Chemistry Reference
In-Depth Information
16.4.2.3.2 2009 Version from Koide's Group. Removal of an allyl group
from allylic ethers has routinely been performed using Pd(PPh
3
)
4
in
organic solvents. In aqueous media, the deallylation might be more
ecient with different phosphine ligands. Therefore, to optimize further
the 2007 method, Koide's group screened commercial phosphine ligands
in the presence of NaBH
4
in 1:9 DMSO-pH 7 buffer and found that John-
Phos and tri(2-furyl)phosphine (TFP) were optimal.
65
Because TFP is more
economical and more soluble in water, this ligand was chosen for further
studies. This 2009 version also showed linearity between the palladium
concentration and fluorescence signals.
They evaluated this 2009 version to determine how quantitative it can be
in the presence of drugs or drug-like compounds. Drug or drug-like samples
were prepared to contain 8.5 ppm of palladium (8.5 mg of palladium per
gram of drug). The expected amount of palladium could be observed in
17 out of 18 samples in either a pH 7 or a pH 10 buffer.
The same group also measured trace palladium after 41 was prepared
by a Suzuki-Miyaura coupling (Figure 16.17). The APE method showed
that the palladium content was 1884
169 ppm. Analyses by ICP-MS and
ICP-OES methods, conducted by two independent institutions, indicated
that the palladium content in the indole derivative was 1813
316 and
2097
108 ppm, respectively.
16.4.2.3.3 2010 Version from Koide's Group. At this point, Koide's
group had several inquiries about the APE method from pharmaceutical
companies. Although their actual data could not be made available to us, it
became clear that the APE method did not meet their expectations
regarding robustness. Simply put, the success rate was only modest (
50%
at best), not meeting criterion 6 in section 16.3. Users might have taken too
much of each sample for analysis; conventionally, the more sample one
uses, the more accurate the analysis is. However, this concept does not
apply to the APE method; as Figure 16.18(a) shows, when trace palladium
in amines was quantified, the deviation from the standard curve became
increasingly noticeable for amine concentrations 42mgmL
1
.
66
This
could be explained by eqn (16.1); reducing the concentration of an API
should allow the equilibrium to shift toward the left. Therefore, the APE
B
Figure 16.17
I ndole derivative 41 was prepared by Suzuki-Miyaura coupling and
subjected to palladium analysis.
Reproduced with permission from Ref. 65. Copyright 2009 American
Chemical Society.
Search WWH ::
Custom Search