M. Yang et al.
Molecular Catalysis 498 (2020) 111247
Cp2TiCl2 may increase the catalytic power of a Lewis acid metal centre
while maintaining relative stability, we took the synthesis of benzo[b]-
[1,4]diazepines as a touchstone. Benzodiazepines are a kind of nitrogen
heterocyclic compound with high pharmacological activity [62–67].
Many acid catalysts have been developed for the synthesis of such
compounds, involving Brønsted acid or Lewis acid catalysts, such as Zn
[(L)Proline]2 [68], Ga(OTf)3 [69], CeCl3 [70], SbCl3‒Al2O3 [71], Au(I)
[72], Yb(OTf)3 [73], and so on. In our previous work, we found that
amide ligand urea facilitates the removal of both Cp rings from titano-
cene [74]. Herein, we disclose the establishment of half-titanocene
complexes with just one Cp ligand, which show excellent catalytic ac-
tivity, water-insensibility and sustainability for the synthesis of benzo
[b]-[1,4]diazepines (Scheme 1). The novel prepared half-titanocene
complex is featured by the removal of one Cp ligand from the titano-
cene dichloride under the dual activation on of a N-donor ligand and
alcohol solvent, the structure of which is elucidated by 1H NMR and ES
(+)-MS analysis. Cyclic experiments demonstrate the sustainability of
the titanocene complex and the employment of an alcohol-water mixed
solvent proves its tolerance towards water.
because only 4%, 21% and 63% product could be obtained, respectively,
for their absence.
To shed light on how the COOH ligand and alcoholic solvent activate
Cp2TiCl2, a titration experiment was conducted by adding activator
methanol and acid ligand to the deuterated chloroform solution of
Cp2TiCl2, followed by the detection of the Cp signal in 1H NMR spectra,
which can be interpreted as the existence of titanocene species. As
shown in Fig. 1, acid ligand (line 2) or methanol (line 3) alone does not
cause any transformation of Cp2TiCl2, as verified by the unchanged
signal for the Cp peak at δ 6.59 ppm (●). Combinative addition of ligand
and methanol was also unable to activate Cp2TiCl2 (line 4). After adding
o-phenylenediamine, two new signals immediately appeared at δ 6.37
ppm (▾) and δ 6.27 ppm (◆), accompanied by the disappearance of the
signal at δ 6.59 ppm (in less than 1 min), suggesting that Cp2TiCl2
transformed into two new titanocene species. The chemical conversion
of Cp2TiCl2 might be attributed to the induced effect of o-phenylenedi-
amine, which was not only added as the reaction substrate, but also
played as a crucial role in the removal of one Cp ligand. As the reaction
time increased, the signal at δ 6.27 ppm gradually decreased and dis-
appeared while the signal at δ 6.37 ppm increased, which indicated that
the real catalytic species has experienced an intermediate state during
the formation process.
2. Results and discussion
To further determine the structure of the real catalytic species, we
analysed 1H NMR spectra and ran an ESI-MS experiment. As can been
seen from Fig. 2b, the proton number ratio for the benzene ring and Cp
ring is 4:5, indicating that the ligand ratio for the acid and Cp is 1:1 in
the titanocene catalytic species. The structure was further elucidated by
ESI-MS analysis. As shown in Fig. 2c, the major signal with an m/z value
of 175.0238 should correspond to Int-II′, where two OMe ligands and
one Cp ligand are bound to the titanium metal centre. The other peak
with an m/z value of 309.0246 (Fig. 2d) should correspond to the ion
peak Int-II′′, in which one Cp ligand, one OMe ligand and one acid ligand
are coordinated to the titanium centre. Taking all the 1H NMR spectra
and ESI-MS spectra together, the catalytic species should be CpTi
Our investigation started with the evaluation of the ligand effect on
the catalyst Cp2TiCl2 through synthesis towards benzo[b]-[1,4]
diazepines from 1,3-ynone and o-phenylenediamine. Initially, we ran
ligand-free conditions, and the target product was only generated with a
low yield of 7% (Table 1, entry 1). Subsequently, a series of COOH li-
gands were examined (entries 2-7). As seen from the experimental re-
sults, the addition of COOH ligands improved the reaction efficiency,
and m-phthalic acid proved to be the best choice, assuring a 40% yield
for the products’ generation. Next, the solvent effect was investigated
systematically, with the results shown in entries 8-17. The worst situa-
tion was when the aprotic solvent toluene was adopted as a solvent
because no product was detected (entry 8). Employment of regular polar
solvents, such as DMF, DMSO, and CH3CN (entries 9-11), assured that
the product yield varied from 29% to 41%, commensurate to that of the
case for 1,4-dioxane (entry 7). Generally, when alcoholic solvents were
used as solvents, the yields were at a better level (entries 12-17), espe-
cially for monohydric alcohols (entries 14-17). In particular, when
methanol and ethanol were employed, 96% yield products were affor-
ded. From a green chemistry point of view, ethanol was chosen as the
optimal solvent (entry 16). To confirm the ligand and solvent activation
effect on the title reaction, attempts deviating from standard conditions
were also performed. As shown in entries 18-20, Cp2TiCl2, ligand and
ethanol are all indispensable for the smooth progression of the reaction
(OMe)2(
η
1-C8H5O4), which is transformed from Cp2TiCl2 under dual
activation of methanol and m-phthalic acid. To ensure that the titano-
cene species obtained by the 1H NMR titration experiment was the real
catalyst, the premade catalyst was used to catalyse the model reaction in
Table 1, a commensurate yield of 97 % was obtained, verifying our
proposal.
The complex of CpTi(OMe)2(
η
1-C8H5O4) we disclosed were synthe-
sized from Cp2TiCl2. This method for the synthesis of the half-titanocene
complex was in low cost, because raw material Cp2TiCl2 was more stable
and cheaper than most half-titanocene complexes. For comparison, the
synthesis methods of reported half-titanocene complexes containing O-
Scheme 1. Titanocene complexes accelerated by a ligand strategy.
2
Yang, Mingming
