Synthesis, crystal structure and electrochemistry of tetrahedral mono-β-diketonato titanocenyl complexes

Inorganica Chimica Acta 360 (2007) 2277–2283 www.elsevier.com/locate/ica

Synthesis, crystal structure and electrochemistry of tetrahedral mono-b-diketonato titanocenyl complexes Elizabeth Erasmus, Jeanet Conradie *, Alfred Muller, Jannie C. Swarts

*

Department of Chemistry, University of the Orange Free State, Bloemfontein 9300, South Africa Received 10 August 2006; received in revised form 5 November 2006; accepted 11 November 2006 Available online 18 November 2006

Abstract The synthesis of a variety of tetrahedral b-diketonato titanium(IV) complexes of the type ½ðC5 H5 Þ2 TiðCH3 COCHCORÞþ ClO4  with R = CF3, OCH3, C6H5, CH3 and Fc is described. The TiIII/TiIV couples and the Fc/Fc+ couple exhibited chemically and electrochemically reversible cyclic voltammetric behaviour. The formal reduction potential of the TiIII/IV couple increased as the group electronegativity of the R group of the b-diketonato ligand increased. Bulk electrolysis showed that one electron was transferred in the TiIII/IV couple and one electron in the ferrocenyl/ferrocenium redox couple in the ligand. The crystal structure for the R = OCH3 complex showed that this b-keto-ester binds through the carbonyl oxygen of the ester group and not the ether oxygen.  2006 Elsevier B.V. All rights reserved. Keywords: Titanocene; b-diketonates; Cyclic voltammetry; Crystal structure

1. Introduction Titanocene dichloride is a convenient starting material for the preparation of many other organometallic compounds of titanium [1] and has found application as a catalyst or catalyst component for a wide variety of hydrometalation and carbometalation reactions as well as other transformations involving Grignard reagents. A key feature of this titanocene dichloride (1) is its complex aqueous chemistry that includes aquated species such as 2 and 3 in Scheme 1 [2]. Cationic b-diketonato dicyclopentadienyl titanium(IV) complexes was first prepared by Doyle and Tobias [3,4]. Although no crystal structure of titanium(IV) complexes of the type [(C5H5)2TiIV(R1COCHCOR2)]+ is known, the structures of the analogous titanocene(III) compounds [(C5H5)2TiIII(R1COCHCOR2)] with R1, R2 = CH3 or

*

Corresponding authors. Tel.: +27 51 4012194; fax: +27 51 4446384 (J. Conradie). E-mail address: [email protected] (J. Conradie). 0020-1693/$ - see front matter  2006 Elsevier B.V. All rights reserved. doi:10.1016/j.ica.2006.11.010

C6H5 [5,6] have been reported. The coordination sphere of these complexes does not change when switching the redox state of the central coordinating metal from TiIII to þ TIIV. For the ½ðC5 H5 Þ2 TiðCH3 COCHCO–OEtÞ ClO4  complex, there are conflicting structural literature reports. Based on the IR results, some authors concluded that bonding at the ester takes place via the keto oxygen [3], while others concluded that bonding takes place via the ‘‘ether’’ type oxygen of the OEt group [7]. Electrochemical studies by Bond et al. on [(C5H5)2TiIII(CH3COCHCOR)] with R = CH3 or C6H5 show a chemically and electrochemically reversible one electron 0 TiIII/IV couple at formal reduction potentials E0 = 0.86 and 0.85 V versus Fc/Fc+ in 0.2 M (nBu4N)(PF6)/butyronitrile, respectively [5]. It was concluded that the formal reduction potentials of these complexes are essentially independent of the b-diketonato ligand when R = CH3 or C6H5. This must be so because the group electronegativities of the CH3 and C6H5 groups are almost the same (vCH3 ¼ 2:34 and vC6 H5 ¼ 2:21) [8]. It remains unexplored as to what the influence of the b-diketonato side group substituents would be if their group electronegativities vary

2278

E. Erasmus et al. / Inorganica Chimica Acta 360 (2007) 2277–2283 2+ . 2ClO4-

2+ . 2ClCl Ti

OH2

H2O

Ti

O

A g C l O4

O

O C H3

R

OH CH3

+ 2AgCl O H2

3 (and other aquated species)

2 (and other aquated species)

O R

Ti

OH2

Cl

3

OH 2

C H3

+ ClO 4

Ti O R

R= CF3,4; OCH3,5; C6H5,6; CH3,7; Fc, 8.

Scheme 1. Synthesis of ½ðC5 H5 Þ2 TiðCH3 COCHCORÞþ ClO4  complexes.

substantially. This study reports the synthesis of some þ known and new ½ðC5 H5 Þ2 TiIV ðCH3 COCHCORÞ ClO4  complexes (Scheme 1) where the R group was chosen to have widely different group electronegativities to enable us to determine the extent of intramolecular electronic communication through an electrochemical study with R = CF3 (4, vCF3 ¼ 3:01), OCH3 (5, vOCH3 ¼ 2:64), C6H5 (6, vC6 H5 ¼ 2:21), CH3 (7, vCH3 ¼ 2:34) and Fc (8, vFc = 1.87). The bonding mode of b-ketoesters to the TiIV core is also settled from the crystal structure determination þ of ½ðC5 H5 Þ2 TiIV ðCH3 COCHCO–OMeÞ ClO4  . 2. Experimental

2.2.1. 1-Methoxy-1,3-butanedionato-j2O,O 0 -bis(g5cyclopentadienyl)titanium(IV)perchlorate (5) A solution of the aquated bis(cyclopentadienyl)titanium(IV) species, 2, was prepared by stirring a solution of titanocene dichloride (250 mg, 1.0 mmol) in water (6 ml) for 30 min before silver perchlorate (376 mg, 1.98 mmol) dissolved in 3 ml water was added. The suspension was then stirred under nitrogen for an hour. Silver chloride was filtered off and washed with water (2 ml). The filtrate was cooled on an ice-bath and an excess methyl acetoacetate (1160 mg, 10.0 mmol) was added dropwise while stirring. The crystalline product was filtered after standing for a day and washed with water and ether to yield 182 mg (48%) of spectroscopically pure crystals of 5 suitable for single crystal X-ray analysis. M.p. = 230 C. IR (cm1) = 1568. 1H NMR (d/ppm, CDCl3) 2.11 (s, 3H, CH3), 3.98 (s, 3H, OCH3), 5.52 (s, 1H, CH), 6.79 (s, 10H, 2 · C5H5). Elemental Anal. Calc. for TiC15H17O7Cl: C, 45.9; H, 4.4. Found: C, 45.9; H, 4.3%. 2.2.2. 1,1,1-Trifluoro-2,4-pentanedionato-j2O,O 0 -bis(g5cyclopentadienyl)titanium(IV)perchlorate (4) This compound was prepared as above but by replacing the methyl acetoacetate with a solution of 154 mg (1.0 mmol) trifluoroacetylacetone in 1.5 ml ice cold THF. A yield of 69 mg (16%) was obtained after three days. M.p. = 230 C. IR (cm1) = 1593. 1H NMR (d/ppm, CDCl3) 2.25 (s, 3H, CH3), 5.96 (s, 1H, CH), 6.88 (s, 10H, 2 · C5H5). Elemental Anal. Calc. for TiC15H14O6ClF3: C, 41.8; H, 3.3. Found: C, 42.1; H, 3.5%.

2.1. Materials and apparatus Solid reagents (Merck, Aldrich and Sigma) were used without further purification. Liquid reactants and solvents were distilled prior to use, water was double distilled. CH3CN was dried over CaH2 and freshly distilled prior to use. 1-Ferrocenylbutane-1,3-dione [9], the titanocene þ complexes [3] ½ðC5 H5 Þ2 TiðCH3 COCHCOC6 H5 Þ ClO4  þ  (6) and ½ðC5 H5 Þ2 TiðCH3 COCHCOCH3 Þ ClO4 (7) were synthesized according to published procedures. Caution: All ClO4  salts should be handled with precautions against possible explosion. Melting points were determined with a Reichert Thermopan microscope with a Kofler hotstage and are uncorrected. NMR measurements were recorded on a Bruker Advance DPX 300 NMR spectrometer. Chemical shifts are reported as d values relative to SiMe4 (0 ppm). IR spectra were recorded from neat samples on a Digilab FTS 2000 Fourier transform spectrometer utilizing a He–Ne laser at 632.6 nm.

2.2.3. 1-Ferrocenoyl-1,3-butanedionato-j2O,O 0 -bis(g5cyclopentadienyl)titanium(IV)perchlorate (8) This compound was prepared as for 5 but by replacing the methyl acetoacetate with a solution of 210 mg (1.0 mmol) 1-ferrocenoyl-1,3-butanedione in 4 ml ice cold THF. A yield of 207 mg (48%) was obtained after three days. M.p. = 86–89 C. IR (cm1) = 1513. 1H NMR (d/ ppm, CDCl3) 2.25 (s, 3H, CH3), 4.41 (s, 5H, C5H5 from Fc), 4.79 (s, 2H, C5H4 from Fc), 4.96 (s, 2H, C5H4 from Fc), 6.22 (s, 1H, CH), 6.75 (s, 10H, 2 · C5H5). Elemental Anal. Calc. for FeTiC24H23O6Cl: C, 52.7; H, 4.2. Found: C, 53.5; H, 4.4%.

2.2. Synthesis

2.2.4. Characterisation data for 1-phenyl-1,3-butanedionatoj2O,O 0 -bis(g5-cyclopentadienyl)titanium(IV)perchlorate (6) Yield: 65%. M.p. = 189–194 C. IR (cm1) = 1547. 1H NMR (d/ppm, CDCl3) 2.44 (s, 3H, CH3), 6.79 (s, 10H, 2 · C5H5), 6.80 (s, 1H, CH), 7.55–7.70 (m, 3H, C6H5), 7.92–7.98 (m, 2H, C6H5).

Safety note: Perchlorate complexes are potentially explosive. We have experienced even small quantities to explode upon heating during melting point determinations. They should be treated with caution and handled in small quantities.

2.2.5. Characterisation data for 2,4-Pentanedionato-j2O,O 0 bis(g5-cyclopentadienyl)titanium(IV)perchlorate (7) Yield: 86%. M.p. > 230 C. IR (cm1) = 1543. 1H NMR (d/ppm, CDCl3) 2.30 (s, 6H, 2 · CH3), 6.14 (s, 1H, CH), 6.73 (s, 10H, 2 · C5H5).

E. Erasmus et al. / Inorganica Chimica Acta 360 (2007) 2277–2283

2.3. Electrochemistry Cyclic voltammetry measurements on 1.0 mmol dm3 solutions of the complexes in dry acetonitrile containing 0.100 mol dm3 tetra-n-butylammonium hexafluorophosphate ((nBu4N)(PF6), Fluka, electrochemical grade) as supporting electrolyte were made under a blanket of purified argon at 25.0 C utilizing a BAS 100 B/W electrochemical workstation interfaced with a personal computer. A three electrode cell, which utilized a Pt auxiliary electrode, a glassy carbon (surface area 0.0707 cm2) working electrode and a Ag/Ag+ (0.010 mol dm3 AgNO3 in CH3CN) reference electrode [10] mounted on a Luggin capillary was used [11,12]. In this work all cited potentials are referenced against the Fc/Fc+ couple as suggested by IUPAC [13]. 0 Ferrocene exhibited E0 = 77 mV versus Ag/Ag+, ipc/ ipa = 0.98, DEp = 74 mV under our experimental conditions [8]. Successive experiments under the same experimental conditions showed that all formal reduction potentials were reproducible within 5 mV. All temperatures were kept constant to within 0.5 C. Bulk electrolysis was carried out utilizing a BAS 100 B/ W voltammograph at 25.0 C in 2.5 cm3 acetonitrile. A three electrode cell, which utilized a Pt wire auxiliary elecTable 1 Crystal data and structure refinement for ½ðC5 H5 Þ2 TiðCH3 COCHCOðOCH3 ÞÞþ ClO4  (5) Empirical formula Formula weight Temperature (K) ˚) Wavelength (A Crystal system Space group Unit cell dimensions ˚) a (A ˚) b (A ˚) c (A a () b () c () ˚ 3) Volume (A Z Dcalc (Mg/m3) Absorption coefficient (mm1) F(0 0 0) Crystal size (mm) h Range for data collection () Index ranges Reflections collected Independent reflections [Rint] Completeness to h = 28.35 (%) Maximum and minimum transmission Refinement method Data/restraints/parameters Goodness-of-fit on F2 Final R indices [I > 2r(I)] R indices (all data) Largest difference in peak and hole ˚ 3) (e A

C15H17ClO7Ti 392.64 173(2) 0.71073 monoclinic P21/c 7.5326(2) 20.8283(5) 10.7927(3) 90 103.839(2) 90 1644.13(7) 4 1.586 0.717 808 0.33 · 0.22 · 0.19 2.18–28.35 10 6 h 6 10, 27 6 k 6 25, 14 6 l 6 14 30,323 4091 [0.0295] 99.8 0.8757 and 0.7977 Full-matrix least-squares on F2 4091/0/219 1.050 R1 = 0.0379, wR2 = 0.0966 R1 = 0.0441, wR2 = 0.1017 0.763 and 0.404

2279

trode (isolated from the sample by means of a 0.1 mol dm3 (nBu4N)(PF6)/CH3CN salt bridge), a sheet of glassy carbon working electrode submersed into a dept that gave an electro-active area of approximately 2 cm2 and a Ag/Ag+ (0.010 mol dm3 AgNO3 in 0.10 mol dm3 (nBu4N)(PF6)/CH3CN) reference electrode mounted on a Luggin capillary were employed. Current readings (I) and the charge transfer (C) were recorded automatically by using BAS 100 B/W Windows software. 2.4. X-ray crystal structure determination for 5 A dark red regular fragment was used for the X-ray crystallographic analysis of compound 5 (Table 1). The initial unit cell and data collection were achieved by the APEX2 software [14] utilizing COSMO [15] for optimum collection of more than a hemisphere of reciprocal space. The frames were integrated using a narrow-frame integration algorithm and reduced with the Bruker SAINT-PLUS [16] and XPREP [16] software packages, respectively. Data were corrected for absorption effects using the multi-scan technique SADABS [17]. The structure was solved by the direct methods package SIR97 [18] and refined using the WINGX software package [19] incorporating SHELXL [20]. The molecular plot was drawn using the DIAMOND program [21] with a 50% thermal envelope probability for non-hydrogen atoms. 3. Results and discussion 3.1. Synthesis of complexes The complexes ½ðC5 H5 Þ2 TiðCH3 COCHCOCF3 Þþ ClO4  (4), ½ðC5 H5 Þ2 TiðCH3 COCHCO–OMeÞþ ClO4  (5), þ ½ðC5 H5 Þ2 TiðCH3 COCHCOC6 H5 Þ ClO4  (6), ½ðC5 H5 Þ2 Tiþ ðCH3 COCHCOCH3 Þ ClO4  (7) and ½ðC5 H5 Þ2 TiðCH3 þ  COCHCOFcÞ ClO4 (8) were obtained by adding either an excess of the b-diketone (in the case of 5 or 7) or a THF solution of a stoichiometric amount of the b-diketone dissolved in THF to an aqueous solution of the aquated bis(cyclopentadienyl)titanium(IV) species (3), Scheme 1. Coordination of the b-diketonate ligand to the titanium centre is promoted by the fact that H2O is a much poorer coordination group than Cl. Removal of the Cl ion by AgCl precipitation drives chloride displacement from the titanium(IV) centre with water molecules to the maximum yield of 3. Compounds 5 and 7 precipitated immediately upon b-diketone addition, while precipitants of 4, 6 and 8 slowly formed from the reaction mixtures at the bottom of the flask at room temperature after more than 24 h. If the AgClO4 is added in excess, the yield of the reaction decreases dramatically probably due to side reactions catalyzed by the excess of AgClO4. The much lower yield of complex 4 may be described to the fact that the electron withdrawing capability of the CF3 groups, makes the b-diketone a poorer electrophile making complex formation more difficult. Another contributing factor may be the

2280

E. Erasmus et al. / Inorganica Chimica Acta 360 (2007) 2277–2283

titanium–fluorine single bond energy of 581 kJ mol1 that is higher than that of the titanium–oxygen single bond (478 kJ mol1) [22] favoring Ti–F bond formation, leading to side products and a lower yield of 4. The synthesized titanocenyl b-diketonato salts are all insoluble in pure water, hexane and ether, slightly soluble in organic solvents such as chloroform, dichloromethane, and acetone and decomposes in ethanol. No decomposition was detected in samples that were stored for up to two years. Solutions in pure and dry acetone or acetonitrile were stable for at least 48 h. 3.2. X-ray structure of 5 For the keto ethyl ester complex analogous to 5, [(C5H5)2Ti(CH3COCHCO–OEt)]+, coordination to the metal via the ‘‘ether’’-type oxygen of the OC2H5 group was suggested [4]. In the crystal structure determination of 5, it was found that ½ðC5 H5 Þ2 TiðCH3 COþ CHCO–OMeÞ ClO4  coordinates through the carbonyl oxygen of the ester group and not through the ether oxygen of the (CO)OMe group. This striking result conclusively proves the interpretations of White [21] who from the IR and NMR spectral data concluded that [(C5H5)2Ti(CH3COCHCO–OEt)]+ must be coordinated through the carbonyl ester oxygen. A perspective view of 5 showing atom labeling is presented in Fig. 1. Crystal data and details of data collection and refinement are given in Table 1. Selected bond lengths and angles for the molecule are listed in Table 2. The asymmetric unit contains one cationic complex [(C5H5)2TiIV(CH3COCHCO(OCH3))]+ and one counter anion ClO4  .

The cation of 5 features a distorted tetrahedral coordination around the Ti atom. The two planar Cp (cyclopentadienyl) rings form an angle of 48.9 between them. The ˚ is not significantly average TiIV–C bond length of 2.362 A shorter than the equivalent bond length in the TiIII complexes. [(C5H5)2TiIII(CH3COCHCOCH3)] showed this ˚ , [(C5H5)2TiIIIaverage bond length to be 2.374 A ˚ and for (C6H5COCHCOC6H5)] showed it to be 2.383 A III ˚ . The [(C5H5)2Ti (CH3COCHCOC6H5)] it was 2.363 A O–Ti–O bond angle of 86.29(7) for 5 is 2–4 larger than that in the [(C5H5)2TiIII(b-diketonato)] complexes. The b-diketonato ligand binds unsymmetrically to the titanium(IV) core with titanium–oxygen bond lengths of ˚ , respectively (see Table 2). These 1.944(2) and 2.009(2) A ˚ shorter than the Ti–O disbond lengths are 0.12–0.14 A tances in the similar [(C5H5)2TiIII(b-diketonato)] complexes ˚ in [(C5H5)2TiIII(CH3COCHC(2.068(5) and 2.068(5) A ˚ in [(C5H5)2TiIIIOCH3)] [5], 2.067 and 2.087 A ˚ in (C6H5COCHCOC6H5)] [6] and 2.078 and 2.085 A III [(C5H5)2Ti (CH3COCHCOC6H5)] [6]). This indicates that the more electron deficient Ti(IV) centre binds oxygen stronger than Ti(III). The longer Ti–O bond length trans to the OCH3 group is consistent with the fact that the O of C@O nearest to the more electronegative OCH3 group of the chelate ring (vOCH3 ¼ 2:64, vCH3 ¼ 2:21) has the smallest trans influence [23]. This is in agreement with the polarization theory, since the oxygen nearest to the more electronegative group (OCH3) will be less polarizable as a result of the electron attracting power of the OCH3 group [24]. ˚, Normal single C–C bond lengths are typically 1.54 A while typical C@C double bonds have a length of ˚ [25]. The large double bond character of the C–C 1.337 A bonds in the chelate ring of the b-diketonato ligand of 5, ˚ , for C13–C14 and 1.354(3) A ˚ for C12–C13, as well 1.401(3) A ˚) as the similar C–O bond lengths (1.277(2) and 1.282(2) A emphasise effective electron delocalisation in the pseudo aromatic core of the b-diketonato ligand. The fact that delocalisation enhances electronic communication between R-groups of the b-diketonato ligand and the Ti centre will be discussed in the electrochemical section below.

Table 2 ˚ ) and angles () for ½ðC5 H5 Þ TiðCH3 COCHCOSelected bond lengths (A 2 ðOCH3 ÞÞþ ClO4  (5) Ti–O1 Ti–O2 O1–C12 O2–C14 O3–C14 O3–C15 C11–C12 C12–C13 C13–C14 Ct1–Ti–Ct2a Fig. 1. A perspective view of ½ðC5 H5 Þ2 TiðCH3 COCHCOðOCH3 ÞÞþ ClO4  (5) showing atom labelling. The ellipsoids are drawn at a 50% probability displacement level.

1.9435(15) 2.0094(15) 1.282(2) 1.277(2) 1.315(2) 1.462(3) 1.495(3) 1.354(3) 1.401(3) 133.74

O1–Ti–O2 C12–O1–Ti C14–O2–Ti C14–O3–C15 O1–C12–C13 O1–C12–C11 C13–C12–C11 C12–C13–C14 O2–C14–O3 O2–C14–C13 O3–C14–C13

86.29(7) 133.16(14) 127.80(13) 117.13(17) 122.26(18) 115.69(18) 122.02(18) 124.48(19) 117.62(18) 125.19(18) 117.18(18)

a Ct1 = centroid of plane defined by C1–C5, Ct2 = centroid of plane defined by C6–C10.

E. Erasmus et al. / Inorganica Chimica Acta 360 (2007) 2277–2283

3.3. Group electronegativity of OMe It has previously been shown that accurate apparent group electronegativities, vR, of R for esters of the type R(C@O)(OCH3) can be obtained from a linear fit between vR and carbonyl IR stretching frequency [8,9]. The straight line generated by the fit of vR and m(C@O) (Fig. 2) fits the equation m(C@O)R = 74.5vR + 1560.5 and was used to determine the effective or apparent group electronegativity

CF3 Cl

-200

ν(C=O) / cm

-1

OCH3

CCl3 CHCl2 CH2Cl CH2Br

1750

H

1700

-400

CH3 C6H5

-600 CF3

Fc C6H5

-800

OCH3

CH3 Fc

1650 1.5

E 0' vs Fc/Fc+ / mV

0

1800

of R = OCH3 as 2.64 on the Gordy scale. This value corresponds reasonably well with vOCH3 ¼ 2:68 as calculated by Huheey [26] using compounds of the type R1R2C@O. The effective vOCH3 value is not the same for all complexes. Utilising PO bond vibrations in compounds of the type R1R2R3P=O, vOCH3 was reported to be 2.89 [27]. 3.4. Electrochemistry All the complexes (4–8) exhibit a characteristic one electron electrochemically reversible TiIII/TiIV couple (see Figs. 3 and 4) with DE < 84 mV. The ½ðC5 H5 Þ2 TiðCH3 þ COCHCOFcÞ ClO4  complex exhibits an additional reversible ferrocenyl ligand redox couple with DE = 77 mV. Peak current ratios (ipc/ipa for the titanium and ipa/ipc for the ferrocenyl couple) for both redox processes is nearly 1. Plots of peak current versus m0.5 (m = scan rate) for all redox processes was linear for all complexes. This result is demonstrated for compound 8 in the inset of Fig. 4 and indicated that the electrochemistry was essentially diffusion controlled. The electrochemical scheme associated with the ferrocenyl complex is

-1000

2

2.5

3

2281

[(C5H 5)2Ti III(CH3COCHCOFc)]

3.5

χR / Gordy scale

+e -

Fig. 2. Left axis: The linear relationship between Gordy scale group electronegativity of an R group, vR, and carbonyl stretching frequencies of the indicated esters, RC(O)OMe. Calibration data from Refs. [8,9] allowed the determination of vOMe from m(C@O) = 1757 cm1 for H3COC(O)OMe. 0 Right axis: The relationship between the formal reduction potential (E0 ) of þ III/IV the Ti couple of the complexes ½ðC5 H5 Þ2 TiðCH3 COCHCORÞ ClO4  and vR of the R groups on the b-diketonato ligands.

-eEo' = -903 mV

[(C5H5)2TiIV(CH3COCHCOFc)]+ +e - -eEo' = 296 mV [(C5H5)2Ti IV(CH 3COCHCOFc+)]2+

200

70

150

50

Free Fc

100

R = C6H5 (6) Free Hfca

50

0

R = CH3 (7)

Current / μA

Relative current / μ A

R = OCH3 (5)

I / μA

R = CF3 (4)

Free Fc ferrocenyl fragment oxidation 0

30

5

10 15 20

1/2

(scan rate)

-1 1/2

/ (mV/s )

10 -10 ferrocenium fragment reduction

-30

R = Fc (8)

TiIII/TiIV

-50 -1500 -50 -1300

40 30 20 10 0

-800

-300

200

Potential / mV vs Fc/Fc

700 +

Fig. 3. Cyclic voltammograms of 1.0 mmol dm3 solutions of ½ðC5 H5 Þ2 TiðCH3 COCHCORÞþ ClO4  , free ferrocene (Fc) and free CH3COCH2COFc in CH3CN containing 0.1 mol dm3 (nBu4N)(PF6) as supporting electrolyte on a glassy carbon working electrode at 25.0 C and a scan rate of 100 mV s1.

-1000

-500

0

500

1000

Potential / mV Fig. 4. Cyclic voltammograms of a 1 mmol dm3 solution of ½ðC5 H5 Þ2 TiðCH3 COCH2 COFcÞþ ClO4  (8), in 0.1 mol dm3 (nBu4N)(PF6)/CH3CN containing free ferrocene as internal standard at scan rates of 50, 100, 150, 200 and 250 mV s1 on a glassy carbon electrode at 25.0 C vs. Fc/Fc+. Inset: The relationship between Ti anodic peak currents and (scan rate)1/2 obeys the Randles–Sevcik equation [28] for 8. Graph slope = 2.14 lA (mVs1)1/2. Scans were initiated at 200 mV in the positive direction.

2282

E. Erasmus et al. / Inorganica Chimica Acta 360 (2007) 2277–2283

Table 3 Cyclic voltammetry data (potentials vs. Fc/Fc+) of 1.0 mmol dm3 CH3CN solutions of the indicated compounds in 0.1 mol dm3 (nBu4N)(PF6)/CH3CN measured on a glassy carbon working electrode at a scan rate of 100 mV s1 0

Complex

vRa (Gordy scale)

vCO (cm1)

Epa (mV)

DEp (mV)

E0 (mV)

ipc (lA)

ipc/ipa

4, R = CF3 5, R = OCH3 6, R = C6H5 7, R = CH3 8, R = Fc Ferrocenyl fragment from 8 Free CH3COCH2COFc Free ferrocene

3.01 2.64b 2.21 2.34 1.87

1593 1561 1547 1543 1513

596 766 820 822 864 334c 278c 37c

78 78 78 84 78 77 79 74

635 805 859 864 903 296c 239c 0c

20.02 20.08 23.66 19.11 21.49 21.53e 24.09e 26.28e

0.94 0.84 0.81 0.96 0.88 0.96d 0.97d 0.98d

Data tabulated are of the TiIII/IV couple unless stated otherwise. a Values from [8,9]. b Result from this study. c Potentials for the ferrocenyl/ferrocenium couple. d ipa/ipc value. e ipa value.

Electrochemical data of the mono-b-diketonato titanium complexes 4–8 are summarized in Table 3. From Table 3 0 and Fig. 2, it can be seen that the E0 for the TiIII/IV couple decreases with decreasing vR value of the b-diketonato side groups. This is expected, because the more electron donating the R group becomes, the more negative the titanium(IV) centre will become and accordingly the easier it will be reduced. Even though the relationship between vR and the formal reduction potential of the titanium centre is not linear, Fig. 2, it still demonstrates good electronic communication between the titanium centre and each of the pendent R-groups to the b-diketonato ligand.

1.6

Q/NF

Fc/Fc+

TiIV/TiIII

1.2 0.8

0.5

0.4

I

I

0.0 0

2000

Current / μA

1.0

0.0 6000

4000

Relative time / s Fig. 5. Coulometry experiments on ½ðC5 H5 Þ2 TiðCH3 COCH2 COFcÞþ ClO4  (8) indicate that the overall amount of electrons transferred per molecule during the oxidation of the ferrocenyl group and the reduction of TiIV in 8 is n = Q/NF = 1 (left axis). The current–time response is shown on the right axis.

A peculiarity of the present titanium system was observed when complex (8) was studied. Initial studies showed that peak currents for the reversible TiIII/IV could be 4–5 times less than for the ferrocenyl/ferrocenium couple. To come up with comparable peak currents for these two couples, extreme electrode care had to be resorted to. A very well polished electrode would give the same peak currents for the titanium and ferrocenyl couples only for the first scan at, e.g. 100 mV s1 scan rate. To get the same current for the ferrocenyl and titanium couples for the ensuing 200 mV s1 and other scans, the electrode had to be polished between successive scans. The ferrocenyl group of the (CH3COCHCOFc) ligand was found to be electroactive at potentials 57 mV higher in the titanium complex than for the free b-diketone. This result clearly showed that the electropositive Ti centre withdraws electron density from the b-diketonato ligand. In another study focussing on [RhI(CH3COCHCOFc)(CO)2], the ferrocenyl couple was found to be shifted to 190 mV versus Fc/Fc+. This is 49 mV lower compared to the free ligand, emphasising the difference between the electron withdrawing properties of the electropositive TiIV centre over the electron donating properties of the electron rich Rh(I) centre. A one electron transfer process for both the TiIII/IV couple and the ferrocene/ferrocenyl couple of 8 was confirmed by bulk electrolysis for both the reduction of Ti(IV) at an applied potential of 1280 mV versus Fc/Fc+ and for the oxidation of the ferrocenyl fragment at an applied potential of 520 mV versus Fc/Fc+ (see Fig. 5 and Table 4).

Table 4 Bulk electrolysis results for the complex ½ðC5 H5 Þ2 TiðCH3 COCH2 COFcÞþ ClO4  (8) Couple III

IV

Ti /Ti Fc/Fc+

Mass (mg)

105 N (mol)

Q (C)

Q/NF

n

Eexp (mV) vs. Fc/Fc+

8.5 8.5

1.56 1.56

1.497 1.511

0.99 1.00

1 1

1280 520

Electrolysis was preformed on N mol of the complex in 0.1 mol dm3 (nBu4N)(PF6)/CH3CN at 25.0 C, using a carbon sheet working electrode. Q = total charge collected and n = number of electrons transferred per molecule. Eexp indicates the potential at which the oxidation of the ferrocenyl ligand or the reduction of the Ti in 8 was performed.

E. Erasmus et al. / Inorganica Chimica Acta 360 (2007) 2277–2283

The formal reduction potentials obtained for (6, R = C6H5) and (7, R = CH3) during this study, 0.859 and 0.864 V, respectively, is mutually consistent with data published by Bond and co-workers, namely 0.85 and 0.86 V versus Fc/Fc+, respectively [5]. We can now, however, ammend the conclusion drawn by these researchers. The TiIII/IV couple will only be independent of the side groups of the b-diketonato ligand in complexes of the type [(C5H5)2Ti(RCOCHCOR 0 )] if the group electronegativities of the R and R 0 side groups are very close to each other. However, if they differ substantially as was the case in this study (vFc = 1.87 and vCF3 ¼ 3:01), the formal reduction potential of the TiIII/IV couple can differ as much as 635  (903) = 268 mV. 4. Conclusions New titanocene complexes of the type ½ðC5 H5 Þ2 Tiþ ðCH3 COCHCORÞ ClO4  with R = CF3, OMe and Fc were synthesised and characterised. A single crystal X-ray determination of the structure of [(C5H5)2Ti(CH3COCHCO–OMe)]+ClO4 indicates that the b-ketoester ligand is bonded to the Ti centre via the carbonyl oxygen and not the ‘‘ether’’-type oxygen of the ester group. This implies that the bonding mode of the b-ketoester is the same as that for b-diketonato ligands. Electrochemical experiments on the [(C5H5)2Ti(CH3COCHCOR)]+ complexes revealed a chemically and electrochemically reversible one electron TiIII/IV couple at a potential dependent on the apparent group electronegativity of the R group of the b-diketonato ligand. Acknowledgements Financial assistance by the South African National Research Foundation under Grant No. 2054243 and the Central Research Fund of the University of the Free State is gratefully acknowledged. We acknowledge A. Kuhn for help in providing crystals suitable for X-ray diffraction of complex 5. Appendix A. Supplementary material CCDC 626381 contains the supplementary crystallographic data for 5. These data can be obtained free of charge via http://www.ccdc.cam.ac.uk/conts/retrieving.html, or from the Cambridge Crystallographic Data

2283

Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: (+44) 1223-336-033; or e-mail: [email protected]. Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ica.2006. 11.010. References [1] N.J. Long, Metallocenes: An Introduction to Sandwich Complexes, Blackwell Science, London, 1998, pp. 23, 148–154. [2] J.H. Toney, T.J. Marks, J. Am. Chem. Soc. 107 (1985) 947. [3] G. Doyle, R.S. Tobias, Inorg. Chem. 6 (1967) 1111. [4] G. Doyle, R.S. Tobias, Inorg. Chem. 7 (1968) 2484. [5] A.M. Bond, R. Colton, U. Englert, H. H} ugel, F. Marken, Inorg. Chim. Acta 235 (1995) 117. [6] S.S. Yun, I. Suh, E.H. Kim, S.J. Lee, Coord. Chem. 51 (2000) 219. [7] D.A. White, Inorg. Nucl. Chem. 33 (1971) 691. [8] W.C. du Plessis, J.C. Erasmus, G.J. Lamprecht, J. Conradie, T.S. Cameron, M.A.S. Aquino, J.C. Swarts, Can. J. Chem. 77 (1999) 378. [9] W.C. Du Plessis, T.G. Vosloo, J.C. Swarts, J. Chem. Soc., Dalton Trans. (1998) 2507. [10] D.T. Sawyer, J.L. Roberts, Experimental Electrochemistry for Chemists, Wiley, New York, 1974, p. 54. [11] D.H. Evans, K.M. O’Connell, R.A. Peterson, M.J. Kelly, J. Chem. Educ. 60 (1983) 291. [12] G.A. Mabbott, J. Chem. Educ. 60 (1983) 697. [13] G. Gritzner, J. Kuta, Pure Appl. Chem. 56 (1984) 461. [14] APEX2 (Version 1.0-27), Bruker AXS Inc., Madison, Wisconsin, USA, 2005. [15] COSMO, Version 1.48, Bruker AXS Inc., Madison, Wisconsin, USA, 2003. [16] SAINT-PLUS, Version 7.12 (including XPREP), Bruker AXS Inc., Madison, Wisconsin, USA, 2004. [17] SADABS, Version 2004/1, Bruker AXS Inc., Madison, Wisconsin, USA, 1998. [18] A. Altomare, M.C. Burla, M. Camalli, G.L. Cascarano, C. Giacovazzo, A. Guagliardi, A.G.G. Moliterni, G. Polidori, R. Spagna, J. Appl. Cryst. 32 (1999) 115. [19] L.J. Farrugia, J. Appl. Cryst. 32 (1999) 837. [20] G.M. Sheldrick, SHELXL97 Program for Crystal Structure Refinement, University of Go¨ttingen, Germany, 1997. [21] K. Brandenburg, H. Putz, DIAMOND, Release 3.1a. Crystal Impact GbR, Bonn, Germany, 2005. [22] O. Kubaschewski, E.L. Evans, C.B. Alcock, Metallurgical Thermochemistry, 4th edn., Oxford, UK, 1967, Table A, p. 304. [23] J.G. Leipoldt, L.D.C. Bok, J.S. van Vollenhoven, A.I. Pieterse, J. Inorg. Nucl. Chem. 40 (1978) 61. [24] A.A. Grinberg, Acta Physiochim. USSR 3 (1935) 573. [25] R.C. Weast, Handbook of Chemistry and Physics, 63th edn., The Chemical Rubber Co., Ohio, pp. F180–F181. [26] J.E. Huheey, J. Phys. Chem. 69 (1965) 3284. [27] P.R. Wells, Prog. Phys. Org. Chem. 6 (1968) 111. [28] A.J. Bard, L.R. Faulkner, Electrochemical Methods: Fundamentals and Applications, Wiley, New York, 1980 (Chapter 6).