A facile organometallic-induced cross-linking ofcopolymers of phosphazene

Carlos Diaz*, Marıa Luisa Valenzuela, Miguel BarbosaDepartamento de Quımica, Facultad de Ciencias, Universidad de Chile Casilla, Santiago 653, Chile

Received 23 August 2003; received in revised form 26 August 2003; accepted 8 September 2003

Abstract

Incorporation of the organometallic fragments MLn ¼ CpFe(dppe)I and CpRu(PPh3)2Cl to the random copoly-

mers [{NP(O2C12H8)}0.8{NP(OC6H4CH2CN)2}0.15{NP(OC6H4CH2CN)(OC6H5)}0.05]n (1) [{NP(O2C12H8)}0.55-

{NP(OC6H4CH2CN)2}0.2{NP(OC6H4CH2CN)(OC6H5)}0.25]n (2) produce the new compounds of the approximate

composition: [{NP(O2C12H8)}x{NP(OC6H4CH2CN�MLn)2}y{NP(OC6H4CH2CN�MLn)(OC6H5)}z]n. The iron

derivatives in solution undergo deprotonation of the dppe, probably caused by a polymeric matrix, which in

turn causes a spontaneous cross-linking affording sparingly, soluble materials. Thermal analysis of the

compounds using DSC and DTA techniques indicates that the incorporation of the organometallic fragment in the

copolymer produces materials, which leave an amount of residue depending on the metal, as well as on

the functionalization degree of the cyanide groups. Incorporation of the organometallic fragments to the polymer

does not produce an improved conductivity in comparison to the insulator polymer without the organometallic

fragment.

# 2003 Elsevier Ltd. All rights reserved.

Keywords: A. Organometallic polymers; B. Chemical synthesis; C. Impedance spectroscopy; D. Thermal properties

1. Introduction

Organometallic fragments supported on polymers have been extensively investigated because of theirimportance for use in catalysis [1]. On the other hand, cross-linking of macromolecules is one of themost important aspects of polymeric materials [2–4]. Most of these cross-linked polymers involvepolysilazanes [5–8], polyphosphazenes [2,3,9] and others [10]. In fact cross-linking is an essential partof photolithography, elastomers technology, biomedical materials research and the conversion ofpolymers to ceramics. In this context, one of the main challenges in the translation of fundamental

Materials Research Bulletin 39 (2004) 9–19

* Corresponding author. Tel.: þ56-2-678-7396; fax: þ56-2-271-3888.

E-mail address: cdiaz@uchile.cl (C. Diaz).

0025-5408/$ – see front matter # 2003 Elsevier Ltd. All rights reserved.

doi:10.1016/j.materresbull.2003.09.030

polymer chemistry into useful materials research involves the development of cross-linking methods.Three methods of polymer cross-linking have been described: (i) use of microcrystallites, (ii) covalentcross-linking and (iii) ionic or coordination cross-links. Only the latter involves transition metals. Inmethods (i) and (iii) the interaction leading to the cross-link is rather weak, while method (ii) isrestricted to polymers having some active side groups, able to react between them, giving covalent linksbetween the chains of the polymer.

Due to intrinsic thermal stability of the polyphosphazenes, their thermal cross-linking processes usuallyproduce preceramic materials [2,3,9]. On the other hand, ceramics containing metals have interestingelectronic and magnetic properties. The incorporation of transition metals having predeterminednanostructures into the ceramic material is not simple. Some methods using organometallic polymers aspreceramic precursors have been suggested [2], but scarce examples have been reported [11,12]. Theorganometallic polymers containing organometallic groups pendent to their polymeric backbone could bepromising preceramic precursors for ceramic materials containing transition metals.

During our research into the incorporation of organometallic fragments to copolyphosphazeneshaving nitrile side groups [13,14], we found that the fragment CpFe(dppe)þ incorporates to the randomcopolymers 1 and 2 with a deprotonation of the dppe. This in turn provokes a spontaneous cross-linkingof the polymer chains, which results in insoluble materials. In contrast, the organometallic fragmentCpRu(PPh3)2

þ is incorporated in a conventional way to the phosphazene polymer backbone. This paperreports these results.

2. Experimental

Infrared spectra were recorded on a FT-Bruker-66V spectrometer in Nujol. NMR spectra wererecorded on a Bruker AMX 300 spectrometer using TMS (H) d ¼ 0:0 ppm as internal standard or85% H3PO4 and downfield positive to the reference as external standard for the 31P measurements.Elemental analyses were performed on Fisons-Carlo Erba EA 1108, apparatus by CEPEDEQ,Universidad de Chile.

Visible absorption spectra were measured on a Varian DMS-90 Spectrophotometer in 0.1 cm lengthcuvettes for the solids and 1 cm length cuvettes for the solutions. TG/DSC measurements wereperformed on a Netzsch STA 409 instrument. The polymer samples were heated at a rate of 10 8C/minfrom ambient temperature to 1000 8C under a constant flow of nitrogen. Electrical conductivitymeasurements were determined in polycrystaline samples using an electrochemical ImpedanceAnalyzer from Autolab PGSTAT-12 with FRA 4.8 program in the range of 0.1–106 Hz; the electricalcontacts were provided with gold layers of about 0.1 mm thickness by means of sputtering using thePelco SC-6 equipment. The polymer samples (31 mg) were compacted under 10 t of pressure toproduce pellets 0:75 mm � 6:7 mm diameter.

All reactions were carried out under dinitrogen with standard Schlenk techniques. The copolymers 1and 2 were prepared from the reaction of [NPCl2]n with HOC6H4–C6H4OH followed by the addition ofthe respective stochiometric amount of HOC6H4CH2CN, using the previously reported carbonatealkaline method [15]. NH4PF6 and CpRu(PPh3)2Cl were purchased from Aldrich. CpFe(dppe)I wasprepared by a previously reported method [16]. The solvents, THF, CH2Cl2, and diethyl-ether weredried and purified by standard methods. An approximate composition of the organometallic polymersare shown in Plate 1.

10 C. Diaz et al. / Materials Research Bulletin 39 (2004) 9–19

3. Preparation of the materials

3.1. Reaction of {[NP(O2C12H8)0.8(OC6H4CH2CN) 0.355(OC6H5)0.045}n with CpFe(dppe)I

A mixture of the polymer 1 (0.05 g) and CpFe(dppe)I (0.1228 g) in dichloromethane (20 ml), inpresence of TlPF6 (0.134 g) were stirred for 8 days at room temperature. The red-orange solution wasfiltered through celite and the solution evaporated under vacuum. The solid residue was washed twicewith a n-hexane/diethylether 4:1 mixture and the red powder was dried under reduced pressure. Yield0.09 g. Analysis found (calc. for {C24.8H19.6F2.4N1.4O2P2.2Fe0.04}n) C 53.77 (58.94); H 3.77 (3.9).N 3.77 (3.9).

3.2. Reaction of {[NP(O2C12H8)0.553(OC6H4CH2CN)0.646(OC6H5)0.2489} with CpFe(dppe)I

A mixture of the polymers 2 (0.05 g) and CpFe(dppe)I (0.1257 g) in dichloromethane l (20 ml), in thepresence of TlPF6 (0.1361 g) were stirred for 8 days at room temperature. The cloudy red-orangesolution was filtered through celite and the solution evaporated under vacuum. The solid residue waswashed twice with a n-hexane/diethylether 4:1 mixture and the red powder was dried under reducedpressure. Yield 0.08 g. Analysis found (calc. for {C30.2H24.8F3.3N1.55O2P2.65Fe3.3}n) C 54.16 (58.8);H 4.19 (4.05); N 1.93 (5.52).

3.3. Synthesis of [NP(O2C12H8)0.8(OC6H4CH2CN�CpRu(PPh3)2)0.355(OC6H5)0.045]n

A solution of CpRu(PPh3)2Cl (0.12 g) in CH3OH (20 ml) was stirred with the polymer 7(0.0413 g) in the presence of NH4PF6 (0.0648 g) at room temperature for 3 days. The solvent wasremoved under reduced pressure and the yellow-orange residue extracted with CH2Cl2 (15 ml) andfiltered through celite. The solution was concentrated under vacuum and n-hexane/dithylether 4:1mixture was added. The yellow solid precipitate was washed twice with the same mixture and dried.Yield 0.1 g.

Analysis found (calc. for {C29.2H22.8F2.4O2N1.4P2.2 Ru0.4}n) C 62.99 (60.52); H 4.41 (3.96); N 2.97(3.38). 31P{1H} NMR (CDCl3): �5.36 [m, P-spiro), �21.2 [m, POC6H4CH2CN�[Ru]], 39.3 (s, PPh3).1H NMR (CDCl3): 7.11–7.34(m, C6H5, C6H4, bispiro), 4.1 (s, C6H5), 3.1(m, CH2).

Plate 1.

C. Diaz et al. / Materials Research Bulletin 39 (2004) 9–19 11

3.4. Synthesis of [NP(O2C12H8)0.553(OC6H4CH2CN�CpRu(PPh3)2)0.646(OC6H5)0.248]n

A solution of CpRu(PPh3)2Cl (0.1396 g) in CH3OH (30 ml) was stirred with the polymer 1 (0.051 g)in the presence of NH4PF6 (0.084 g) at room temperature for 3 days. The solvent was removed underreduced pressure and the yellow-orange residue extracted with CH2Cl2 (15 ml) and filtered throughcelite. The solution was concentrated under vacuum and n-hexane/dithylether 4:1 mixture was added.The yellow solid precipitate was washed twice with the same mixture and dried. Yield 0.11 g.

Analysis found (calc. for {C35.65H28.6F3.3O2N1.55P2.65 Ru0.55}n) C 60.9 (60.21); H 4.32 (4.05); N 3.41(3.05). 31P {1H} NMR (CDCl3): �4.99[m, P-spiro), �20.63 [m, POC6H4CH2CN�[Ru]], 39.3 (s, PPh3).1H NMR (CDCl3): 7.13–7.36 (m, C6H5, C6H4, bispiro), 4.1 (s, C6H5), 3.26(m, CH2).

4. Results and discussion

4.1. Reaction of the copolymers 1 and 2 with CpFe(dppe)I

Selected IR data for the compounds 1–6 are in Table 1. The polymers 1 and 2 reacted readily withCpFe(dppe)I in CH2Cl2 in the presence of TlPF6 to give orange-brown solids. Elemental analysis wereslightly different according to the ideal formula for 5 and 7 (Plate 1). This can be due to the presence ofsome contaminant in the solid produced in the reaction leading to the insoluble solid after the cross-linking. The solution for compound 5 was not clear, it was cloudy. On the other hand, as observedfrequently in some metallic derivatives of polyphosphazenes [17,18], this polymer, initially soluble inCH2Cl2 when handled for the separation, lead to insolubilization of the polymer. This did not allow anadequate NMR characterization of the sample. However, in THF a somewhat soluble fraction of thepolymer was obtained, then their NMR spectrum was visualized. Multinuclear NMR (Fig. 1(A)) clearlyshows, for the iron derivatives, the presence of the fragment CpFe(dppe) deprotonated at the methyleneof the bis(diphenylphosphino)ethane ligand. Deprotonation of [CpFe(dppe)(C=C(CH3)2]þ in a basicmedia has been reported [19]. In fact, the absence of 31P NMR signals at �90 ppm, as well as thepresence of the signals around 50 and 42 ppm, indicate the presence of dppe deprotonated ligated to

Table 1

Relevant IR solid state data for copolymers 1–6

Compounds n(CN) n(C–OP) n(PN) n(P–OC) n(C5H5) n(PF6)

1 2250 1273 1246 1171 –

1196

2 2250 1243 1189 1117 –

3a 2216 1268 1245 1120 1097 839

4 2250 1271 1246 1160 1070 –

1196

5a 2216 1246 1189 1122 1096 837

1174

6 2249 1267 1246 1168 1095 841

1198

aOther bands not observed for the other polymer were observed around 3300, 2700, and 1600 cm�1 (see text).

12 C. Diaz et al. / Materials Research Bulletin 39 (2004) 9–19

iron [19]. Signals around �5.03, �6.07 and �15.11 correspond to three types of phosphorus in thebackbone P¼¼N polymer. The signal corresponding to PF6 is missing in the 31P NMR spectrum.However, the presence of the PF6 anion in the solid state is suggested by the intense n(PF6) band around840 cm�1 in its IR spectrum. On the other hand, the appearance of IR bands in solid state in the range:2721, 2680 and 2614 cm�1; 2216 and 1697 cm�1 not present in the free polymer nor in theorganometallic fragments, clearly indicate the presence of groups of the type C=N� R(R0) (R ¼ H orarilo) [20,21]. The solution 13C NMR spectrum of the material, although of not good quality, exhibitsthe expected signals of the polymer, being the CN and the CH2 carbon signal at 120.14 and 19.44;14.14, respectively (free polymer: 120.95 and 26.5; 22.55). The two different CH2 resonances are dueto OC6H4CH2CN groups linked to different (P¼¼N) units, which is consistent with the presence ofdeprotonated dppe ligated to iron. Two =C–H carbon resonance signals typical of carbon bridgehead[19] (d13C ¼ 56:9 ppm in CpFePPh2CH2CH PPh2C=C(CH3)2 [19]) were observed at 61.12 and59.14 ppm. The deprotonation of dppe causes a cross-linking between the polymeric chains producinginsoluble materials with probable structures (Fig. 1(A)). Table 1 shows IR bands of the n(PN) vibrationsof the polyphosphazene backbone [22] observed normally around 1200 cm�1.

Polymer 3 is soluble in CH2Cl2 and hence, an adequate NMR characterization was made (Fig. 2).Their 31P NMR spectrum exhibited the signals typical of bis(biphenylphosphino)ethano mono-deprotonated iron cyclopentadienyl fragments, linked to two different [P¼N] sites in the polymerbackbone. A small amount of the fragment CpFe(dppe)þ appears to be present in the polymer, asevidenced by a small signal at 92.35 ppm typical of dppe coordinated to nitriles [23,24]. The

Fig. 1. Schematical representation of the possible structure of the organometallic iron fragment linked to polymer 5, showing

some of their NMR data (A) and the probable structure of the cross-linking polymer (B).

C. Diaz et al. / Materials Research Bulletin 39 (2004) 9–19 13

approximate ratio 1:3 of the two 31P-signal at 89.66 and 87.66 ppm is consistent with one dppedeprotonated linked to iron cyclopentadienyl fragment in each unit containing the cyanide spacer. Astrong signal at 12.24, and two weak at 2.9 and �1.6 are assigned to the three different P¼¼N units in thepolymers. A strong heptuplet signal at �144 ppm indicates the presence of PF6

� consistent with then(PF6) at 859 cm�1, observed in the IR spectrum of this polymer in solid state. The 13C NMR spectrumexhibits in addition to the normal expected signal of the polymer precursors, the signals correspondingto two typical C5H5 groups of CpFe(dppe) fragments linked probably to two different sites in the P¼¼Npolymer backbone. They appear at 81.8 and 72.16 ppm. Two signals at 61.12 and 59.7 also suggest thepresence of bridgehead carbon resonance [19] =CH–. The 1H NMR spectrum exhibited severalcomplex groups of signals with remarkable fine structures in the phenyl region. Three multiplets ofapproximately equal intensity at 7.1, 7.3 and 7.4 ppm are typical of aromatic protons of the fragment(Cp)FePPh2CH2CH(PPh2) [19]. Aromatic hydrogens of the polymers are observed as multiplets at 6.45,6.59, 6.97 and 7.61 ppm. These data are consistent with a polymeric chain having differentcyclopentadienyl iron moieties as is represented schematically in Fig. 2. These groups may be anchoredto the same or different chain, as indicated by the interrogation mark. Additionally, the appearance ofIR bands around 3391, 2216 and 2718 cm�1; 2676 and 1786 cm�1 assigned to the presence of themoiety C=N� R(R0) (R ¼ H or arilo) [20,21] are in agreement with the proposed structure of thepolymers (Fig. 2).

Fig. 2. Schematical representation of the possible structure for the different type iron of organometallics fragments linked to

the polymer 3 backbone. NMR data are indicated in rectangle. Cross-linking between the chains are not shown.

14 C. Diaz et al. / Materials Research Bulletin 39 (2004) 9–19

4.2. Reaction of the copolymers 1 and 2 with CpRu(PPh3)2Cl

Contrary to the polymers containing Fe, the reaction of polymers 1 and 2 with CpRu(PPh3)2Cl inpresence of NH4PF6 and in MeOH as solvent, afforded the expected polymers having the unalteredorganometallic fragment CpRu(PPh3)2 anchored to the cyanide group. Elemental analysis, as well asspectroscopic data (Table 1 and Section 2), are in agreement with the proposed formula.

Although analysis of polymer 6 is satisfactory, that of polymer 4 is somewhat deviate, factnot unusual in organometallic polymers with pendent coordinated groups [17,18]. Thus, a smallamount of free CpRu(PPh3)2Cl was detected spectroscopically. Numerous attempts to purify thesample by usual methods, as found for other organometallic polymers, lead to decomposition orinsolubilization.

In addition to the 31P-signal corresponding to the N¼¼P(O2C12H8)/N¼¼P(OC6H4CH2CN)2,NP(OC6H5)(OC6H4CH2CN) moiety in the polymer with the correct ratio 0.8/0.2 for 4 and 0.55/0.45for 6, the spectrum exhibited the typical signal of the CpRu(PPh3)2 moiety groups. However, the signalcorresponding to PF6 is missing in polymer 4. This is in agreement with the IR spectra, where then(PF6) band observed normally, around 840 cm�1, is absent. This means that in the formation of thepolymer 4, the chloride probably remains as the counterion, instead of the PF6 anion. This can be due tothe well known ionization of the chloride from the CpRu(PPh3)2Cl in MeOH [25,26].

CpRuðPPh3Þ2Cl ! CpRuðPPh3Þ2þ þ Cl�

4.3. UV-Vis spectra of the copolymers

The UV-Vis spectra of the polymer complexes clearly confirm the incorporation of the organometallicfragments to the polymeric backbone. The polymers 1 and 2 do not absorb above 350 nm, while theorganometallic polymers exhibit absorption in the range 350–700 nm. In fact the iron organometalliccopolymers 3 and 5 exhibit a continuum increasing absorption from red (�700) to blue, which can be thetail of strong absorption in the ultraviolet zone (two intense maxima at 230 and 250 nm). However, ashoulder around 438 nm was observed for both polymer complexes, which could be due to the presence ofsome CpFe(dppe)NCRþ chromophores [23,24]. The continuous decreasing absorption could mask anyweak transitions due to the presence of other (CpFe(dppe), deprotonated)þ species in the polymer chain.Thus, the bispiro iron model complex [N3P3(O2C12H8)2(OC6H4CH2CN�Fe(Cp)(dppe))2][PF6]2 exhibitsthree absorptions centered at: 630, 530 and 450 nm [23]. On the other hand UV-Vis absorption patterns ofthe polymer metal complexes 4 and 6 are similar to that of their respective oligomer ruthenium complexes[23,24] suggesting and corroborating that the chromophores CpRu(PPh3)2 are coordinated, unaltered tothe polymer chain.

4.4. Thermal analysis of the polymers and their complexes

The glass transition temperatures (Tg) for the organometallic polymers and polymers without metalwere determined by differential scanning calorimetry and are listed in Table 2. For the polymer 6, theruthenium derivative raised the Tg value as observed for other polyphosphazenes with pendentorganometallic fragment chains [27]. The bulky group CpRu(PPh3)2 appears to restrict the torsion ofthe polymer. The thermal behavior of polymers 1 and 2 and their iron and ruthenium complexes were

C. Diaz et al. / Materials Research Bulletin 39 (2004) 9–19 15

also examined by thermogravimetric analysis. From Fig. 3 and Table 2 it can be seen that theincorporation of an organometallic fragment to the polymer decreases the temperature of the onset ofdecomposition. Thus, polymers 3, 4 and 6 underwent a gradual weight loss in the temperature range 3:300–500 8C; 4: 280–530 8C; 6: 260–600 8C. In contrast, the polymers 1, 2 and 5 underwent a rapidweight loss at ca. 1: 400 8C; 2: 390 8C; 6: 300 8C. This can be attributed to the volatilization of thelow molecular weight species (mainly cyclic oligomers) previously formed in the temperature range100–200 8C [9]. With regard to the effect of the CpFe(dppe) and CpRu(PPh3)2 units on the thermalproperties of the polymers, in Table 2 it can be seen that the coordination of the organometallicmoieties on the polymer 1 decreases the percentage of nonvolatile residue in the case of the ironderivatives, while with the ruthenium polymers an increase of the pirolitic residue is observed.However, the coordination of either organometallic fragment to polymer 2 decreases the residuepercentage after the pyrolysis.

Table 2

TG and DSC data for polymers and their complexes

Polymers Dec (8C) % Residue (8C) Tg (8C)

1 396 35 (960) 123

2 400 46 (980) 85.4

3 298 20 (980) –

4 280 40 (980) –

5 280 18 (980) 112.6

6 260 35 (980) 106.2

Fig. 3. Thermal analysis of the copolymer 1 and their ruthenium organometallic derivative 4.

16 C. Diaz et al. / Materials Research Bulletin 39 (2004) 9–19

It appears that the iron organometallic complexes decrease the amount of residue after pyrolysis,which can be due to the cleavage of the cross-linking material with the temperature. On the contrary,the ruthenium organometallic complex increases or decreases the residue depending on the charge ofthe CN groups of the polymers. An increase in pirolitic residue can be due to the effective cross-linkingprovoked by the effective link of the ruthenium fragments to the chains of the polymers.

The lower thermal stability of the iron polymers 3 and 5 in comparison to the respective rutheniumpolymers 4 and 6 can be due to two factors: (i) facile loss of a triphenylphosphine ligand from theCpRu(PPh3)2 moiety, as reported to occur with temperature [25,26], would produce a reactiveCpRu(PPh3)2 group to coordinate another polymer chain leading to further cross-linking and (ii) theweakness of the bond: polymer–CH=N(CpFe(dppe))—(dppe-2H)-CpFe-N=CH–polymer (Fig. 1).

4.5. Conductivity measurements

To explore the possibility of electron transfer behavior of the copolymer after incorporation ofan organometallic fragment, electrical conductivity measurements were carried out by impedancespectroscopy. Values in the range of s ¼ 10�9 to 10�11 S cm�1 were obtained for both the copolymerand the copolymer that contains the organometallic fragments (Table 3). These values are typical ofinsulators. Similar electrical conductivities have been reported for other polyphosphazenes (e.g.[PN(OC6H5)2]n) [30].

The electrical conductivity of organometallic polymers has not been fully investigated. Relatively fewstudies have been reported [28–32]. However, some data suggest that the conductivity depends on threefactors: (a) the conductive properties of the polymer backbone [28,32]. (b) the p-conjugated behavior ofthe spacer that is connected to the polymer backbone, and (c) the electronic ‘richness’ of theorganometallic fragment. Although the N¼¼P backbone in copolymers 1 and 2 is insulating, electricalconductivity may be expected due to charge hopping between the metal centers; namely, the externalsphere mechanism that was proposed for polyphosphazenes that contain pendant ferrocene groups [27].In this example, the fragment, Cp2Fe, is electrochemically active as well as electron rich. In contrast, theinsulator behavior of the iron and ruthenium organometallic polymers may well be due to a coilconformation of the polymer that does not permit external sphere electron transfer between the metalcenters. In fact, solution properties studied by size exclusion chromatography using simultaneouslymultiangle light scattering and differential refractive index detectors, indicated that polyspiropho-sphazenes, including random copolymers similar to 3, behave as a random coil [33,34]. Further studiesto correlate electrical conductivity with polymer structure and conformation are needed and are currentlyunderway. Alternatively, another explanation of the insulator behavior of the organometallic polymermay be the low metal content of the polymer and the non-conductive polyphosphazene matrix.

Table 3

Impedance spectroscopy conductivity’s for the copolymers and their organometallics derivatives

Polymers s (S cm�1)

1 1.2 � 10�11

2 1.6 � 10�10

3 4.6 � 10�10

5 4.5 � 10�10

6 1.16 � 10�9

C. Diaz et al. / Materials Research Bulletin 39 (2004) 9–19 17

4.5.1. General discussion: possible mechanism of the cross-linking reactionThe reaction of the copolymers 1 and 2 with CpFe(dppe)I does not afford the expected polymer having

the anchored unaltered fragment. Instead, the reaction produced a deprotonation of the dppe ligand witha concommitant cross-linking of the polymer. We believe that this unexpected reaction can be due to amatrix effect, because of the analogue reaction of the organometallics CpFe(dppe)I with the monomerligand HOC6H4CH2CN gives the expected complex containing the unaltered iron fragment[CpFe(dppe)�NCCH2C6H4OH][PF6]. On the other hand the reaction of the same iron organometallicswith the oligomers [N3P3(O2C12H8)2(OC6H4CH2CN)2] and with [N3P3(OC6H4R)5(OC6H4CH2CN)]R ¼ H, But also afford the expected complexes [N3P3(O2C12H8)2(OC6H4CH2CN�Fe(Cp)(dppe))2][PF6]2

and [N3P3(OC6H4R)5(OC6H4CH2CN�CpFe(dppe))][PF6], respectively. We tentatively attribute theobserved deprotonation behavior of the Fe systems to the occurrence of several nitrogen ligands neareach other, which creates a basic-like micro-environment. This in turn provokes cross-linking of thepolymer chains. In fact, several ‘‘macromolecular effects’’ involving chemical reactions have beenreported [35].

5. Conclusions

1. The reaction of the copolymers 1 and 2 with the iron fragments afford new materials of somecomplex composition in which groups CpFe(dppe)þ and their deprotonated derivatives are anchoredto the polymeric chain producing insoluble cross-linked solids.

2. The reaction of the copolymers 1 and 2 with the ruthenium organometallic fragments afford solublecopolymers with unaltered CpRu(PPh3)2

þ groups anchored to the polymer chains.3. The thermal stability of the organometallic polymers 3, 4, 5 and 6 was low when compared with that

of the free polymeric ligands 1 and 2, which decomposed at 150 8C. However, the final residueabove 800 8C was somewhat high for the ruthenium polymers due to their more effectivecoordinative cross-linking behavior than their iron copolymers, which exhibited low residues.

Acknowledgements

This work was supported by FONDECYT (1000672 and partially 1030515), and DID (PG-165/2001).

References

[1] W.E. Leardbeater, M. Marco, Chem. Rev. 102 (2002) 3217.

[2] H.R. Allcock, Chem. Mater. 6 (1994) 1476.

[3] H.R. Allcock, Ch.G. Cameron, Macromolecules 27 (1994) 3125.

[4] D. Mecerreyes, V. Lee, C.J. Hawker, J.L. Hedrick, A. Wursch, W. Volksen, T. Magbitang, E. Huang, R.D. Miller, Adv.

Mater. 13 (2001) 204.

[5] W.R. Schmid, D.M. Narsavage-Heald, D.M. Jones, P.S. Marchetti, D. Raker, G.E. Maciel, Chem. Mater. 11 (1999) 1455.

[6] R.J.P. Corriu, J.J.E. Moreau, P. Thepot, M.W. Chiman, Chem. Mater. 8 (1996) 100.

[7] Y.D. Blum, K.B. Schwartz, R.D. Laine, J. Mater. Sci. 24 (1989) 1707.

[8] M. Weinmann, A. Zern, F. Aldinger, Adv. Mater. 13 (2001) 1704.

[9] H.R. Allcock, G.S. McDonnell, G.H. Riding, I. Manners, Chem. Mater. 2 (1990) 425.

18 C. Diaz et al. / Materials Research Bulletin 39 (2004) 9–19

[10] T. Wideman, L.G. Sneddon, Chem. Mater. 8 (1996) 3.

[11] R. Peterson, D.A. Foucher, B.Z. Tang, A. Lough, N.P. Raju, J.E. Greedan, I. Manners, Chem. Mater. 7 (1995) 2045.

[12] E.J. Houser, T.M. Keller, Macromolecules 31 (1998) 4038.

[13] C. Dıaz, P. Castillo, J. Inorg. Organomet. Polym. 11 (2002) 183.

[14] C. Dıaz, P. Castillo, G.A. Carriedo, P. Gomez-Elipe, F.J. Garcıa-Alonso, Macromol. Chem. Phys. 203 (2003) 1912.

[15] G.A. Carriedo, L. Fernandez-Catuxo, F.G. Garcıa-Alonso, P. Gomez-Elipe, P.A. Gonzalez, Macromolecules 29 (1999)

5320.

[16] C. Dıaz, R. Latorre, Bol. Soc. Chil. Quım. 37 (1992) 211.

[17] H.R. Allcock, A.A. Dembeck, E.H. Klingeberg, Macromolecules 24 (1991) 5208.

[18] H.R. Allcock, I. Manners, M.N. Mang, M. Parvaez, Inorg. Chem. 29 (1990) 522.

[19] R.D. Adams, A. Davison, J.P. Selegue, J. Am. Chem. Soc. 101 (1979) 7232.

[20] W.P. Fehlhamer, M. Frit, Chem. Rev. 93 (1993) 1243.

[21] R. Bar, P.W. Fehlhamer, J. Organomet. Chem. 353 (1988) 197.

[22] G.A. Carriedo, F.I.G. Alonso, P.A. Gonzalez, J.R. Menendez, J. Raman Spectrosc. 29 (1998) 327.

[23] C. Dıaz, I. Izquierdo, F. Mendizabal, N. Yutronic, Inorg. Chim. Acta 294 (1999) 20.

[24] C. Dıaz, I. Izquierdo, Polyhedron 18 (1999) 1479.

[25] P.M. Treichel, D.A. Komar, P.I. Vincenti, Synth. React. Inorg. Met. Org. Chem. 14 (1984) 383.

[26] J.M. Bruce, F.S. Wong, B.W. Skelton, A.H. White, J. Chem. Soc. Dalton (1981) 1398.

[27] H.R. Allcock, E.N. Silverberg, C.J. Nelson, W.D. Coggio, Chem. Mater. 5 (1993) 1307.

[28] R.P. Kingsborough, T.M. Swager, Prog. Inorg. Chem. 48 (1999) 123.

[29] P. Nguyen, P. Gomez-Elipe, I. Manners, Chem. Rev. 99 (1999) 1515.

[30] H.R. Allcock, Th.X. Neenan, Macromolecules 19 (1986) 1495.

[31] M.O. Wolf, M.S. Wrigton, Chem. Mater. 6 (1994) 1526.

[32] G. Zotti, S. Zecchin, G. Schiavon, Chem. Mater 7 (1995) 2309.

[33] G.A. Carriedo, F.J. Garcıa-Alonso, P. Gomez-Elipe, J.L. Garcıa Alvares, M.P. Tarazona, M.T. Rodriguez, E. Saiz, J.T.

Padron, Macromolecules 33 (2000) 3671.

[34] J. Burdalo, M.P. Tarazona, G.A. Carriedo, F.J. Garcıa-Alonso, P.A. Gonzalez, Polymers 40 (1999) 4251.

[35] D.C. Sherrington, P. Hodge, Syntheses and Separation Using Functional Polymers, John Wiley & Sons, New York, 1988.

C. Diaz et al. / Materials Research Bulletin 39 (2004) 9–19 19