Intramolecular communication and electrochemical observation of the 17-electron ruthenocenium cation in fluorinated ruthenocene-containing β-diketones; Polymorphism of C10H21 and C 10F21 derivatives

Article abstract of DOI:10.1039/c3nj00354j

Ruthenocene-containing β-diketones, RcCOCH₂COR with Rc = ruthenocenyl and R = C₁₀F₂₁ (1), CF₃ (2), C₆F₅ (3), C₁₀H₂₁ (4), CH₃ (5) and C₆H₅ (6), were synthesised by Claisen condensation of the appropriate methyl ester with acetylruthenocene, and their spectroscopic, electrochemical and thermal properties were compared. A new synthetic route utilising 1,2,3-benzotriazol-1-ylethanone (9) or 1,2,3-benzotriazol-1-yl(ruthenocenyl)methanone (10) as a reactant, rather than the conventional esters, was found to be more efficient for β-diketone synthesis. The apparent acid dissociation constants, pK_a′, of the new ruthenocene-containing β-diketones are 7.14(4) (1, R = C ₁₀F₂₁), 9.92(3) (3, R = C₆F₅) and 10.06(2) (4, R = C₁₀H₂₁). Peak anodic potentials of the ruthenocenyl group of 1-6, pK_a′ values and the FTIR ν(CO) stretching frequencies of the precursor esters, RCOOCH₃, correlated linearly with the Gordy scale group electronegativity, χ_R, of the C₁₀F₂₁ (χC₁₀F₂₁ = 3.04), C ₆F₅ (χC₆F₅ = 2.46), C ₁₀H₂₁ (χC₁₀H₂₁ = 2.43) and other R-groups. An electrochemical study in the non-interacting solvent and electrolyte system CH₂Cl₂/0.1 mol dm^-3 [N( ⁿBu₄)][B(C₆F₅)₄] revealed electrochemically irreversible one-electron transfer Rc/Rc⁺ couples in the potential range 650 < E_pa < 1110 mV. Strikingly, reduction of the unstable monomeric 17 electron species, [Rc⁺(C ₅H₅)(C₅H₄COCH₂COR)], was clearly observed for all β-diketones possessing fluorinated R-groups, even at slow (100 mV s^-1) scan rates. The enol isomer of the fluorinated β-diketones had >90% abundance under the conditions of study and the first order rate constant of enol to keto conversion varied between 220 and 50 000 s^-1 depending on solvent (CDCl₃ or CD₃CN) and R-groups. Thermal analysis (DSC) of 1 and 4 showed no liquid crystalline mesophase behaviour but definite polymorphism was observed. β-diketones 1 and 4 exist as low temperature polymorphs below 42°C or 12°C respectively. The high temperature polymorphs converted to isotropic liquids at 83°C (compound 1) or 52°C (compound 4).

Full text of DOI:10.1039/c3nj00354j

NJC
View Article Online
View Journal | View Issue
PAPER
Intramolecular communication and electrochemical
observation of the 17-electron ruthenocenium cation
in fluorinated ruthenocene-containing b-diketones;
polymorphism of C10H21 and C10F21 derivatives†
Cite this: NewJ.Chem., 2013,
37, 2862
Elizabeth Erasmus and Jannie C. Swarts*
Ruthenocene-containing b-diketones, RcCOCH2COR with Rc = ruthenocenyl and R = C10F21 (1), CF3 (2),
C6F5 (3), C10H21 (4), CH3 (5) and C6H5 (6), were synthesised by Claisen condensation of the appropriate
methyl ester with acetylruthenocene, and their spectroscopic, electrochemical and thermal properties were
compared. A new synthetic route utilising 1,2,3-benzotriazol-1-ylethanone (9) or 1,2,3-benzotriazol-1-
yl(ruthenocenyl)methanone (10) as a reactant, rather than the conventional esters, was found to be more
0
efficient for b-diketone synthesis. The apparent acid dissociation constants, pKa , of the new ruthenocene-
containing b-diketones are 7.14(4) (1, R = C10F21), 9.92(3) (3, R = C6F5) and 10.06(2) (4, R = C10H21). Peak
0
anodic potentials of the ruthenocenyl group of 1–6, pKa values and the FTIR n(CQO) stretching frequencies
of the precursor esters, RCOOCH3, correlated linearly with the Gordy scale group electronegativity, wR, of the
C10F21 (wC
= 3.04), C6F5 (wC F = 2.46), C10H21 (wC
= 2.43) and other R-groups. An electrochemical study
10F21
10H21
6
5
in the non-interacting solvent and electrolyte system CH2Cl2/0.1 mol dmꢀ3 [N(nBu4)][B(C6F5)4] revealed
electrochemically irreversible one-electron transfer Rc/Rc+ couples in the potential range 650 o Epa
o
1110 mV. Strikingly, reduction of the unstable monomeric 17 electron species, [Rc+(C5H5)(C5H4COCH2COR)],
was clearly observed for all b-diketones possessing fluorinated R-groups, even at slow (100 mV sꢀ1) scan
rates. The enol isomer of the fluorinated b-diketones had >90% abundance under the conditions of study
and the first order rate constant of enol to keto conversion varied between 220 and 50 000 sꢀ1 depending
on solvent (CDCl3 or CD3CN) and R-groups. Thermal analysis (DSC) of 1 and 4 showed no liquid crystalline
mesophase behaviour but definite polymorphism was observed. b-diketones 1 and 4 exist as low tempera-
ture polymorphs below 42 1C or 12 1C respectively. The high temperature polymorphs converted to isotropic
liquids at 83 1C (compound 1) or 52 1C (compound 4).
Received (in Montpellier, France)
4th April 2013,
Accepted 20th June 2013
DOI: 10.1039/c3nj00354j
www.rsc.org/njc
FcCOCH2COR, have been studied electrochemically to investigate
the intramolecular communication between the ferrocenyl moiety
1. Introduction
Free b-diketones, RCOCH2COR0, are important intermediates in
the organic synthesis of numerous heterocyclic compounds.1 Many
exhibit antioxidant, antibacterial or antitumor activities.2 They, and
a wide range of b-diketonato transition metal complexes, have been
subjected to catalytic,3 kinetic,4 structural,5 electrochemical,6 and
medical studies.7 As far as the metallocene-containing b-diketones
are concerned, due to the reversible one-electron redox properties
of the ferrocenyl group,8 ferrocene (Fc)-containing b-diketones,
and the b-diketonato side groups, R.9 The presence of the
ferrocenyl group led to enhanced b-diketone cytotoxicity against
HeLa and CoLo cancer cell lines,10 and the fluorinated species
FcCOCH2COCF3 was twice as reactive against the platinum-
resistant large human lung carcinoma COR L23/CPR cell line
as cisplatin.10 Enolisation of the ferrocene-containing b-diketone
in solution predominantly generated the Fc–CO–CHQC(OH)–R
tautomer and the kinetics of keto–enol isomerisation have been
reported.11
Much less attention has been paid to derivatives of the
analogues and equally electron-donating metallocene, ruthenocene
(RcH = RuII(C5H5)2), because it is more costly than ferrocene and
derivatisation is synthetically more challenging. The general
organic chemistry of the ruthenocenyl group, Rc, closely resembles
Department of Chemistry, University of the Free State, Bloemfontein, 9300,
Republic of South Africa. E-mail: swartsjc@ufs.ac.za; Fax: +27-51-4446384;
Tel: +27-51-4012781
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c3nj00354j
c
2862 New J. Chem., 2013, 37, 2862--2873
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013

View Article Online
Paper
NJC
that of ferrocene. However, electrochemically, the Rc+/Rc = argon was purchased from Sigma-Aldrich. THF was dried by
[RuIII(C5H5)(C5H4)]+/[RuII(C5H5)(C5H4)] couple cannot be observed refluxing under a nitrogen atmosphere over a sodium wire and
in the same wide range of solvents and supporting electrolytes as distilled directly before use. Acetylruthenocene,21 methyl ferro-
the Fc+/Fc couple of the ferrocenyl group, because usually over- cenoate,5a ruthenocenoyl chloride,22 complexes 2, 5, 617 and
oxidation to a RuIV species occurs and because the 17 electron [N(nBu4)][B(C6F5)4]15 were prepared as described before. Water
RuIII(C5H5)2 species is inherently unstable in any electron-rich was double distilled. Melting points were determined using a
+
environment.12 The weak coordinating and solvating properties of Reichert Thermopan microscope with a Kofler hot-stage and
the solvent CH2Cl2 as well as the weak tendency to form ion pairs are uncorrected. NMR spectra were recorded on a Bruker
of the type Rc+ꢁꢁꢁAꢀ with Aꢀ being the anion of the supporting Avance DPX 300 NMR spectrometer. Chemical shifts are
electrolyte [N(nBu4)][B(C6F5)4]13 or [N(nBu4)][B(C6H3(CF3)2)4]14 reported as d values relative to SiMe4 (0 ppm). IR spectra were
allowed observation of one-electron RcH oxidation to RcH+ rather recorded from neat samples on a Digilab FTS 2000 Fourier
than two-electron oxidation to a Ru(IV) Rc2+ species as well as transform spectrometer utilizing a He–Ne laser at 632.6 nm.
detection of the reduction of the RcH+ species. Once formed, RcH+
Elemental analyses were performed by the University of the
dimerizes on the CV time scale in CH2Cl2–[N(nBu4)][B(C6F5)4] to Free State’s Analytical Chemistry division using a Leco Truspec
produce different low and high temperature dimers,13–15a in Micro instrument for C, H and N analysis, or by Canadian
equilibrium with the parent unstable seventeen electron Rc+ Microanalytical Service, Ltd.
species. A quasi-reversible one-electron oxidation of ruthenocene
has also been observed in Lewis acid–base molten salts.16
2.2. Synthesis
Ruthenocene-containing b-diketones (RcCOCH2COR, with R =
CF3, C6H5, CH3, Rc and Fc) have been synthesised and their
keto–enol isomerisation has been studied.17 The electrochemistry
of these b-diketones in CH3CN–[N(nBu)4][PF6] showed an irrever-
sible two-electron oxidation of the Rc group to most likely form
[(C5H4COCH2COR)(C5H5)RuIVꢁCH3CN]2+.17 Only the b-diketone,
RcCOCH2COFc, was studied in CH2Cl2 utilizing [N(nBu)4]-
[B(C6F5)4] as the supporting electrolyte. A one-electron oxidation
of the ruthenocenyl group was observed but no clear Rc+ reduction
could be observed. However, a one-electron Rc/Rc+ couple was
observed for ruthenocene-containing chalcones RcCOCHCHR
in CH2Cl2–[N(nBu4)][B(C6F5)4];18 they are structurally related to
ruthenocene-containing b-diketones.
In our on-going studies on the cytotoxicity of metallocene
derivatives,10,19 it was consistently observed that fluorinated
compounds show enhanced cytotoxic activity against various
cancer cell lines.10 Other researchers also detected the high
biological reactivity of fluorinated compounds.20 A need arose to
enlarge the array of available fluorinated ruthenocene-containing
b-diketones to further investigate this biological trend and to
determine how other physical properties of such ruthenocene
derivatives may also be influenced by the presence of fluoro
substituents.
2.2.1. RcCOCH2COR complexes 1, 3 and 5. The ruthenocene-
containing b-diketones, RcCOCH2COR with R = C10F21 (1), C6F5
(3) and C10H21 (4), were synthesized, Scheme 1, utilizing the
general Claisen condensation procedure outlined for 1 below.
Care was taken to separate the aldol self-condensation product
from acetylruthenocene.23
1-Ruthenocene-4-perfluoroundecylprop-1,3-dione, 1. Under
Schlenk conditions, acetylruthenocene (0.60 g, 2.19 mmol)
was dissolved in the minimum dry THF (ꢂ1 ml) and stirred
for a few minutes. Lithium diisopropylamide (1.8 mol dmꢀ3
,
1.2 ml, 2.16 mmol) was added slowly, while the solution was
kept cool in an ice-bath and stirred for a further 30 min at room
temperature. The colour changed from yellow to orange. Methyl
perfluoroundecanoate (1.27 g, 2.19 mmol) was added dropwise
while cooling in ice and the solution was left to stir for 16 h at
room temperature. Excess diethyl ether was added and the
precipitate was filtered, and the solid suspended and stirred in
an ice-cold solution of HCl (1 mol dmꢀ3, 100 ml) for 30 min.
The product was extracted with diethyl ether, washed with
water (3 ꢃ 100 ml), dried over MgSO4 and the solvent removed
under reduced pressure. Chromatography of the residue using
hexane : diethyl ether (1 : 1) (Rf = 0.55) as the eluent afforded
1.55 g (86%) pure 1, m.p. 72 1C. dH (300 MHz, CDCl3): enol
isomer (100%) 4.57 (s; 5H; C5H5); 4.89 (t; 2H; 0.5C5H4); 5.17
(t; 2H; 0.5C5H4); 5.59 (s; 1H; CH). FTIR: nmax/cmꢀ1 1716 (CQO),
1615 (CQO). Anal. calc., for C23H11O2F21Ru: C, 33.7, H, 1.4%.
Found: C, 33.3, H, 1.4%.
Towards this end we report here the synthesis, characterisation
and electrochemical properties of RcCOCH2COR compounds with
R = CF3, C6F5 and C10F21, and compare their physical and
chemical properties with the CH3, C6H5 and C10H21 analogues.
A
new synthetic route towards ruthenocene-containing
b-diketones is also demonstrated. The results from a thermal
phase study of the C10-complexes as well as the kinetics of keto–
enol isomerisation are reported.
Characterisation data for 1-ruthenocene-4-(2,3,4,5,6-penta-
fluorophenyl)prop-1,3-dione, 3. Rf = 0.85 in hexane : diethyl ether
(1 : 1), yield 8.3%, m.p. 170 1C. dH (300 MHz, CDCl3): enol
isomer (100%) 4.63 (s; 5H; C5H5); 4.90 (t; 2H; 0.5C5H4); 5.19
(t; 2H; 0.5C5H4); 5.95 (s; 1H; CH). FTIR: nmax/cmꢀ1 1743 (CQO).
Anal. calc., for C19H11O2F5Ru: C, 48.8, H, 2.4%. Found: C,
48.6, H, 2.5%.
2. Experimental
2.1. General
Ruthenocene (Strem), CH3COCl, benzotriazole, aluminum
trichloride and other solid reagents (Merck, Sigma-Aldrich) were
Characterisation data for 1-ruthenocene-4-undecylprop-1,3-
used without further purification. Dry CH2Cl2 stored under dione, 4. Rf = 0.92 in hexane : diethyl ether (1 : 1), yield 23%
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013
New J. Chem., 2013, 37, 2862--2873 2863

View Article Online
NJC
Paper
Scheme 1 Two different synthetic routes for the preparation of ruthenocene-containing b-diketones 1–6. LiDPA = lithium diisopropylamide.
m.p. 60 1C. dH (300 MHz, CDCl3): enol isomer (100%) 0.89 dry THF (2 ml) was added at ꢀ78 1C. The mixture was allowed
(m; 3H; CH3); 1.26 (m; 14H; 7CH2); 1.85 (m; 2H; CH2); 2.26 to warm to room temperature and stirred overnight, before
(m; 2H; CH2); 4.58 (s; 5H; C5H5); 4.78 (t; 2H; 0.5C5H4); 5.14 quenching with water (20 ml). Diethyl ether (60 ml) was added
(t; 2H; 0.5C5H4); 5.64 (s; 1H; CH). FTIR: nmax/cmꢀ1 1735 (CQO), and the organic layer was separated, washed with water (2 ꢃ
1617 (CQO). Anal. calc., for C23H32O2Ru: C, 62.5, H, 7.3%. 50 ml), dried and solvent removed to yield the product, which
Found: C, 61.9, H, 6.9%.
could be used without further purification. Yield 90% (23 mg).
2.2.2. 1-Ruthenocene-4-methylprop-1,3-dione, 5. Complex Characterisation data as per Method A above.
5 was also synthesised utilising a new synthetic methodology
2.2.3. Synthesis of 1,2,3-benzotriazol-1-ylethanone, 9. A
from benzotriazole derivatives 9 or 10 as reactants, Scheme 1. revised procedure of the original Katritzky24 method was used.
To a solution of 1-H-benzotriazole (1 g, 8.39 mmol) in dry
CH2Cl2 (20 ml) were added triethyl amine (1.012 g, 10 mmol)
and acetyl chloride (724 mg, 9.22 mmol) under an argon
Method A. From 1,2,3-benzotriazol-1-ylethanone, 9. Under
Schlenk conditions, acetylruthenocene (100 mg, 0.366 mmol)
dissolved in dry THF (ꢂ1 ml) was added at ꢀ78 1C to a stirred
atmosphere at 0 1C while stirring. The mixture was stirred for
solution of lithium diisopropylamide (1.8 mol dmꢀ3, 0.2 ml,
a further 30 min and then quenched with HCl (2 mol dmꢀ3
,
0.36 mmol) in dry THF (0.5 ml). The resulting mixture was
stirred at ꢀ78 1C for 1 h. A solution of 1,2,3-benzotriazol-1-
ylethanone, 9 (54 mg, 0.33 mmol) in dry THF (0.2 ml) was added at
ꢀ78 1C. The mixture was allowed to warm to room temperature
and stirred overnight, before quenching with water (5 ml). The
mixture was diluted with ether (15 ml) and the organic layer was
separated, washed with water (2 ꢃ 20 ml), dried and the solvent
removed to give 86 mg (82%) of analytically pure 5, m.p. 111 1C.
10 ml). The organic phase was separated, washed with HCl
(2 M, 2 ꢃ 10 ml) and water (2 ꢃ 50 ml). The solvent was
removed to yield 96% (1.3 g) of 9 as white crystals, mp 92–95 1C.
dH (300 MHz, CDCl3): 3.04 (s; 3H; CH3); 7.53 (t, 1H; 0.25C5H4);
7.67 (t, 1H; 0.25C5H4); 8.14 (t, 1H; 0.25C5H4); 8.31 (t, 1H; 0.25C5H4).
FTIR: nmax/cmꢀ1 1705 (CQO), 1664 (CQO). Anal. calc., for
C8H7ON3: C, 59.6, H, 4.4%. Found: C, 59.4, H, 4.5%.
2.2.4. Synthesis of 1,2,3-benzotriazol-1-yl(ruthenocene)-
methanone, 10. To a solution of 1-H-benzotriazole (11 mg,
0.09 mmol) in dry CH2Cl2 (1 ml) were added triethyl amine
(11 mg, 0.11 mmol) and ruthenocenoyl chloride (30 mg, 0.1 mmol)
under an argon atmosphere at 0 1C while stirring. The mixture
was stirred for a further hour and then quenched with HCl
(2 mol dmꢀ3, 1 ml). The organic phase was separated and
successively washed with aqueous HCl (2 mol dmꢀ3, 2 ꢃ 5 ml)
and water (2 ꢃ 5 ml). The solvent was removed to give 28 mg
(84%) of 9 as white crystals, mp 102–105 1C. dH (300 MHz, CDCl3):
4.60 (s; 5H; C5H5); 4.99 (t; 2H; 0.5C5H4); 5.84 (t; 2H; 0.5C5H4); 7.51
(t; 1H; C6H4); 7.66 (d; 1H; C6H4); 8.14 (d; 1H; C6H4); 8.33 (t; 1H;
C6H4). FTIR: nmax/cmꢀ1 1715 (CQO), 1685 (CQO). Anal. calc., for
C17H13ON3Ru: C, 54.3, H, 3.5%. Found: C, 54.1, H, 3.4%.
dH (300 MHz, CDCl3): keto isomer – 2.28 (s, 3H, CH3); 3.71 (s; 2H;
CH2); 4.62 (s; 5H; C5H5); 4.82 (t; 2H; 0.5C5H4); 5.12 (t; 2H; 0.5C5H4);
enol isomer – 2.05 (s, 3H, CH3); 4.58 (s; 5H; C5H5); 4.78 (t; 2H;
0.5C5H4); 5.09 (t; 2H; 0.5C5H4); 5.64 (s; 1H; CH). FTIR: nmax/cmꢀ1
1715 (CQO), 1647 (CQO). Anal. calc., for C14H14O2Ru: C, 53.3, H,
4.5%. Found: C, 53.1, H, 4.4%.
Method B. From 1,2,3-benzotriazol-1-yl(ruthenocenyl)methanone
(10). Under Schlenk conditions, dry acetone (4.6 mg, 0.08 mmol)
dissolved in dry THF (ꢂ1 ml) was added to a solution of lithium
diisopropylamide (1.8 mol dmꢀ3, 0.1 ml, 1.8 mmol) in dry THF
(2 ml), while stirring at ꢀ78 1C. The resulting mixture was
stirred at this temperature for 1 h. A solution of 1,2,3-benzo-
triazol-1-yl(ruthenocene)methanone, 10, (30 mg, 0.08 mmol) in
c
2864 New J. Chem., 2013, 37, 2862--2873
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013

View Article Online
Paper
NJC
0
2.3. Apparent acid dissociation constants, pKa
always assigned an integral value of one. The CH2 signal of the
keto form at 3.5–4.1 ppm was then integrated and the % keto-
isomer was calculated from the relationship (I = integral value):
Ruthenocene-containing b-diketones 1, 3 and 4 were dissolved
in the mixed solvent system water : acetonitrile = 9 : 1 (by
volume) containing 0.1 mol dmꢀ3 NaClO4ꢁH2O to maintain a
constant ionic strength. Acetonitrile was used as co-solvent,
since the b-diketones are insoluble in pure water and since it
was found elsewhere25 that acetonitrile interferes the least with
the determination. UV-spectra of the neutral b-diketone
(0.2 mmol dmꢀ3 b-diketone at pH = 2, pH was adjusted with
HClO4) and the deprotonated b-diketonato (0.2 mmol dmꢀ3
b-diketone at pH = 12, pH was adjusted with NaOH) were
obtained from these dissolved analyte systems. From these
spectra a wavelength, see Table 1, was chosen where the change
in absorbance between the neutral b-diketone and the
b-diketonato anion was convenient and large. Titrations of
10 ml solutions containing b-diketone were then performed
with 0.1 mol dmꢀ3 and 1 mol dmꢀ3 NaOH or with 0.1 mol dmꢀ3
and 1 mol dmꢀ3 HClO4 depending on whether an acidic or a
basic titration was performed. The change in absorbance with
changing pH was measured on a Cary 50 Probe UV/Visible
spectrophotometer utilising an Orion model SA 720 pH meter,
equipped with a glass electrode. The pH meter was calibrated
using buffers at pH 4.00, 7.00 and 10.00. The increase in
ðI of keto signalÞ ꢃ 100
% keto isomer ¼
(2)
ðI of keto signalÞ þ 2ðI of enol signalÞ
Once the % keto isomer was known, a plot of % keto isomer
vs. time was constructed. The rate constant was extracted from
a plot of data according to eqn (3).
ꢀ
ꢁ
% keto isomer at t ¼ 0
% keto isomer at t ¼ t
C0 ꢀ C
Ct ꢀ C
1
1
ln
¼ ln
¼ kobs
t
(3)
with kobs = k1 + k
.
17 The equilibrium constant Kc was obtained
ꢀ1
from the relationship
% enol isomer
% keto isomer
k1
Kc ¼
¼
(4)
(5)
k
ꢀ1
This Kc is applicable to the equilibrium.
k1
RcCOCH2COR
Ð RcCOCHCðOHÞR
k
ꢀ1
Once kobs and Kc were known, k1 and k could be deter-
ꢀ1
k1
0
mined by simultaneously solving kobs = k1 + k and Kc ¼
.
ꢀ1
volume during the titration was no more than 5%. The pKa
k
ꢀ1
values were obtained by fitting the obtained data to eqn (1),25,26
using the least squares fitting program MINSQ.27
2.5. Electrochemistry
Measurements on ca. 2.0 mmol dmꢀ3 solutions of the complexes in
dry air-free dichloromethane containing 0.10 mmol dmꢀ3 tetra-
butylammonium tetrakis(pentafluorophenyl)borate, [N(nBu4)]-
[B(C6F5)4], as the supporting electrolyte were conducted under
a blanket of purified argon at 25 1C utilizing a BAS 100 B/W
electrochemical workstation interfaced with a personal computer.28
A three electrode cell, which utilized a Pt auxiliary electrode, a
glassy carbon working electrode and an in-house constructed
0
AHA10ꢀpH þ AA10ꢀpK
a
AT
¼
(1)
0
10ꢀpH þ 10ꢀpK
a
In eqn (1), AT = total absorption, AHA = absorption of the
neutral b-diketone and AA = absorption of the deprotonated
(basic) b-diketonato anion.
2.4. Isomerisation kinetics
2.4.1. Monitoring of the conversion of the enol- to keto- Ag/AgCl reference electrode29 were employed. Temperature was
isomer. The slow formation of keto isomers from an enol kept constant within 0.5 1C. Experimentally potentials were refer-
enriched solution until the equilibrium position was reached enced against a Ag/AgCl reference electrode, but results are pre-
1
was monitored by H NMR following the disappearance of the sented referenced against free ferrocene (FcH) as an internal
methine proton at ca. 5.5–6.1 ppm and the appearance of the standard as suggested by IUPAC.30 To achieve this, each experiment
CH2 signal of the keto form at ca. 3.5–4.1 ppm with time was performed first in the absence of ferrocene and then repeated
utilising a Bruker Avance DPX 300 NMR spectrometer.
in the presence of o1 mmol dmꢀ3 ferrocene. Data were then
2.4.2. Calculation of % keto isomer and determination of manipulated on a spreadsheet to set the formal reduction poten-
the value of Kc. Integration of the 1H NMR spectra was done in tials of the FcH/FcH+ couple at 0 V. Under our conditions data of
such a way that the enol methine proton at ca. 5.5–6.1 ppm was FcH/FcH+ drifted within a series of experiments so each CV or LSV
Table 1 pKa values (determined at lexp) and molar extinction coefficients, e (in units of dm3 molꢀ1 cmꢀ1) of RcCOCH2COR complexes in the mixed solvent CH3CN
0
(10%)–H2O (90%) at 25 1C; b-diketonato = [RcCOCHCOR]ꢀ
0
R
wR
pKa
lexp/nm (e)
b-Diketone lmax (e)
b-Diketonato lmax (e)
Ref.
1: C10F21
2: CF3
3: C6F5
4: C10H21
5: CH3
6: C6H5
7: Rc
3.04
3.01
2.46
2.43
2.34
2.21
1.87
1.89
7.14 (4)
307 (3915)
315 (—)
320 (2078)
312 (9809)
319 (—)
365 (—)
327 (—)
255 (2718)
5960 (260)
260 (2079)
250 (9809)
16 890 (258)
13 610 (333)
5840 (259)
5500 (430)
255 (3915)
315 (—)
320 (3867)
250 (10 563)
318 (—)
347 (—)
331 (—)
This work
17
This work
This work
17
17
17
17
7.36 (3)
9.92 (3)
10.06 (2)
10.22 (1)
11.31 (4)
12.9
8: Fc
12.7
365 (—)
339 (—)
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013
New J. Chem., 2013, 37, 2862--2873 2865

View Article Online
NJC
Paper
was corrected individually. The FcH/FcH+ couple has an E10 = methyl esters vs. previously determined wR values of a series of
421 mV vs. Ag/AgCl with DE = 77 mV and ipc/ipa = 0.99. other R-groups Fig. 1.9,25 wR values of the appropriate R groups were
Successive experiments under the same experimental condi- then read off to be wC
= 3.04, wC
= 2.43 and wC F = 2.46
10H21
6 5
10F21
tions showed that all formal reduction and oxidation potentials respectively. Table 1 summarises wR values relevant to this study.
were reproducible within 5 mV.
A least squares fit of the data showed that n(CQO) is
dependent on wR, according to the equation
2.6. DSC phase studies
n(CQO) = 74.932wR + 1559.3; R2 = 0.9983.
(6)
Phase transitions of the two long chain ruthenocene-containing
b-diketones, 1 and 4 were investigated by Differential Scanning
Calorimetry (DSC) to obtain enthalpies of phase changes. Samples
of 5–10 mg were heated and cooled at rates of 10 1C minꢀ1 between
ꢀ60 and 125 1C using a TA Instruments DSC 10 thermal analyser
fitted with a Du-Pont Instruments mechanical cooling accessory
and a TA Instruments Thermal Analyst data processing unit. Three
successive heating and cooling cycles were recorded. The last two
heating and cooling cycles exhibited thermal events within 95%
reproducibility.
A linear relationship between n(CQO) and wR is only possi-
ble if there is good communication between R and the carbonyl
group in the esters RCOOMe.
0
3.3. pKa determination
0
The apparent acid dissociation constants, pKa , of the new
b-diketones, 1, 3 and 4 were determined by spectroscopically
3. Results and discussion
3.1. Synthesis and characterisation
Ruthenocene-containing b-diketones RcCOCH2COR with R =
[C10F21 (1), CF3 (2), C6F5 (3), C10H21 (4), CH3 (5), C6H5 (6)] were
prepared by lithium diisopropylamide-induced Claisen conden-
sation of acetylruthenocene and an ester containing the desired
R-group (Scheme 1). The yields obtained varied (see the Experi-
mental section), but the general trend that was observed is that
the greater the electronegativity of the R group, the higher
the yield obtained. This observation is consistent with more
electron density being removed from the carbonyl group of
RCOOMe, making them more susceptible to nucleꢀophilic
attack from the in situ generated nucleophile RcCOCH2
.
An alternative method of obtaining ruthenocene-containing
b-diketones was also explored. This method, pioneered by Katritzky
et al.24,31 but hitherto unexplored for metallocene derivatives,
involves the reaction of an acid chloride with benzotriazole to give
an N-acylbenzotriazole. The obtained N-acylbenzotriazole can in
turn be converted into a b-diketone under basic conditions.
To investigate the applicability of this general method to
ruthenocene-containing b-diketones, RcCOCH2COCH3 (5) was
synthesised according to this method. Acetylchloride and
ruthenocenylchloride were reacted with benzotriazole in sepa-
rate reactions to liberate the N-acylbenzotriazoles 9 and 10
respectively in high yields. By reacting 9 and 10 with acetyl-
ruthenocene or acetone under basic conditions (Scheme 1),
RcCOCH2COCH3, 5 was obtained in yields of 82 or 90%
respectively. These yields are much higher than the 43% overall
yield obtained from the Claisen condensation of ethyl acetate
and acetylruthenocene.17
3.2. Group electronegativities
The Gordy scale group electronegativities, wR,32 for the R-groups
C10F21, C10H21 and C6F5 are unknown. Hence they were obtained by
inserting the value of the carbonyl stretching frequency of the
appropriate methyl esters, RCOOMe (R = C10H21, C10F21 and C6F5)
into the linear plot of known carbonyl stretching frequencies of
Fig. 1 Relationship between Gordy scale wR values and pKa0, Epa of the Rc/Rc+
couple (enol signal) of RcCOCH2COR compounds as well as the carbonyl
stretching frequency for esters of the type RCOOCH3. R-groups are indicated
on the graphs. Points X (keto signal) andE (enol signal but not fitting the trend
set by other compounds) were not used to fit the line.
c
2866 New J. Chem., 2013, 37, 2862--2873
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013

View Article Online
Paper
NJC
0
monitoring UV/vis changes during an acid–base titration of pKa values for ruthenocene-containing b-diketones may be
these b-diketones. The term ‘‘apparent’’ acid dissociation estimated from the equation
0
constant with symbol pKa is utilised because no attempt
pKa = ꢀ4.7447wR + 21.584; R2 = 0.9959.
(7)
0
was made to separate pKa values for the keto- and enol isomer
of each b-diketone. A suitable wavelength at which each
pKa could be determined was identified from the overlay
3.4. Keto–enol isomerisation kinetics
0
When the ruthenocene-containing b-diketones are stored in the
solid state, they convert over several weeks 100% to the enol
form. When such a sample is dissolved in a suitable solvent,
during a slow kinetic process, some of the enol isomer
is converted to the keto isomer to eventually reach a dynamic
solution equilibrium position. Keto–enol isomerisation
kinetics of b-diketones 1–8 (complex 7 has R = Rc and 8 has
R = Fc; the synthesis are described in ref. 17) was studied by
1H NMR in CD3CN; data for the new b-diketones 1, 3 and 4 were
also obtained in CDCl3 at 20 1C, see Table 2. Compounds 7 and
8 were studied before17 but their data are included in this study
to contrast the effect of strong electron-withdrawing fluori-
nated b-diketone substituents with strong electron-donating
metallocenyl substituents.
Time-based plots illustrating the increase in percentage keto
isomer as the equilibrium position is approached after dissol-
ving a completely enolized b-diketone at 20 1C in CD3CN are
shown in Fig. 4. The insert highlights the validity of a first-order
treatment of the data applicable to 6 as an example. The
measured half-lives, t1/2, of the reactions suggested an activity
series of isomerisation in CD3CN as follows (R values in
brackets; 3 is not included in the activity series because Kc is
so large, no meaningful measurements can be made):
(fastest) 7(Rc) > 6(C6H5) > 8(Fc) > 5(CH3) > 2(CF3) > 1(C10F21) >
4(C10H21) (slowest).
spectra of the deprotonated b-diketonato anion and the
neutral b-diketone species; Fig. 2 shows this for 3. Fig. 3 shows
the absorbance–pH profiles used for the pKa determina-
tions of 10, 3 and 4. Results are tabulated in Table 1. The
listed pKa -values were obtained from the non-linear least
squares fit of the UV/Vis absorbance–pH data (Fig. 3) inserted
into eqn (1).
0
The Gordy scale group electronegativity of the R groups of
0
the ruthenocene-containing b-diketones and their pKa values
are inversely proportional to each other, Fig. 1, because the
pseudo-aromatic b-diketone becomes more electron-deficient
with R-substituents with higher wR values. This means
it removes electron density from the b-diketone backbone,
making it more acidic according to the reaction
Ka
RcCOCH2COR þ H2O
Ð
½RcCOCHCORꢄꢀ þ H3Oþ.
Also, once a proton is removed from the b-diketone to
generate the b-diketonato anion, [RcCOCHCOR]ꢀ, a more electro-
negative R-group such as C10F21 will stabilize the negative
charge on the b-diketonato anion better than a more electron-
donating R-group. This again implies a smaller pKa in the above
equation. A linear least squares fit of the data of Fig. 1 shows
In CDCl3 this activity series is
(fastest) 7(Rc) > 8(Fc) > 5(CH3) > 1(C10F21) > 3(C6F5) > 6(C6H5) >
2(CF3) > 4(C10H21) (slowest).
In general, the equilibrium constant, Kc, decreases as the
group electronegativity of the R-group decreases. However, the
rate constants kobs, k1 and kꢀ1 have a tendency to increase with
decreasing wR. In both solvents, compound 4 with R = C10H21
isomerises the slowest and 7 with R = ruthenocenyl isomerises
the fastest. The ratio between the slowest and the fastest half-
lives for CDCl3 is 49 856/220 = 226 but for CD3CN it is only
18 381/4620 = 4; this is a difference of two orders of magnitude.
These differences are attributed to CD3CN being capable of
interacting (solvate) with polar compounds to a much large
degree than CDCl3.
Fig. 2 UV/vis spectra of 3 (ꢁ ꢁ ꢁ, at pH 2.00) and 3ꢀ (—, at pH 12.00) at 25 1C.
I = 0.1 mol dmꢀ3 NaClO4 in water containing 10% CH3CN.
From Table 2 it is evident that as the group electronegativity
of the R-group, wR, increases, DG becomes more negative for the
keto–enol equilibriums. This implies that thermodynamically,
in solution, the enol isomer is the most stable species for
compounds with electron-withdrawing R-groups. This is also
evident from the % enol isomer present.
3.5. Electrochemistry
The electrochemistry of the ruthenocene-containing b-diketones
RcCOCH2COR, where R = CF3 (2), CH3 (5), Rc (7) and Fc (8) in the
coordinating solvent and electrolyte system CH3CN–[N(nBu4)][PF6]
Fig. 3 Absorbance–pH profiles of 2.0 mmol dmꢀ3 solutions of RcCOCH2COR
compounds 1,
3 and 4 in 10% acetonitrile–90% water at ionic strength
0.100 mol dmꢀ3 (NaClO4) at 25 1C; R groups are indicated on the graphs.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013
New J. Chem., 2013, 37, 2862--2873 2867

View Article Online
NJC
Paper
Table 2 Gordy scale group electronegativities, equilibrium constants,a Gibbs energies (kJ molꢀ1), half lives and kinetic rate constants (sꢀ1) for the keto–enol
equilibriuma of RcCOCH2COR in CD3CN and CDCl3 at 20 1C
DGb
% Enol at equilibrium t1/2 (s) 105 kobs
105 k1
106
k
Ref.
a
c
d,e
d
d
Comp.: R-group (solvent used) wR
Kc
ꢀ1
1: C10F21 (CD3CN)
1: C10F21 (CDCl3)
2: CF3 (CD3CN)
2: CF3 (CDCl3)
3: C6F5 (CD3CN)
3: C6F5 (CDCl3)
4: C10H21 (CD3CN)
4: C10H21 (CDCl3)
5: CH3 (CD3CN)
5: CH3 (CDCl3)
6: C6H5 (CD3CN)
6: C6H5 (CDCl3)
7: Rc (CD3CN)
7: Rc (CDCl3)
3.04
8.90
10.6
14.6
12.3
ꢀ5.33
89.9
91.4
93.6
92.5
16 500
2659
4.24(4)
26.06(3)
4.60(1)
2.17(5)
3.78
23.8
4.31
2.01
4.24
22.5
2.95
1.6
This work
This work
This work
17
This work
This work
This work
This work
This work
17
This work
17
This work
17
3.04
3.01
3.01
ꢀ5.76
ꢀ6.53
ꢀ6.11
15 065
31 900
2.46 >19 f
oꢀ7.16 f >95 f
—
—
—
—
f
f
f
f
2.46
2.43
2.43
2.34
2.34
2.21
2.21
1.87
1.87
1.87
1.89
12.5
5.54
ꢀ6.15
ꢀ4.16
ꢀ3.72
0.16
92.6
84.7
82.1
48.3
68.5
70.2
85.0
42.6
41.2
42.5
47.0
2996
18 381
49 856
8988
700
7296
11 000
4620
220
23.13(2)
3.77(1)
1.39(5)
7.71(1)
98(2)
8.5(4)
6.3(5)
15.0(6)
320(10)
9.1(5)
21.4
3.19
1.14
3.72
68.3
5.97
5.36
6.38
132
17.1
5.80
2.49
39.9
297
25.3
9.42
86.2
1880
52.0
563
4.59
0.93
2.17
2.36
5.67
0.74
0.70
0.74
0.89
ꢀ1.89
ꢀ2.08
ꢀ4.23
0.73
0.87
0.74
8: Fc (CD3CN)
8: Fc (CDCl3)
7600
650
3.90
50.7
17
17
0.29
107(3)
a
b
c
Applicable to eqn (5); Kc = (% enol isomer)/(% keto isomer) at equilibrium (eqn (4)). DG = ꢀRT ln Kc and applicable to eqn (5). t1/2 = (ln 2)/kobs
.
d
e
Obtained by simultaneously solving the equations kobs = k1 + kꢀ1 and Kc = k1/kꢀ1, see the Experimental section. Values in parentheses indicate
f
errors in the last digit(s). Not all errors are known, because the error in Kc is unknown. % The keto content at equilibrium too low for meaningful
measurements.
establish if generation and specifically observation of the reduc-
tion of the 17 electron cation, [(C5H5)RuIII(C5H4–COCH2COR)]+,
in a one-electron transfer step is possible.
Fig. 5 shows the CVs of 5 at 25 and ꢀ5 1C at scan rates of up
to 2000 mV sꢀ1, CVs of 1–7 at a scan rate of 200 mV sꢀ1 are
shown in Fig. 6 and electrochemical data are summarised in
Table 3.
In Fig. 5, two ruthenocene-based oxidation peaks 1 and 2 are
observable for 5 at slow scan rates. Complex 5 exists as a
mixture of the keto- and enol isomers [Rc–CO–CH2–CO–CH3 "
Rc–CO–CHQC(OH)–CH3] in slow equilibrium with each other.
Peak 1 was assigned to the one-electron oxidation of the
ruthenocene fragment of the enol isomer and peak 2 is associated
Fig. 4 Bottom: 1H NMR determined time-based plots showing the conversion
from the enol to keto isomer for the indicated compounds at 293 K in CD3CN.
Top: a kinetic plot of data for 6. The slope corresponds to the observed first order
rate constant kobs = k1 + kꢀ1, see the Experimental section, eqn (3).
were reported by Kemp and co-workers.17 In this medium, the
17eꢀ Rc+ fragment does not survive long enough to detect its
reduction on the CV time scale. They observed in CH3CN–
[N(nBu4)][PF6] a two-electron oxidation step for the RuII centre.
In this study, the electrochemistry of the ruthenocene-containing
b-diketones, 1–8, were studied in the non-coordinating solvent
and electrolyte system, CH2Cl2–[N(nBu4)][B(C6F5)4],33 inter alia to
Fig. 5 Comparative cyclic voltammograms of 2.0 mmol dmꢀ3 RcCOCH2COCH3
(5) solutions in CH2Cl2/0.1 mol dmꢀ3 [N(nBu4)][B(C6F5)4] on a glassy carbon-
working electrode at scan rates of 50, 100, 200, 300, 400, 500 and 2000 mV sꢀ1
and temperatures of 25 1C and ꢀ5 1C.
c
2868 New J. Chem., 2013, 37, 2862--2873
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013

View Article Online
Paper
NJC
In principle b-diketone 5 can in both enol and keto forms also
undergo these dimerisation processes after 5+ has been gener-
ated. This would lead to four dimeric species having closely
overlapping reduction processes. Due to the instability of 5+
2+
and the expected associated dimers, e.g. (5)2 it was not
possible to unambiguously identify these dimeric species
although peaks X and Y in the CVs of 5 (Fig. 5) may be
associated with them. The broadness of peaks X and Y is
consistent with more than one dimeric RuIII species being
reduced at each of these peaks.
In contrast to 5 the CVs of 1, 2, and 3 (Fig. 6) only showed
one oxidation peak (peak 1) during the anodic sweep because 1
exists more than 90% in the enol form at equilibrium (Table 1).
The keto content of these complexes is too low to be clearly
identifiable in the CVs shown. For these complexes, even at low
(200 mV sꢀ1) scan rates, reduction of the monomeric electro-
chemically generated 17 electron species, 1+, 2+ and 3+, is
observed at wave 3 in the cathodic cycle. Clearly the electron-
withdrawing effect of the fluorinated R-groups in 1–3 stabilized
1+–3+ to the extent that their reduction is observable on the CV
Fig. 6 Cyclic voltammograms of 2.0 mmol dmꢀ3 solutions of RcCOCH2COR in
CH2Cl2/0.1 mol dmꢀ3 [N(nBu4)][B(C6F5)4] on a glassy carbon-working electrode at time scale. The Rc/Rc+ couples for all the b-diketones are
25 1C and a scan rate of 200 mV sꢀ1. R is indicated on each CV. For 6, peaks 1 and
electrochemically quasi-reversible (90 o DE o 150 mV) and
2 were not resolved. Wave Y is associated with the enol for 1–7 and X with keto
chemically irreversible (ipc/ipa { 1, Table 3). Electrochemically
form.
reversible one-electron transfer processes are characterized by
DE = Epa ꢀ Epc = 59 mV.15
with the keto isomer. Kemp et al.17 also described the observation
Importantly, peak Y is observed in the cathodic half-wave of
of these two isomers in CH3CN–[N(nBu4)][PF6] and were able to 1–3 but peak X is absent. Upon noting that 1–3 have only enol
study the kinetics of keto–enol conversion with cyclic voltam- isomer that is electrochemically observable in solution (i.e. keto
metry in this electrochemical medium. They conclusively content is too low to detect, see Table 3), it follows that the
showed that the Rc group of the enol isomer is oxidised before reduction peak at Y belongs to the reduction of the enol form and
that of the keto isomer.
not the keto form. This implies that if peak Y is indeed associated
Our use of the non-nucleophilic and non-coordinating sup- with dimeric forms of 1+–3+, then, as for free ruthenocene,13,14
porting electrolyte, [N(nBu4)][B(C6F5)4], allowed us to demon- enol dimeric ruthenocenium cations of the b-diketones are
strate for the first time the reduction of the electrochemically reduced at lower potentials (peaks labelled Y) than their keto
generated unstable 17 electron monomeric b-diketonato counterparts. Keto-dimer reductions are most likely associated
species, (C5H5)RuIII(C5H4OCH2COCH3), 5+, albeit only at a fast with peaks labelled X. For complexes 4–7, oxidation of mixtures of
scan rate (2000 mV sꢀ1, peak 3 in Fig. 5).
keto and enol isomers could be observed electrochemically at
For the parent metallocene, free ruthenocene, electrochemi- waves 1 and 2 (Fig. 6). No reduction of a monomeric Rc+ cation
cally generated [(C5H5)2RuIII]+ is reported to form a dimeric was detectable at 200 mV sꢀ1 scan rates for 4–7 (wave 3 is absent),
species [(C5H5)2RuIII–RuIII(C5H5)2]2+ and a species in which but cathodic waves X and Y would be consistent with reduction of
Ruꢁ ꢁ ꢁ(C5H4) bonds exist in CH2Cl2–[N(nBu4)][B(C6F5)4].13,14 keto and enol dimeric species.
Table 3 Electrochemical data in CH2Cl2/0.1 mol dmꢀ3 [N(nBu4)][B(C6F5)4] (potentials vs. FcH/FcH+) of the ruthenocenyl moiety of 2.0 mmol dmꢀ3 solutions of
RcCOCH2COR complexes at T = 25 1C and a scan rate of 200 mV sꢀ1. Data in brackets are in CH3CN/0.1 mol dmꢀ3 [N(nBu4)][PF6]a
0
Epa1/mV
DEp1/mV
E1ꢅ =mV
ipa1/mA
ipc/ipa
Epa2/mV
ipa2/mA
EpcX/mV
EpcY/mV
b
b
b
1: C10F21
2: CF3
1108
119
1048
10.3
7.3 (—)
2.3
2.5
6.3 (—)
4.9
7.5 (—)
7.6 (—)
9.6 (—)
0.4
0.5 (—)
0.2
—
—
—
—
—
ꢀ646
ꢀ490
ꢀ524
ꢀ464
ꢀ851
ꢀ524
ꢀ979
b
b
b
981 (609)
921
113 (—)
113
924 (—)
864
—
—
b
b
b
3: C6F5
4: C10H21
5: CH3d
5: CH3
6: C6H5
7: Rc
—
—
680c
175c
593c
o0.01
882
1083 (584)
10.1
9.0
186
86
97
ꢀ13
205
472
859 (473)
863
839 (451)
750e (213, 403)
970e (484)a
168c (—)
93c
775c (—)
816c
o0.01 (—)
o0.01
949
3.8
b
b
144c (—)
767c (—)
o0.01 (—)
o0.01 (—)
o0.01 (—)
—
(605)
—
b
b
b
—
—
905 (403, 480)
6.9
—
8: Fca
—
—
1046b (636)a
—
—
b
b
b
b
a
b
Potentials in CH3CN/0.1 mol dmꢀ3 [N(nBu4)][PF6] are from ref. 17. Not possible to determine with confidence due to small intensity, poor
resolution or absence of peaks. At a scan rate of 2000 mV sꢀ1, inaccurate values due to poor intensity of ipc. Values are estimates only. At ꢀ51.
c
d
e
Inaccurate values due to poor resolution between peaks 1 and 2. Values are estimates only.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013
New J. Chem., 2013, 37, 2862--2873 2869

View Article Online
NJC
Paper
The electrochemistry of 8 with R = Fc in CH2Cl2/0.1 mol dmꢀ3
[N(nBu4)][B(C6F5)4] was described in detail by Kemp et al.,17 and
we refrain from elaborating again here on it.
The peak anodic potentials of the first Rc/Rc+ couple of
RcCOCH2COR (peak 1) increase as the Gordy scale group
electronegativity of the R group increases (Fig. 1 and Table 3),
because as the R group becomes more electron-withdrawing,
more electron density is removed from the ruthenocenyl group.
This makes the ruthenium centre relatively more positive and more
difficult to oxidize according to the reaction RuII - RuIII + eꢀ. In
the Fig. 1 plot of wR vs. Epa, wFc = 1.87 cannot be used when plotting
Epa values for 8 because by the time the ruthenocene group gets
oxidised, the ferrocenyl group has already been oxidised to Fc+.
Hence in the Fig. 1 plot, wFc+ = 2.829,25 was used because the
measured ruthenocenyl oxidation at wave 1 is for the species
Fc+COCH2CORc, 8+, and not neutral FcCOCH2CORc, 8. Since the
Rc group now experiences the electron-withdrawing power of Fc+,
the oxidation of Rc is measured at a potential more positive than
for 3 with R = C6F5 having wC F = 2.46, but less positive than for 1
6
5
and 2 (R = C10F21, wC
= 3.04 and R = CF3, wCF = 3.01 respectively),
10F21
3
Table 3 and Fig. 1. The relative trend identifiable from the Fig. 1
Epa vs. wR plot (Epa increase with increasing wR) highlights good
electronic communication between the Rc and the R-groups in
RcCOCH2COR.
A linear least squares fit of the data of Fig. 1 (the point for
R = C10H21 was not included for the fit as it is obviously
not following the trend suggested by the other compounds),
shows ruthenocenyl group Epa values of the enol isomer of
ruthenocene-containing b-diketones may be estimated from
Gordy scale group electronegativities, wR, utilising the equation
Epa = 270.91wR + 228.2; R2 = 0.8964
(8)
The poor R2 value associated with the above equation is
attributed to the positions of the C10F21 and CF3 (and also
C
10H21) compounds that did not follow the trend set by the
Fig. 7 Differential scanning calorimetric thermograms of 1 and 4. A heating and
cooling rate of 10 1C minꢀ1 was used. Three successive heating and cooling cycles
produced identical thermograms; the second cycle is shown. Phase transition
enthalpies are indicated next to each peak label in bold. The temperatures shown
are for the phase transition onset (e.g. melting point or crystallization point) as
well as peak maximum temperature.
other compounds accurately.
3.6. Phase studies of long-chain ruthenocene-containing
b-diketones
Since long alkyl chains are known to induce liquid crystalline
behaviour,34 differential scanning calorimetry (DSC) and
polarised light optical microscopy studies were performed on
1 and 4. No mesophase thermotropic liquid crystal properties Keto–enol tautomerism may also contribute to the broadness of
were found but different solid state polymorphs could be this peak. Melting of the high temperature polymorph sets in at
observed.
Cooling of 1 (R = C10F21) in the isotropic liquid resulted in
83.39 1C (peak M).
The cooling cycle of 4 (Fig. 7, top) shows slow crystallisation
the onset of crystallization at 42.23 1C (right hand base of peak kinetics sets in at 2.71 1C (peak F1). A further solid state
F, Fig. 7, bottom). During the rest of the 10 1C minꢀ1 cooling reorganisation appears to take place at ꢀ36 1C (peak labelled F2).
cycle, no further phase change can be observed. In contrast, on Peak F2 is assigned to a glass transition even though it’s
the heating cycle, 1 showed definite and reproducible poly- detection was not replicated during the heating cycle.
morphism at peaks S1 and S2. At peak S1 (Tmax = 46.16 1C), the
On the heating cycle, polymorphism is associated with peaks
lower temperature polymorph melted but then immediately S1 (Tmax = 17.39 1C) and S2 (Tmax = 21.78 1C) of 4. At T o 17.39 1C,
solidified at peak S2 (Tmax = 57.50 1C) into a high temperature the low temperature polymorph exists. The small amplitude
polymorph. The broadness of the peak is associated with the peak S1 may be associated with the beginning of melting of this
long –C10F21 side chain that requires more time to settle via polymorph. It is followed immediately by cold crystallisation of
rotations and other motions into the optimum conformation. this phase into the crystalline high temperature polymorph at
c
2870 New J. Chem., 2013, 37, 2862--2873
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013

View Article Online
Paper
NJC
peak S2. Melting of the high temperature polymorph sets in at converted to the high temperature polymorph at ca. 12 1C while
52.07 1C (peak M). This transition was confirmed as a melting for the C10F21 complex this transition was initiated at ca. 41 1C.
point by polarized light optical microscopy.
No evidence of any liquid crystal properties for 1 or 4 could
be detected by DSC measurements or variable temperature
Acknowledgements
The authors acknowledge financial support from NRF (grant
81829) and the UFS during the course of this study.
polarized light microscopy. It is concluded that intermolecular
attractive forces between these b-diketone molecules, unlike in
C10 substituted phthalocyanines,34b are not strong enough to
support a liquid crystalline mesophase. Liquid crystalline
mesomorphism may be observed if further functional groups,
such as esters, are introduced in the long chain alkyl substi-
tuents on the methane position of 4.35
References
1 (a) A. L. Baumstark, A. Choudhary, P. C. Vasquez and
M. Dotrong, J. Heterocycl. Chem., 1990, 27, 291–294;
(b) A. P. West, D. Van Engen and R. A. Pascal, J. Org. Chem.,
1992, 57, 784–786; (c) C. P. Felix, N. Khatimi and
A. J. Laurent, J. Org. Chem., 1995, 60, 3907–3909; (d) M. A. P.
Martins, S. Brondani, V. L. Leidens, D. C. Flores, S. Moura,
N. Zanatta, M. Horner and A. F. C. Flores, Can. J. Chem.,
2005, 83, 1171–1177.
4. Conclusion
Three new ruthenocene-containing b-diketones of the type
RcCOCH2COR with R = C10F21 (1), C6F5 (3) and C10H21 (4) were
synthesised by base-catalysed Claisen condensation of acetyl-
ruthenocene with an appropriate ester. A different but more
efficient technique of synthesising metallocene containing
b-diketones utilising 1,2,3-benzotriazol-1-ylmethanone deriva-
tives was developed and demonstrated in the synthesis of
RcCOCH2COCH3, 5. Acid dissociation constants for the new
ruthenocene-containing b-diketones were determined to be
2 (a) A. V. Kelin, Curr. Org. Chem., 2003, 7, 1691–1711;
(b) A. V. Kelin and A. Maioli, Curr. Org. Chem., 2003, 7,
1855–1886.
3 (a) W. R. Cullen, S. J. Rettig and E. B. Wickenheiser,
J. Organomet. Chem., 1989, 370, 141–154; (b) T. E. Banach,
C. Berti, M. Colonna, M. Fiorini, E. Marianucci, M. Messori,
F. Pilati and M. Toselli, Polymer, 2001, 42, 7511–7516.
4 (a) S. S. Basson, J. G. Leipoldt, A. Roodt, J. A. Venter and
T. J. Van der Walt, Inorg. Chim. Acta, 1986, 119, 35–38;
(b) J. Conradie and J. C. Swarts, Organometallics, 2009, 28,
1018–1026; (c) E. Fourie, J. C. Swarts, D. Lorcy and N. Bellec,
Inorg. Chem., 2010, 49, 952–959; (d) T. G. Vosloo, W. C. du
Plessis and J. C. Swarts, Inorg. Chim. Acta, 2002, 331,
188–193; (e) M. M. Conradie and J. Conradie, Inorg. Chim.
Acta, 2002, 361, 208–218.
5 (a) D. E. Graham, G. J. Lamprecht, I. M. Potgieter, A. Roodt
and J. G. Leipoldt, Transition Met. Chem., 1991, 16, 193–195;
(b) E. Erasmus, J. Conradie, A. Muller and J. C. Swarts, Inorg.
Chim. Acta, 2007, 360, 2277–2283; (c) S. Otto, A. Roodt,
J. J. C. Erasmus and J. C. Swarts, Polyhedron, 1998, 17,
2447–2453; (d) G. J. Lamprecht, J. C. Swarts, J. Conradie
and J. G. Leipoldt, Acta Crystallogr., Sect. C: Cryst. Struct.
Commun., 1993, 49, 82–84; (e) K. G. Van Aswegen,
J. G. Leipoldt, I. M. Potgieter, G. J. Lamprecht, A. Roodt
and G. J. Van Zyl, Transition Met. Chem., 1991, 16, 369–371.
6 (a) A. K. Diallo, J.-C. Daran, F. Carret, J. Ruiz and D. Astruc,
Angew. Chem., Int. Ed., 2009, 48, 3141–3145; (b) A.
Hildebrandt, D. Schaarschmidt and H. Lang, Organometal-
lics, 2011, 30, 556–563; (c) J. Conradie, T. S. Cameron,
M. A. S. Aquino, G. J. Lamprecht and J. C. Swarts, Inorg.
Chim. Acta, 2005, 358, 2530–2542.
0
pKa = 7.14(4) for 1 (R = C10F21), 9.92(3) for 3 (R = C6F5) and
10.06(2) for 4 (R = C10H21). The Gordy scale group electronega-
tivities of the fluorinated molecular fragments C10F21 (wC
3.04) and C6F5 (wC F = 2.46) and the C10H21 alkyl group (wC
=
=
10F21
10H21
6
5
2.43) were determined. Isomerisation from enol to the keto
form of the b-diketones 1–8 in CDCl3 and CD3CN were shown
by 1H NMR spectroscopy to become slower as the group
electronegativity of the R-group becomes larger. A solvent effect
is noted in that isomerisation in CDCl3 becomes in the exces-
sive case 226 times slower when changing R from C10H21 to Rc,
but it becomes only four times slower in CD3CN for the same
compounds. This difference is attributed to the stronger solva-
tion power of CD3CN over CDCl3. The electrochemistry of the
ruthenocene-containing b-diketones in CH2Cl2/0.1 mol dmꢀ3
[N(nBu4)][B(C6F5)4] revealed a one-electron oxidation Rc/Rc+
couple in the voltage range 680 o Epa o 1110 mV depending
on the R-group. The unstable 17 electron Rc+ species are more
stabilized in b-diketones when the R-group is strongly electron-
withdrawing (R = C10F21, CF3, C6F5) than when R becomes
electron-donating. When R is electron-donating as in R = CH3,
only traces of the Rc+ species at ꢀ5 1C at a scan rate of
2000 mV sꢀ1 are observed. Conversely, when R = C10F21,
reduction of the seventeen-electron species Rc+ is observed at
room temperature and a scan rate of 100 mV sꢀ1. R-groups
which are electron-donating exhibit the lowest oxidation potentials
while those compounds with electron withdrawing R-groups are
oxidized at the highest potentials. Phase studies revealed that
neither of the long chain b-diketones RcCOCH2COCR (with R =
7 F. Caruso, C. Pettinari, F. Marchetti, M. Rossi, C. Opazo,
S. Kumar, S. Balwani and B. Ghosh, Bioorg. Med. Chem.,
2009, 17, 6166–6172.
8 (a) T. J. Muller, J. Conradie and E. Erasmus, Polyhedron,
2012, 33, 257–266; (b) A. Hildebrandt, T. Ru¨ffer, E. Erasmus,
J. C. Swarts and H. Lang, Organometallics, 2010, 29,
4900–4905; (c) V. N. Nemykin, G. T. Rohde, C. D. Barrett,
C
10H21 or C10F21) exhibits any liquid-crystalline mesophase
properties. Solid state polymorphism was, however, observed.
The low temperature polymorph of the C10H21 complex
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013
New J. Chem., 2013, 37, 2862--2873 2871

View Article Online
NJC
Paper
R. G. Hadt, C. Bizzarri, P. Galloni, B. Floris, I. Nowik, 22 P. C. Reeves, Org. Synth., 1977, 56, 28–31.
R. H. Herber, A. G. Marrani, R. Zanoni and N. M. Loim, 23 J. J. C. Erasmus, G. J. Lamprecht, J. C. Swarts, A. Roodt and
J.
Am.
Chem.
Soc.,
2009,
131,
14696–14978;
A. Oskarsson, Acta Crystallogr., Sect. C: Cryst. Struct. Commun.,
1996, 52, 3000–3002.
(d) V. Chandrasekhar and R. Thirumoorthi, Organometallics,
2007, 26, 5415–5422; (e) S. Ogawa, H. Muraoka, K. Kikuta, 24 A. R. Katritzky and A. Pastor, J. Org. Chem., 2000, 65,
F. Saito and R. Sato, J. Organomet. Chem., 2007, 692, 60–69. 3679–3682.
9 W. C. du Plessis, J. J. C. Erasmus, G. J. Lamprecht, 25 A full account of determining pKas of water-insoluble
J. Conradie, T. S. Cameron, M. A. S. Aquino and
J. C. Swarts, Can. J. Chem., 1999, 77, 378–386.
10 J. C. Swarts, T. G. Vosloo, S. J. Cronje, W. C. du Plessis, C. E.
metallocene-containing b-diketones in different mix solvent
systems may be found in: W. C. du Plessis, T. G. Vosloo and
J. C. Swarts, J. Chem. Soc., Dalton Trans., 1998, 2507–2514.
J. van Rensburg, E. Kreft and J. E. van Lier, Anticancer Res., 26 R. G. Wilkins, The Study of Kinetics and Mechanism of
2008, 28, 2781–2784.
11 W. C. du Plessis, W. L. Davis, S. J. Cronje and J. C. Swarts,
Inorg. Chim. Acta, 2001, 314, 97–104.
12 (a) D. E. Bublitz, G. Hoh and T. Kuwana, Chem. Ind., 1959, 78,
635–636; (b) T. Kuwana, D. E. Bublitz and G. Hoh, J. Am. Chem.
Reactions of Transition Metal Complexes, Allyn and Bacon,
Boston, 1974, p. 42. Note that equation 1.207 in this
reference was adapted by replacing k (rate constant) with
A (absorbance), and noting that [H+] = 10ꢀpH, and KAH
=
10ꢀpK
.
a
Soc., 1960, 82, 5811–5817; (c) S. P. Gubin, L. I. Smirnova, 27 L. Helm, MINSQ, Non-Linear Parameter Estimation and Model
L. I. Denisovich and A. A. Lubovich, J. Organomet. Chem.,
1971, 30, 243–255; (d) L. I. Denisovich, N. V. Zakurin,
Development, Least Squares Parameter Optimisation V3.12,
MicroMath Scientific Software, Salt Lake City, UT, 1990.
A. A. Bazrukova and S. P. Gubin, J. Organomet. Chem., 1974, 28 Normally one would not use 2 mmol dmꢀ3 analyte
81, 207–216.
concentrations if the supporting electrolyte is present at
0.1 mol dmꢀ3; an analyte concentration of 1 or even
0.5 mmol dmꢀ3 (i.e. at least a 100 fold electrolyte excess)
would normally be more appropriate. However, the aim of
this study was to observe the reduction wave of the Rc+
fragment, which was present in very weak intensity. Upon
choosing these low analyte concentrations for 4–6, even at
2000 mV sꢀ1 scan rate, ipc disappeared in the background,
while for 1–3, the current was weak at slow scan rates. To
overcome this, we chose 2 mmol dmꢀ3 analyte concen-
tration to observe ipc for Rc+ reduction for all six studied
compounds (see Table 3) knowing that the penalty is larger
DE values. A discussion on optimum experimental electro-
chemical conditions and the use of data obtained from such
experiments may be found in ref. 15a.
13 S. Trupia, A. Nafady and W. E. Geiger, Inorg. Chem. Com-
mun., 2003, 42, 5480–5482.
14 (a) M. G. Hill, W. M. Lamanna and K. R. Mann, Inorg. Chem.,
1991, 30, 4690–4692; (b) J. C. Swarts, A. Nafady,
J. H. Roudebush, S. Trupia and W. E. Geiger, Inorg. Chem.,
2009, 48, 2156–2165.
15 (a) H. J. Gericke, N. I. Barnard, E. Erasmus, J. C. Swarts,
M. J. Cook and M. A. S. Aquino, Inorg. Chim. Acta, 2010, 363,
2222–2232; (b) R. J. LeSuer, C. Buttolph and W. E. Geiger,
Anal. Chem., 2004, 76, 6395–6401.
16 R. J. Gale and R. Job, Inorg. Chem., 1981, 20, 42–45.
17 K. C. Kemp, E. Fourie, J. Conradie and J. C. Swarts, Organo-
metallics, 2008, 27, 353–362.
18 E. Erasmus, Inorg. Chim. Acta, 2011, 378, 95–101.
19 (a) R. F. Shago, J. C. Swarts, E. Kreft and C. E. J. van Rensburg, 29 K. G. von Eschwege and J. C. Swarts, Polyhedron, 2010, 29,
Anticancer Res., 2007, 27, 3431–3433; (b) C. E. J. van Rensburg, 1727–1733.
E. Kreft, J. C. Swarts, S. R. Dalrymple, D. M. MacDonald, 30 G. Gritzner and J. Kuta, Pure Appl. Chem., 1984, 56, 461.
M. W. Cooke and M. A. S. Aquino, Anticancer Res., 2002, 22, 31 A. R. Katritzky, Y. Zhang and S. K. Singh, Synthesis, 2003,
889–892; (c) E. Fourie, E. Erasmus, J. C. Swarts, A. Jakob,
2795.
H. Lang, G. K. Joone and C. E. J. Van Rensburg, Anticancer 32 Gordy scale group electronegativities are empirical numbers
Res., 2011, 31, 825–829.
that express the accumulative tendency of a group of atoms
or a fragment of a molecule to attract electron densities of
the bonding electrons of a covalent bond as a function of
the number of valence electrons, n, and the covalent radius,
r (in Å), of groups, see (a) R. E. Kagarise, J. Am. Chem.
Soc., 1955, 77, 1377–1379; (b) P. R. Wells, Progress in
Physical Organic Chemistry, Wiley, New York, 1968, vol. 6,
pp. 111–145.
20 (a) M. Boehringer, H. Fischer, M. Henning, D. Hunziker,
J. Huwyler, B. Kuhn, B. M. Loeffler, T. Luebbers, P. Mattei,
R. Narquizian, E. Sebokova, U. Sprecher and H. P. Wessels,
Bioorg. Med. Chem., 2010, 20, 1106–1108; (b) P. Mattei,
M. Boehringer, P. Di Giorgio, H. Fischer, M. Henning,
J. Huwyler, B. Kocer, B. Kuhn, B. M. Loeffler,
A. MacDonald, R. Narquizian, E. Rauber, E. Sebokova and
U. Sprecher, Bioorg. Med. Chem., 2010, 20, 1109–1113; 33 Leading references describing the advantages of
(c) L. A. Hardegger, B. Kuhn, B. Spinnler, L. Anselm,
R. Ecabert, M. Stilhle, B. Gsell, R. Thoma, J. Diez, J. Benz,
J. M. Plancher, G. Hartmann, D. W. Banner, W. Haap and
F. Diederich, Angew. Chem., Int. Ed., 2011, 50, 314–318.
21 M. D. Rausch, E. O. Fischer and H. Grubert, J. Am. Chem.
Soc., 1960, 82, 76–82.
[N(nBu4)][B(C6F5)4] as the supporting electrolyte because it
limits ion pair formation of the type [cation]+ꢁ ꢁ ꢁ[B(C6F5)4]ꢀ
include ref. 15, as well as (a) E. Fourie, J. C. Swarts,
I. Chambrier and M. J. Cook, Dalton Trans., 2009, 1145–1154;
(b) A. Donoli, A. Bisello, R. Cardena, F. Benetollo, A. Ceccon and
S. Santi, Organometallics, 2011, 30, 1116–1121; (c) M. J. Cook,
c
2872 New J. Chem., 2013, 37, 2862--2873
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013

View Article Online
Paper
I. Chambrier, G. F. White, E. Fourie and J. C. Swarts, Dalton
NJC
Soc., 1982, 104, 5245–5247; (c) A. S. Cherodian,
A. N. Davies, R. M. Richardson, M. J. Cook,
N. B. McKeown, A. T. Thomson, J. Feijoo, G. Ungar and
K. J. Harrison, Mol. Cryst. Liq. Cryst., 1991, 196, 103–114;
(d) M. J. Cook, M. F. Daniel, K. J. Harrison, N. B. McKeown
and A. J. Thomson, J. Chem. Soc., Chem. Commun., 1987,
1086–1088.
Trans., 2009, 1136–1144; (d) F. Barriere, N. Camire, W. E. Geiger,
U. T. Mueller-Westerhoff and R. Sanders, J. Am. Chem. Soc.,
2002, 124, 7262–7263; (e) A. K. Diallo, C. Absalon, J. Ruiz and
D. Astruc, J. Am. Chem. Soc., 2011, 133, 629–641.
34 (a) A. N. Cammidge, I. Chambrier, M. J. Cook, E. H. G.
Langner, M. Rahman and J. C. Swarts, J. Porphyrins Phthalo-
cyanines, 2011, 15, 890–897; (b) C. Piechocki, J. Simon, 35 J. Han, L. F. Zhang and W. Wan, J. Organomet. Chem., 2003,
A. Skoulious, D. Guillon and J. P. Weber, J. Am. Chem.
672, 86–93.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013
New J. Chem., 2013, 37, 2862--2873 2873