, followed by 1- bromododecane (or hexadecane) while keeping the temperature low. 4,4' -ditridecyl-2,2' -dipyridxne (a). A 1.6 M solution of n-butyllithium m hexane (5.12 ml, 8.2 mmol) was added dropwise to a suspension of diisopropylamme (1.13 ml, 8.0 mmol) in THF (5 ml) at -78 ⁰ C and allowed stirring for 30 mm. At - 78°C, 4, 4' -dimethyl-2, 2' -dipyridine (0.73 g, 4 mmol) in THF (30 ml) was slowly (30 mm) added and stirred for additional 2 h. The suspension was warmed to 0 ⁰ C, 1-dodecylbromide (2.63 g, 10 mmol) m THF (5 ml) added. After stirring for one hour at 0°C, the solution was allowed to reach room temperature while stirring for 16 hours. Methanol (10 ml) was added and the solution stirred for 10 minutes to quench it. Water (100 ml) was added and the organic phase extracted with diethylether (3 x 100 ml) . The combined organic fractions were dried over Na ₂ SO ₄ , filtered and the solvent evaporated. The product can be purified upon recrystallisation m pentane to obtain an off-white powder which was filtered and dried under vacuum. Yield: 1.44 g (69%) .

¹ H NMR (CD ₂ Cl ₂ ): δ 0.86 (t, 6H, ² J _HH = 6.6 Hz CJf ₃ (CH ₂ ) ₁ 2) , 1.30-1.70 (br m, 44H, CH ₃ (CW ₂ ) 11) , 2.70 (t, 4H, ² J _HH = 8.1 Hz, CW ₂ Ar), 7.16 (dd, 2H, ² J _m = 4.8 Hz, ⁴ J _HH = 1.5 Hz, H*-5,5'), 8.26 (d, 2H, ² J _HH = 0.9 Hz, H*-3,3'), 8.56 (dd, 2H, ² J _HH = 4.8 Hz, ⁴ J _HH =

0.6 Hz H*-6,6'). ¹³ Ci ¹ H): δ 14.34 (CH ₃ ), 22.91 (CH ₃ CH ₂ ), 29.55 (CH ₂ CH ₂ CH ₂ Ar), 29.87 (CH ₃ (CH ₂ ) ₂ (CH) ₆ ) , 30.66 (CH ₂ CH ₂ Ar), 32.13 (CH ₃ CH ₂ CH ₂ ), 35.79 (CH ₂ Ar), 121.81 (C5), 124.30 (C3), 148.91 (C6), 153.66 (C4), 155.78 ( C2 ) . 4 ,4' -dxheptadecyl-2 ,2' -dipyridme (b) . A 1.6 M solution of n-butyllithium in hexane (6.75 ml, 10.8 mmol) is added dropwise to a suspension of dusopropylamine (1.48 ml, 10.5 mmol) m THF (5 ml) at -78°C, warmed up to 0 ⁰ C and allowed stirring for 30 mm. A solution of 4 , 4' -dimethyl-2, 2' -dipyridme (0.73 g, 4 mmol) in THF (30 ml) was then dropwise added and stirred for additional 2 h. followed by addition of 1- hexadecylbromide (3.06 g, 10 mmol) m THF (5 ml) . The resulting suspension was stirred for additional 1 h at 0 ⁰ C. The bath was then removed and allowed reaching room temperature overnight. Upon addition of methanol (10 ml) the reaction was quenched, water (100 ml) was added and the organic phase extracted with diethylether (3 x 100 ml). The combined organic fractions were dried over MgSO ₄ , filtered and volatiles removed by rotary evaporation. Resulting combination of organic layers was dried with anhydrous Na ₂ SO ₄ , filtered, evaporated and the residue recrystallised from heptane to obtain an off-white powder which was filtered and dried under vacuum. Yield: 59.3 g (73%) .

¹ H NMR: δ 0.86 (t, 6H, ² J _m = 6.9 Hz, CH ₃ (CH ₂ ) _I6 ), 1.30- 1.70 (br m, 6OH, CH ₃ (CH ₂ )Is), 2.68 (t, 4H, ² J _m = 7.8 Hz, CH ₂ Ar), 7.13 (dd, 2H, ² J _HH = 5.1 Hz, ⁴ J _HH - 1.5 Hz, H*-5,5' ) , 8.23 (d, 2H, ² JHH = 0.9 Hz, H*-3,3' ) , 8.55 (dd, 2H, ² J _HH = 4.8 Hz, ⁴ J _H H = 0.6 Hz, H*-6,6' ) - ¹³ C( ¹ H) δ 14.36 (CH ₃ ) , 22.93 (CH ₃ CH ₂ ) , 29.93 (CH ₃ (CH ₂ ) ₂ (CH) ii) , 30.68 (CH ₂ CH ₂ Ar) , 32.15 (CH ₃ CH ₂ CH ₂ ) , 35.79 (CH ₂ Ar) , 121.72 (C5) , 124.26 (C3) , 149.01 (C6) , 153.52 (CA) , 155.96 (C2) .

[Ru (4, 4' -ditridecyl-2,2'-dipyridine) ₂ Cl ₂ ] (Ia) and [Ru(4,4' -diheptadecyl-2,2' -dipyridine) ₂ C1 ₂ ] (Ib). Reacting RuCl ₃ with two equivalents of the corresponding ligand in DMF at 120 ⁰ C results m the RuCl ₂ (bpyR ₂ ) ₂ complexes Ia and Ib (scheme 1). These complexes have been purified by means of size-exclusion chromatography. The use of stoichiometric amounts of ligand is crucial m order to avoid the formation of side products containing an undesired number of ligands that cause ma]or decreases of the yields.

[Ru(4 _f 4'-d-Ltridecyl-2,2 ^/ -dipyrxdine)2Cl ₂ ] (Ia). A mixture of RuCl ₃ -3H ₂ O (83.7 mg, 0.32 mmol), 4 , 4 ' -di (tridecyl) - 2,2'-dipyridine (0.333 g, 0.64 mmol) and LiCl (80.5 mg, 1.9 mmol) in DMF (10 ml) was heated to 110 ⁰ C during 6 h. The reaction was cooled down, acetone (30 ml) added and the solution was further cooled to -2O ⁰ C. The precipitate formed was filtered off and washed with water (3 x 20 ml) . The residue was purified twice on Sephadex LH20 with a acetonitπle/isopropanol (1:9) mixture to obtain Ia as a purple solid. Yield 0.32 g (82%) . ¹ H NMR gave broad signals. ¹ H NMR: δ 0.88 (m, 12H,

CE ₃ (CH ₂ )I ₂ ), 1.18-1.80 (br m, 88H, CH ₃ (CH ₂ ) u), 2.89 (br s, 8H, CJf ₂ Ar), 7.20-7.30 (br m, 4H, H-5,5'), 7.54-7.61 (br m,, 4H, H- 3,3'), 8.33 (br s, 4H H-6, 6' ) . (FAB, ⁺ ): m/z = 1141.93 [M-2C1 ^" ] ⁺ . Anal. Calcd for C ₇₂ H ₁₂₀ Cl ₂ N ₄ Ru: C, 71.25; H, 9.97; N, 4.62. Found: C, 70.78; H, 9.66; N, 4.88.

[Ru(4,4'-diheptadecγl-2,2' -dxpyridine) ₂ C1 ₂ ] (Ib). A mixture of RuCl ₃ -3H ₂ O (63 mg, 0.24 mmol), 4 , 4 ' -diheptadecyl-2 , 2 ' - dipyridme (0.305 g, 0.48 mmol) and LiCl (59.35 g, 1.4 mmol) in DMF (10 ml) was heated to HO ⁰ C during 8 h. The reaction was cooled down; acetone (30 ml) added and then cooled further to - 20 ⁰ C. The precipitate formed was filtered off and washed with water (3 x 20 ml) . The residue was purified twice on Sephadex LH20 using lsopropanol/acetonitπle mixture (9:1) to obtain Ib as a purple solid. Yield 0.283 g (82%). ¹ H NMR gave broad signals. ¹ H NMR: δ 0.87 (m, 12H,

CH ₃ (CH ₂ ) ₁₆ ), 1.20-1.80 (br m, 120H, CH ₃ (CH ₂ ) ₁₅ ) , 2.88 (t, 8H, ² J _HH = 7.9 Hz, CH ₂ Ar), 7.20-7.30 (br m, 4H, ² J _m = 6.6 Hz, H*-5,5'), 7.52-7.60 (br m,, 4H, H3,3'), 8.31 (br s, 4H H6,6'). (FAB, ⁺ ): m/z = 1522.16 [M-2C1 ^" ] ⁺ . Anal. Calcd for C ₈₈ Hi ₅₂ Cl ₂ N ₄ Ru: C, 73.49; H, 10.65; N, 3.90. Found: C, 73.40; H, 10.54; N, 3.83.

[Ru (4, 4' -dxtrxdecyl-2,2' -dipyridine) ₂ (H ₂ O) ₂ ] (NO ₃ J ₂ (2a) and [Ru(4,4'-dihexadecyl-2,2'-dxpyrxdxne) ₂ (H ₂ O) ₂ ] (NOa) ₂ (2b). We subsequently reacted Ia and Ib with a stoichiometric amount of AgNO ₃ , first at a low (-5 to 10 °C) and then at a mild (20 to 60 ⁰ C) temperature, and then added traces of water, to replace the chlorides by solvent (water) molecules and weakly coordinating anions (NO ₃ ), thereby generating charged complexes 2a and 2b. As a result, these complexes can self-aggregate in apolar solvents

and their reactivity is improved at the same time, because the water ligands will dissociate better than the chlorides.

Compound 2a: MS (FAB ⁺ ) : m/z = 1240,91 [M-NO ₃ ] ⁺ , 1223,93 [MH-H ₂ O-NOs] ⁺ . Anal. Calcd. for C ₇₂ Hi ₂₄ N ₆ O ₈ Ru: C, 66.37; H, 9.59; N, 6.45. Found: C, 66.64; H, 9.71; N, 6.33. Compound 2b: MS (FAB ⁺ ) : m/z = 1465.16 [M-NO ₃ J ⁺ , 1448.17 [MH-H ₂ O-NO ₃ ] ⁺ . Anal. Calcd. for C ₈₈ Hi ₅₆ N ₆ O ₈ Ru: C, 69.20; H, 10.30; N, 5.50. Found: C, 68.95; H, 10.17; N, 5.37.

Example 2

Polymerization of styrene

Styrene polymerizations were carried out in a closed Schlenk vessel under N ₂ atmosphere and can advantageously work in the presence of O ₂ . The appropriate amount of catalyst was dissolved in degassed styrene (final volume, 3 ml) and immediately placed in a preheated oil bath.

Tables 1 and 2 collect the performance of catalysts 2a and 2b, respectively, in the polymerization of styrene. The yields of polystyrene obtained are good, ranging from 50-99%, improving as temperatures rise. The most remarkable aspect is that the polymers produced have high molecular weights (> 10 ⁶ ) .

Table 1. Polymerization of styrene with 2a ^a results

Entry T ( ⁰ C) cone. time conv

M _w ^c MjM _n ^c (M) (h) (%) ^b

1 50 10 ^"3 24 48 220 000 1.6

2 50 lO ^"4 24 33 960 000 1.6

3 50 5 x 10 ^~5 24 30 1 100 000 1.6

4 50 lo- ⁵ 24 23 1 320 000 1.6

5 60 10 ^"4 24 50 470 000 1.6

6 70 10 ^"3 24 96 170 000 1.6

7 70 5 x lO ^"4 24 78 210 000 1.6

8 70 lo- ⁴ 24 61 390 000 1.6

9 70 5 x 10 ^"5 24 53 450 000 1.5

10 70 2 x 10 ^~5 24 49 540 000 1.6

11 70 lo- ⁵ 24 47 630 000 1.7

12 90 lo- ⁴ 12 64 100 000 1.6

13 90 lo- ⁴ 24 84 120 000 1.7

14 110 10 ^"4 4 53 60 000 1.5

15 110 10 ^"4 24 98 70 000 1.6

16 ^d 50 lo- ⁴ 24 24 250 000 1.6

17 ^e 70 10 ^"4 24 35 120 000 1.8

18 ^f 70 lO ^"4 24 51 150 000 1.9

19 ^g 70 ₁₀ -4 24 7 40 000 1.8 ^a> All reactions carried out in neat styrene unless otherwise stated. ^b) From GC, using hexadecane as IS. ^c) From SEC chromatography. ^d) styrene: toluene 1:1 (v/v) . ^e) With CCl ₃ Br as initiator, 40 mM. ^f> With Cl ₃ CCOOCH ₃ as initiator, 40 mM. ^g) With CHI ₃ as initiator, 40 mM.

Table 2. Polymerization of styrene with 2b ^a results

Entry T ( ⁰ C) cone. time conv

M _w MJM _n (M) (h) ⁽ % ⁾

1 50 io- ^J 24 54 180 000 1.6

2 50 10 ^~4 24 38 680 000 1.6

3 50 5 x 10 ^"5 24 35 920 000 1.6

4 50 10 ^~5 24 25 1 500 000 1.6

5 60 10 ^~4 24 57 330 000 1.6

6 70 lo- ³ 24 81 145 000 1.6

7 70 5 x 10 ^~4 24 78 210 000 1.6

8 70 10 ^~4 24 67 240 000 1.5

9 70 5 x 10 ^~5 24 61 300 000 1.5

10 70 2 x 10 ^"5 24 56 440 000 1.6

11 70 10 ^~5 24 49 610 000 1.7

12 90 10 ^"4 12 64 100 000 1.6

13 90 10 ^"4 24 90 105 000 1.7

14 110 lo- ⁴ 4 58 40 000 1.5

15 110 10 ^"4 24 99 50 000 1.6

16 ^d 50 lO ^"4 24 25 260 000 1.4

17 ^e 70 lo- ⁴ 24 35 120 000 1.8

18 ^f 70 io ^~4 24 51 150 000 1.9

19 ^g 70 lo- ⁴ 24 6 40 000 1.8 a) All reactions carried out in neat styrene unless otherwise stated b) From GC, using hexadecane as IS. ^c) From SEC chromatog- raphy. ^d) styrene: toluene 1:1 (v/v) . ^e) With CCl ₃ Br as initiator, 40 mM. ^f) With Cl ₃ CCOOCH ₃ as initiator, 40 mM. ⁹⁾ With CHI ₃ as initiator, 40 mM.

In contrast to conventional systems, it is surprising that reagents that are normally used as initiators do not improve the performance of the reactions (entries 17-19) , they seem to retard the reaction (entries 17 and 18) and in some cases even to deactivate (entries 19) the catalytic performances of the catalysts. These initiators are generally toxic and it is advantageous that a method can work in the absence of these chemicals.

Figure 1 depicts the relationship between the catalyst concentration and the polystyrene conversion. Clearly, the

conversions improve as the catalyst's concentration is increased. Higher conversions are obtained with more concentrated samples, as also seen in entries 1 to 4 as in living radical polymerizations. Figures 2 and 3 show the evolution of the molecular weights of the polymers obtained at different concentrations and at constant temperature. The polymers with the highest molecular weights are obtained at the lowest catalyst concentrations. A drawback of operating at low concentrations is that the polymerization rates are slower

Assuming again that the molecular weight is proportional to the sguare root of the concentration of the catalyst, then molecular weights will be higher when reactions are performed in bulk solutions with respect to solutions where styrene is diluted. This is indeed the case as evident from entry 16; lower molecular weights are achieved m toluene/styrene systems with respect to reactions under bulk conditions. Therefore, good results are achieved when working in neat substrates, but the system still works when monomer is diluted.

As deducted from tables 1 and 2, conversions are higher when reactions are carried out at high temperatures. This relationship between conversions and reaction temperatures at constant concentrations is depicted in figure 4. According to the mechanism followed by controlled radical transfer polymerization reactions, the temperature plays an important role as it can be tuned to obtain either high molecular weight polymers or high conversions. If the catalyst concentration is kept constant, conversions should increase as the reaction temperature is raised. Indeed, as expected for radical transfer polymerizations, the performance of our catalysts improves at higher temperatures.

An additional effect of the temperature in controlled radical polymerizations is that the average molecular weight of the polymers decreases as the temperature increases. As depicted in figure 5, this is indeed the case: lower molecular weights are obtained with both catalysts when higher temperatures are employed.

However, one can establish a certain influence of aggregates at high concentrations. The essential difference between solutions of 2a and 2b in neat styrene is that only the latter undergoes self-assembly in solution and these aggregates are likely to influence the outcome of the reaction. This difference can be seen in figure 4. Such a representation is typically a linear decay in conventional (homogeneous) systems, which is the case when 2a is used as catalyst. On the contrary, solutions of 2b show an exponential decay. This means that at high concentrations, metallo-aggregates of 2b favour the formation of smaller polymers with respect to solutions of 2a. Therefore, the performance of the catalysts in view of their employment in homogeneous systems or emulsions, is different.

The polymerization reactions can thus be carried out in emulsions or homogeneously, as desired, in order to tune the molecular weights of the polymers when working at catalyst concentrations between 10 ^~5 and ICT ³ . The catalyst concentrations employed are lower than those typically employed in living radical polymerization protocols. All in all, our catalysts are good for polymerization reactions and follow the atom-transfer radical mechanism typical of living polymerizations. The polydispersities obtained are somewhat higher. However, the catalyst loads employed are lower, the molecular weights obtained are very high compared to traditional protocols and the catalysts do not need the assistance of toxic initiators.

Example 3

Polymerization of methylmetacrylate and butylacrylate Polymerizations in neat methylmetacrylate and neat butylacrylate with catalyst 2a were carried out according to Example 2, using a catalyst concentration of 10 ^"4 M and a temperature of 50 °C. The molecular weights of the resulting polymers as measured using SEC chromatography were 1 130 000 for polymethylmetacrylate and 2 250 000 for polybutylacrylate .