[ Pobierz całość w formacie PDF ]

observed via near real-time PVM images was also in Figures 2b 2d. Evaluation of the off-line NIR S-G
qualitatively confirmed by the off-line chemical imag- first derivative spectra provided evidences of minor
ing microscopic characterization results as shown in component s presence in the final coprecipitates. For
Figures 8a 7d. The morphology difference and crys- this purpose, the S-G first derivative spectra were
tallinity difference, to a large extent, will dictate the grouped into two groups: (1) GroupEud with weight
selections of formulation strategy and downstream fraction of Eudragit L100 larger than 0.9 (WEud > 0.9);
processing.41 42 (2) Groupnap with weight fraction of naproxen larger
than 0.3(Wnap > 0.3). Then signatures of naproxen
in GroupEud and Eudragit L 100 in Groupnap were
NIR Characterization of the Final Coprecipitates
sought.
Figures 2a 2d summarize the NIR characterization On the basis of Beer s law,45 when the sample path-
results of the final coprecipitates obtained in the length can be approximated as a constant, the spec-
exploratory study and pure naproxen and Eudragit tral value is proportional to the species concentration.
L100 as well. Simple qualitative comparison of the Supposing there is no significant interaction between
NIR raw spectra reveals the following findings: two components Z1 and Z2, the spectrum of a prod-
uct Y consisting of components Z1 and Z2 will be a
(1) The NIR spectra of the coprecipitation products linear superposition46 47 of the spectra of Z1 and Z2.
from batches with low drug/polymer ratios (I Consequentially, the spectral value of Y will be in be-
and J) are similar to that of the pure component tween the spectral values of Z1 and Z2 and will not go
Eudragit L100. beyond the spectra values of Z1 and Z2. If there is a
(2) Similarly, the NIR spectra of the coprecipitation closed spectral area bordered by the spectra of Z1 and
products from batches with high drug/polymer Z2, then the spectrum of Y should lie in this closed
ratios (E and D) are similar to that of naproxen. spectral area and will not go beyond it.
(3) A unique and strong characteristic NIR ab- For GroupEud, there were two identified wavelength
sorbance peak around 1910 nm for Eudragit ranges, (1100, 1140) and (1680, 1720) nm, which can
L100 was presented at the NIR spectra of provide evidences of naproxen as minor component
the coprecipitation products from batches with presented in the final products. Around 1127 nm there
low drug/polymer ratios (I and J), but not at was a downward characteristic peak of naproxen
the NIR spectra of the coprecipitation products (Fig. 2b). Eudragit L100 has spectral values lower
from batches with high drug/polymer ratios (D than that of naproxen. A closed area bordered by
and E). the spectra of naproxen and Eudragit L100 can be
(4) There are a few weak characteristic NIR peaks found in Figure 2b. For most of the final coprecipita-
for naproxen, located around 1142, 1184, 1378, tion products from formulations dominant with Eu-
2044, 2164, 2386, 2418, 2466 nm. These weak dragit l100, their S-G first derivative NIR spectra
characteristic peaks were presented at the NIR fall within this closed area. Similar situations were
spectra of the coprecipitation products from observed for wavelength range of (1680, 1720) nm
batches D and E (with high drug/polymer ra- (Fig. 2c), where two closed areas, (1682.5, 1689.5) and
tios) but not for batches I and J (with low drug/ (1689.5, 1715) nm, were bordered by the spectra of
polymer ratios). naproxen and Eudragit L100. For all of the final copre-
cipitation products from formulations dominant with
The above findings are in qualitative agreement Eudragit l100, their S-G first derivative spectra fall
with the PVM observations, which show that amor- within these two closed areas. Therefore, the spectral
phous nature was dominant for batches I and J, evidences of naproxen in wavelength ranges of (1100,
whereas crystalline nature was dominant for batches 1140) and (1680, 1720) nm confirmed the existence of
D and E. However, NIR detection limit may present naproxen in the final products in GroupEud, in spite
challenge regarding detecting the presence of minor of its low percentage. Similarly, the presence of Eu-
component due to its low percentage in those starting dragit L100 in the all of the final products in Groupnap
formulations. Spectral preprocessing algorithms may was spectrally confirmed as shown in Figure 2d.
DOI 10.1002/jps JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 100, NO. 5, MAY 2011 [ Pobierz całość w formacie PDF ]