This art icle was downloaded by: [ Leonardo Meneghini]
On: 07 Novem ber 2011, At : 08: 50
Publisher: Taylor & Francis
I nform a Lt d Regist ered in England and Wales Regist ered Num ber: 1072954 Regist ered
office: Mort im er House, 37- 41 Mort im er St reet , London W1T 3JH, UK
Journal of Liquid Chromatography &
Related Technologies
Publicat ion det ails, including inst ruct ions f or aut hors and
subscript ion inf ormat ion:
ht t p: / / www. t andf online. com/ loi/ lj lc20
CHEMOMETRIC EVALUATION OF
DARIFENACIN HYDROBROMIDE USING A
STABILITY-INDICATING REVERSED-PHASE
LC METHOD
L. Z. Meneghini
C. F. Codevilla
a
a
, C. Junqueira
, P. E. Fröehlich
a
a
, A. S. Andrade
b
& A. M. Bergold
, F. R. Salazar
a
,
a
a
Programa de Pós-Graduação em Ciências Farmacêut icas, Faculdade
de Farmácia, Universidade Federal do Rio Grande do Sul-UFRGS,
Port o Alegre, Brasil
b
Laborat ório de Pest icidas e Resíduos Vet erinários-Minist ério da
Agricult ura do Brazil, Port o Alegre, Brazil
Available online: 03 Nov 2011
To cite this article: L. Z. Meneghini, C. Junqueira, A. S. Andrade, F. R. Salazar, C. F. Codevilla, P. E.
Fröehlich & A. M. Bergold (2011): CHEMOMETRIC EVALUATION OF DARIFENACIN HYDROBROMIDE USING
A STABILITY-INDICATING REVERSED-PHASE LC METHOD, Journal of Liquid Chromat ography & Relat ed
Technologies, 34: 18, 2169-2184
To link to this article: ht t p: / / dx. doi. org/ 10. 1080/ 10826076. 2011. 585486
PLEASE SCROLL DOWN FOR ARTI CLE
Full t erm s and condit ions of use: ht t p: / / www.t andfonline.com / page/ t erm s- and- condit ions
This art icle m ay be used for research, t eaching, and privat e st udy purposes. Any
subst ant ial or syst em at ic reproduct ion, redist ribut ion, reselling, loan, sub- licensing,
syst em at ic supply, or dist ribut ion in any form t o anyone is expressly forbidden.
The publisher does not give any warrant y express or im plied or m ake any represent at ion
t hat t he cont ent s will be com plet e or accurat e or up t o dat e. The accuracy of any
inst ruct ions, form ulae, and drug doses should be independent ly verified wit h prim ary
sources. The publisher shall not be liable for any loss, act ions, claim s, proceedings,
dem and, or cost s or dam ages what soever or howsoever caused arising direct ly or
indirect ly in connect ion wit h or arising out of t he use of t his m at erial.
Journal of Liquid Chromatography & Related Technologies, 34:2169–2184, 2011
Copyright # Taylor & Francis Group, LLC
ISSN: 1082-6076 print/1520-572X online
DOI: 10.1080/10826076.2011.585486
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
CHEMOMETRIC EVALUATION OF DARIFENACIN HYDROBROMIDE
USING A STABILITY-INDICATING REVERSED-PHASE LC METHOD
L. Z. Meneghini,1 C. Junqueira,1 A. S. Andrade,2 F. R. Salazar,1
C. F. Codevilla,1 P. E. Fröehlich,1 and A. M. Bergold1
1
Programa de Pós-Graduação em Ciências Farmacêuticas, Faculdade de Farmácia,
Universidade Federal do Rio Grande do Sul-UFRGS, Porto Alegre, Brasil
2
Laboratório de Pesticidas e Resı´duos Veterinários-Ministe´rio da Agricultura
do Brazil, Porto Alegre, Brazil
& A set of chemometric tools including in silico simulation and design of experiments (DOE) was
used to develop and validate an RP-LC=UV method for darifenacin hydrobromide. Retention factor
(k0 ) from darifenacin and resolution (R) between darifenacin and degradation product were assessed
by an exploratory study using organic modifier type, organic modifier content, mobile phase pH, flow
rate, and column oven temperature as factors in a full factorial 25 design. The significant factors
(organic modifier content, mobile phase pH, and column oven temperature) were selected and
applied in a central composite design (CCD) with axial points positioned in a ¼ 1.681. The optimum condition was obtained with Derringer’s desirability. The validation was performed with a
C8 (150 mm 4.6 mm, 5 lm) column, wavelength 205 nm, 1.1 mL min1 flow rate, mobile phase
with acetonitrile (MeCN) and orthophosphoric acid (OPA) 0.01% (v=v) pH 3.0 in a ratio 28:72
and column maintained at room temperature (25 1 C). Forced decomposition was performed in
agreement with the guidelines and collected data were confirmed with LC-MS=MS. The obtained
k0 and R were 3.22 and 2.62, respectively. Additionally a Plackett-Burman design was done to study
robustness, and the results agreed with the guideline requirements for a stability-indicating method.
Keywords darifenacin hydrobromide, Derringer’s desirability, design of experiments,
high performance liquid chromatography, stability-indicating
INTRODUCTION
Development and optimization may be excessively time-consuming in
analytical separations, when the technique uses a broad range of factors
for this purpose, and more than one response must be optimized concurrently.[1] Reversed-phase liquid chromatography (RP-LC) is a traditional
Address correspondence to Ana Maria Bergold, Programa de Pós-Graduação em Ciências
Farmacêuticas, Faculdade de Farmácia, Universidade Federal do Rio Grande do Sul-UFRGS. Av.
Ipiranga, 2752, CEP 906610-000, Porto Alegre, RS, Brazil. E-mail: ana.bergold@ufrgs.br
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
2170
L. Z. Meneghini et al.
technique for pharmaceutical dosage. It utilizes many experimental
conditions (concentration of organic modifier, pH, flow rate, and others)
associated with more than one response (retention time, capacity factor,
resolution, among others). Since a one-factor-at-a-time optimization could
be a complicated process, a chemometric approach involving optimization
and validation steps is recommended for a quick, economic, precise, and
accurate LC method.[2–4]
This paper aimed at developing and validating a simple stabilityindicating LC=UV method, employing the design of experiments (DOE)
methodology to determine darifenacin hydrobromide (2,3-dihydrobenzofuran-5-yl)ethyl)-3-pyrrolidinylg-2,2-diphenylacetamide hydrobromide) in
an extended-release tablet formulation with 15 mg of the drug. Darifenacin
hydrobromide (DF) (Figure 1) is an anticholinergic drug used in patients
with an overactive bladder who do not achieve symptom relief and quality
of life improvement with conservative management.[5] In the literature,
LC=APCI-MS=MS was used for the determination of DF enantiomers in
human plasma.[6] In addition, other chiral chromatography methods have
been used for the determination of DF in bulk drug and extended release
tablets.[7,8] However, they employed techniques that require a special column and complex mobile phase unusual in routine analysis. Furthermore,
no degradation products resulting from forced-degradation in the
validation study have been reported.
DOE provides a variety of designs and chemometric tools for the LC
application. In the development stage, two-level factorial design is useful
for screening, but it does not provide a detailed model of the system
investigated.[1,9–11]
The central composite design (CCD) is often used to find the optimum
condition and to provide a mathematical model to predict how one or more
responses relate to the values of various factors.[9,12,13] When the situation
involves simultaneous consideration of multiple responses, a multi-criteria
decision making (MCDM) chemometric tool is necessary. There are many
deals for MCDM, including the desirability functions (Derringer’s desirability). After an appropriate building response surface model for each
response, this technique converts the latter into an individual desirability
FIGURE 1 Darifenacin hydrobromide (molecular weight 507.5).
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
Chemometric Evaluation of Darifenacin Hydrobromide
2171
function (di) to achieve a global optimization.[13–15] Once the method has
been developed it must be validated. In official compendia it is encouraged
that some requirements, like robustness, be performed with DOE. Many
types of factorial designs using small variations in the method parameters
can be performed to observe method reliability in routine analysis.[16–18]
In the present work, initial chromatographic conditions in the RP-LC
were investigated by in silico simulation. An exploratory two-level factorial
design was used to assess some factors that could affect the DF retention
factor and the resolution. Next, a CCD design was applied for a response
surface methodology (RSM) study. Derringer’s desirability has been used to
define the global optimum conditions in RSM models and to predict the
best chromatographic system. In addition, a Plackett-Burman (PB) design
was performed to test robustness in the validation stage.
EXPERIMENTAL
Materials and Reagents
Darifenacin hydrobromide (99.70%) was purchased from Waterstone
Technology (USA). Methanol, acetonitrile, and tetrahydrofuran LC grade
are from J.T Baker (USA). Orthophosphoric acid, formic acid, and acetic
acid were obtained from Merck (Germany). The 0.45-mm nylon filters were
purchased from Millipore (USA). Analysis was performed on a Shimadzu
chromatographic system with LC20A binary pump, SPD-AVvp UV=VIS
photodiode array detector, SIL 20-AC autosampler, and CTO20A oven. Alternatively, for the robustness study, a Shimadzu system with LC20A quaternary
pump, SPD-AV UV=VIS photodiode array detector, and Rheodyne 7725i
manual injector was used. Samples from forced degradation study were analyzed by LC Agilent 1100 series (quaternary pump) coupled MS=MS API
5000 (LINAC) in electrospray positive mode. The chromatographic separations were achieved with Shimadzu C8 column (150 mm 4.6 mm, 5 mm) for
the experimental designs and validation; Merck C8(250 mm 4.6 mm
5 mm), Merck C8 (125 mm 4.6 mm 5 mm) columns in the robustness
study; Macherey-Nagel C18 (150 mm 4.6 mm 5 mm), Macherey-Nagel
C18 (250 mm 4.6 mm 5 mm), and Merck C8 (250 mm 4.6 mm 5 mm)
columns for the in silico simulations.
Softwares
Osiris 4.2.0.0 version (DATALYS Inc., France) was used for the in silico
chromatographic simulations. The experimental designs were analyzed by
Design-Expert 7.0.0 version (Stat–Ease Inc., USA) and the other statistical
tests (linearity, ANOVA) by STATISTICA 5.0 version (Stat-Soft Inc., USA).
2172
L. Z. Meneghini et al.
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
Chemometric Methods
Initial chromatographic conditions were assessed by simulation using
Osiris. This software provides many interfaces (tabs) and it was necessary
to perform some runs under specific conditions (inputs) to obtain predictions (outputs) from modifications in the chromatographic parameters.
The mobile phase (MP) – pH tab was used with different stationary phases
(RP-C18, RP-C8); eluent composition (acetonitrile-water, methanol-water)
and pH range (2.5-3.5) were the inputs and the DF retention time was the
output. After this trial, two chromatographic systems were performed with
samples from the stress testing. In order to improve the chromatographic
parameters retention factor (k0 ) and resolution (R), an exploratory 25
general factorial design (Table 1) was applied. In factorial design 2k, k is
the number of factors and 2 is the number of levels for each factor. It was
therefore possible to select significant factors and to apply a rotatable
CCD design with eight factorial points, six axial points (positioned in
a ¼ 1.681), and six central points. Once the design has been performed,
the first step in RSM is to find a suitable approximation for the true
functional relationship between response and the set of independent
variables.[9] For an experimental design with three factors, a low-order
polynomial model is usually employed:
y ¼ b0 þ b1 x1 þ b2 x2 þ b3 x3 þ b12 x1 x2 þ b13 x1 x3 þ b23 x2 x3
ð1:0Þ
In this first order response surface model, y represents the estimated
response (e.g., resolution); b0, b1, b2, bm is a set of unknown parameters
where b0 is the average experimental response, b1 to b3 are the estimated
effects (main effects) on the factors (x1,x2,x3), and b12 to b23 are estimated
effects with interaction. After model assessment a higher order (quadratic
or cubic) polynomial model and=or mathematical transformation in the
response is sometimes necessary.[19]
TABLE 1 General Full-Factorial Design
Levels
Factors
Units
(1)
(þ1)
A: Organic Modifier Type
Category acetonitrile tetrahydrofuran
B: Organic Modifier Content
% (v=v)
20
40
C: Mobile Phase pH
–
2.5
5.0
0.90
1.20
D: Flow Rate
mL min1
E: Column Oven Temperature
C
25.0
50.0
(1) and (þ1) are ‘‘low’’ and ‘‘high’’ levels.
Chemometric Evaluation of Darifenacin Hydrobromide
2173
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
Stress Studies
Forced decomposition was performed in agreement with the guidelines
and specific literature[16,20] in methanolic solutions containing drug
substance or drug product. Intentional degradation was carried out by
exposing 5 mL of reference=test stock solution to 5 mL of hydrochloric acid
or sodium hydroxide at different concentrations. This procedure was
repeated using additional 50 C heat in a water bath. The solutions were
transferred to a 10-mL volumetric flask, allowed to attain room temperature, and then neutralized with acid or base (when necessary). Oxidative
degradation was performed by mixing equal volumes of standard stock solution with 13% hydrogen peroxide solution in a 10-mL volumetric flask.
The samples were allowed to reach ambient temperature and diluted with
methanol. The solutions were kept in a dry oven at 50 C and 80 C for different lengths of time to perform a thermal stress study. Photolytic studies
were carried out with a 2 mL sample in a quartz container and exposed to
light (UV 352 nm) in a photostability chamber (controlled temperature).
Blank solutions were prepared by the aforementioned procedure, wherein
instead stock solutions methanol was used. The analytical data of the
method were collected by PDA detector for the peak purity evaluation.
When secondary peak was detected, an ESI LC=MS=MS analysis was
performed for confirmation.
Validation
Validation was performed according to the ICH guidelines Q2(R1)[16]
and US Pharmacopeia.[21]
System Suitability and Solution Stability
The stability of DF in solution was assessed at room temperature, under
ultrasound effects, and refrigeration condition at different periods. Prior to
beginning the method validation stage, the accuracy and precision of LC
data collected were checked. For this, ten injections were performed and
the chromatographic parameters were evaluated by RSD.
Linearity and Range
Solutions of DF were prepared at seven concentrations (0.005, 0.010,
0.015, 0.020, 0.025, 0.030, 0.035 mg mL1) spanning the range of 25–
175% of the target DF assay concentration (0.020 mg mL1) for the linearity
experiments. Ordinary least-squares method was applied to calibration
curve construction y ¼ ax b, where y is the response (peak area), x is the
concentration, and a and b are the slope and intercept, respectively. The
2174
L. Z. Meneghini et al.
model was evaluated by determination coefficient regression significance,
lack-of-fit, and residual analysis.
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
Precision
Fifty tablets were weighed and finely powdered. A portion of powder was
accurately weighed, placed for 15 min in an ultrasound bath and diluted to
0.020 mg mL1 in mobile phase. Then, the sample solutions were filtered
through a 0.45-mm membrane and transferred to vials for LC quantification.
Nine experiments were carried out in triplicate on three different days with
different analysts and instruments. The RSD was assessed for repeatability
and intermediate precision.
Accuracy
Method accuracy was calculated as a percentage of recovery by assaying
the known added amount of analyte in the samples. The concentrations
were 75%, 100%, and 125% from nominal concentration. The recoveries
RSD were evaluated.
Limit of Detection (LOD) and Limit of Quantification (LOQ)
LOQ was determined by two approaches: based on signal-to-noise ratio
(RS=N), with 10:1 ratio and based on standard deviation of response and
slope (SDLOQ). This limit was subsequently validated by the assay of a set
of samples. Likewise, LOD was assessed on RS=N (generally 3:1) and SDLOD;
additionally the LOD was justified with an experimental test. The equations
for LOQ and LOD involving standard deviation of the slope and intercept
have been estimated by the following equations:
SDLOQ ¼ 10:0 s=S
TABLE 2
SDLOD ¼ 3:3 s=S
ð2:0Þ
Factors and Levels Investigated in the Placket-Burman Design
Levels
Factors
A: MeCN Content(%) in MP
B: pH Aquose Component in MP
C: Flow Rate
D: Wavelength
E: Autosampler Injection Volume
F: Column Oven Temperature
G: Acid Type
H: Column Type
I–L: Dummies 1.2. 3
Units
(1)
(þ1)
(0)
% (v=v)
–
mL=min
nm
mL
C
Categoric
Categoric
27
2.90
1.0
206
10
23
Acetic
Merck,
120 mm
–
29
3.10
1.2
210
30
27
Formic
Merck,
250 mm
–
28
3.00
1.1
208
20
25
OPA
Shimadzu,
150 mm
–
–
(1), (þ1), (0) are ‘‘low’’, ‘‘high’’ and nominal levels, respectively.
Chemometric Evaluation of Darifenacin Hydrobromide
2175
where s is the standard deviation of the response (blank) and S is the slope
of calibration curve.
Robustness
Eleven factors were selected to construct the centered Plackett-Burman
design (Table 2), but three factors, named dummy factors represent unreal
physical changes in the chromatographic system. Nominal levels are the
conditions used for method validation. Pareto chart (with t statistics) and
half-normal probability plot were used to indicate relevant effects.
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
RESULTS AND DISCUSSION
Forced Decomposition Study
The results are shown in Table 3. A stress condition with hydrogen
peroxide 13% for 72 hr resulted in significant degradation of DF drug
substance and drug product. The data were collected from a PDA detector
and confirmed with LC-MS=MS, (Figure 2). Mass balance between
non-degraded DF solution and degraded samples agreed (information
not shown).
DOE
In silico chromatographic simulation demonstrated that C8 (250 mm
4.6 mm 5 mm, Merck) was the best choice for the column. An isocratic
elution employing a mixture of acetonitrile (MeCN), methanol (MeOH)
and orthophosphoric acid (OPA) (pH 2.5, 0.01%v=v) at a ratio 20:40:40
resulted in adequate parameters for the drug product assay. This HLPC
TABLE 3 Conditions Applied to the Forced Degradation Study
Assay
Stress Agent
HCl 0.1 M
HCl 1 M
NaOH 1 M
NaOH 1 M
Heat 50 C
Heat 80 C
Heat 50 C þ HCl 0.1 M
Heat 50 C þ HCl 0.1 M
H2O2 1.3%
H2O2 13%
UV 352 nm
UV 352 nm
Time (hours)
Drug Substance
Drug Product
24
72
24
72
24
72
14
14
72
72
24
72
99.98
99.75
99.34
100.26
99.70
100.24
99.57
100.12
99.62
92.54
100.38
100.24
99.64
99.11
99.13
99.34
99.10
99.32
99.83
98.83
99.39
94.23
99.32
100.22
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
2176
L. Z. Meneghini et al.
FIGURE 2 Mass spectrum comparison between DF (a) and peak from stress degradation (b); PDA
scanning from DF peak (l) and product from stress degradation (2). (Color figure available online.)
system was performed with samples from stress studies. Peak impurity
was detected in the 13% hydrogen peroxide solution of the samples. New
MeCN: OPA 0.01% ratios were attempted but not enough peak purity
was achieved. As selectivity is a function of type and concentration of
organic modifier,[11,22] the full-factorial 25 was used to assess the differences
between MeCN and tetrahydrofuran (THF) and other conditions. Significant effects with the k0 response were found for all factors analyzed
(Figure 3a). Flow rate did not present a significant mean effect or interaction effect for R (Figure 3b). No significant R difference was found between
THF and MeCN; than the last one was chosen to be investigated in the CCD
(less expensive and damaging for the apparatus). The results of CCD
performed in twenty runs are shown in Table 4.
Statistical analysis demonstrated that a second order polynomial was
most appropriate for the relationship between factors and responses, as
the response surface graphs show in the Figures 4 and 5. Based on the k0
results (Figure 4) it can be seen that the k0 value increases in the region
with lower pH, higher MeCN content and higher column oven temperature. For R (Figure 5), the lowest values of pH, MeCN content and column
oven temperature results in the highest resolution value. The polynomials
2177
Chemometric Evaluation of Darifenacin Hydrobromide
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
FIGURE 3 Perturbation plots from full-factorial 25 showing the significant effect of the examined
factors on the responses (a) Retention factor, (b) Resolution. Where the organic modifier type is
(1) for the MeCN and (þ1) for the THF, B is an organic modifier content, C is mobile phase pH,
D is a flow rate and E is a column oven temperature.
were obtained after ‘‘backward elimination’’ of the nonsignificant
coefficients.
Derringer’s desirability was used to optimize the k0 and R. The first step in
this procedure was to convert each response into an individual desirability
function (di) which varies over the range 0 di 1, where di ¼ 1 is the best
TABLE 4
Experimental Design for DCC
Factors and Levelsa
Runs
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
a
MeCNb
22.0
28.0
22.0
28.0
22.0
28.0
22.0
28.0
20.0
30.0
25.0
25.0
25.0
25.0
25.0
25.0
25.0
25.0
25.0
25.0
(1)
(þ1)
(1)
(þ1)
(1)
(þ1)
(1)
(þ1)
(-1.681)
(þ1.681)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
pHc
3.0
3.0
5.0
5.0
3.0
3.0
5.0
5.0
4.0
4.0
2.3
5.7
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
(1)
(1)
(þ1)
(þ1)
(1)
(1)
(þ1)
(þ1)
(0)
(0)
(1.681)
(þ1.681)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
(0)
Responses
Oven Temperatured
26.0
26.0
26.0
26.0
45.0
45.0
45.0
45.0
35.5
35.5
35.5
35.5
19.5
51.5
35.5
35.5
35.5
35.5
35.5
35.5
(1)
(1)
(1)
(1)
(þ1)
(þ1)
(þ1)
(þ1)
(0)
(0)
(0)
(0)
(1.681)
(þ1.681)
(0)
(0)
(0)
(0)
(0)
(0)
(1), (þ1), (0) are ‘‘low’’, ‘‘high’’ and nominal levels, respectively.
Acetonitrile content (%) in mobile phase.
c
pH of the mobile phase.
d
Column oven temperature in degree celsius.
b
k0
R
11.40
4.41
19.73
6.01
7.34
3.12
13.93
4.32
16.83
3.23
5.30
13.70
8.39
4.31
6.45
6.08
6.46
6.92
7.21
6.87
3.68
2.51
1.90
2.53
3.25
2.15
0.73
1.68
2.08
1.88
3.07
1.57
3.03
2.17
2.60
2.37
2.42
2.62
2.60
2.55
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
2178
L. Z. Meneghini et al.
FIGURE 4 RSM graphs showing the effect of factors on retention factor (k0 ): (a) MeCN content (A)
and pH (B) in mobile phase; (b) MeCN content (A) and column oven temperature (C); (c) mobile
phase pH (B) and column oven temperature (C). (Color figure available online.)
FIGURE 5 RSM graphs showing the factor effects on resolution (R): (a) MeCN content (A) and mobile
phase pH (B); (b) mobile phase pH (B) and column oven temperature (C). (Color figure available
online.)
Chemometric Evaluation of Darifenacin Hydrobromide
2179
alternative and di ¼ 0 when the response is outside the optimal region.
Then the design variables are chosen to maximize global desirability (D),
D ¼ ðd1 d2 . . . dm Þ1=m
ð3:0Þ
where there are m responses and d1,d2 . . . dm are the individual desirabilities. If the objective or target T for the response y is a maximum value:
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
d¼
8
< 0;
yL
T L
:
1;
r
y<L
LyT
y>T
ð3:1Þ
when weight r ¼ 1, the desirability function is linear. Choosing r > 1 places
more emphasis on being close to the target value, and choosing 0 < r < 1
makes this less important. If the target for the response is a minimum value:
d¼
8
< 1;
U y
U T
:
0;
r
y<T
T yU
y>U
ð3:2Þ
the two-sided desirability function (3.3) assumes that the target is located
between the lower (L) and the upper (U) limits, and is defined as:
8
0;
>
r 1
>
>
>
< yL
T L
d¼
r 2
>
U y
>
>
>
: U T
0;
y<L
LyT
ð3:3Þ
LyU
y>U
FIGURE 6 Desirability’s Derringer RSM graph showing optimum operability region of the chromatographic method where (A) is MeCN content and (B) is pH in mobile phase. (Color figure available
online.)
2180
L. Z. Meneghini et al.
TABLE 5 Parameters Used in Derringer’s Desirability
Criteria
Response=
Parameter
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
k0
R
Column oven temperature
Lower
Limit
Upper
Limit
Goal
Importance
2
2
23
5
4
28
Target ¼ 3.5
Range
Target ¼ 25 C
2
2
1
The results of Derringer’s function are demonstrated in Figure 6. The
conditions proposed for criteria selection have been established in the
appropriate parameters for routine analysis (Table 5). Room temperature
was chosen to be used for method simplification. Ten solutions have been
obtained from the response surface producing a desirability value (D) of
TABLE 6
Requirements Assessed for the RP-HPLC=UV Method Validation
System Suitability Test (n ¼ 10)
Area
RSDa
Parameter
R2
b ¼ 743
a ¼ 70109
Rt b
1397478
8.45
0.03
0.47
Linearity
CLg
Seh
–
12877
13152
6283
588
281
k0
Tc
Nd
3.22
0.12
1.25
0.35
p-value
–
0.91
<0.01
RS=N
79.69
Re
4712
2.62
0.22
0.46
LQ analysisf
SDLOD
Experimental
11.78
400.00
LD analysisf
3.89
20.00
29.30
Model Assessment
Parameter
Statistical Test
Regression
Lack-of-Fit
Autocorrelation in Residuals
Homoscedasticity
ANOVA; F-value
ANOVA; p-value
Durbin-Watson; D-value
Levene; p-value
Barllell; p-value
Ryan-Joiner; p-value
Normality Test
Kolmogorov–Smirnoff; p-value
Anderson-Darling; p-value
Dixon; s-value
Outlier Assumption
a
Relative standard deviation.
Retention time.
c
Tail.
d
Theorical plates.
e
Resolution.
f
Concentration in ng mL1.
g
Confidence limit.
h
Standard error.
b
Critical Value
Calculated Value
3.87
0.05
dL ¼ 1.22; dv ¼ 1.42
0.05
0.05
0.05
62240.80
0.99
2.72
0.72
0.15
0.12
0.15
0.05
0.05
0.976
0.30
0.680
2181
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
Chemometric Evaluation of Darifenacin Hydrobromide
0.726, where the MeCN content, mobile phase pH, and column oven temperature considering the global optimum condition were 28%, 3.0, and
25.0 C, respectively; the predicted k0 and R were 4.46 and 2.45, respectively.
The mobile phase pH and organic modifier content play an important
role for elution and selectivity in RP=LC mechanisms and could have a
dramatic effect on separation of the ionizable compounds.[11,22] The 8.80
pKa and basic behavior in aqueous solution were predicted for DF by
ALOGPS2.1 simulator and the molecule until 7.8 would be in the ionized
form. The changes in flow rate did not demonstrate any effect on the resolution parameter, probably because the DF has a poor basic behavior,
without a free primary amine which could interact with the free silanols
in the stationary phase.[11,23]
Method Validation
The samples showed stability in 45 d under refrigeration (methanolic
solution) and 2 d at room temperature (mobile phase solution). The
parameters utilized for the present validation were 205 nm (wavelength),
1.1 mL min1 flow rate, mobile phase with MeCN and OPA (pH 3.0,
0.01% v=v) in at a 28:72 ratio and column maintained at room temperature
(25 1 C). The results are presented in Table 6, Table 7; and the chromatogram is shown in Figure 7. Resolution between peaks of darifenacin and
degradation product was adequate, resulting in peak purity (UV=PDA) for
the darifenacin peak. As seen in Table 6, LOQ and LOD results, using different approaches, are not similar. The greatest difference occurred between
TABLE 7 Analysis for Accuracy and Precision for the Method Validation
Accuracy
Range (%)
Mean (%)
RSDa (%)
97.45–99.18
98.33
0.88
Precision
Conditionb
Repeatability
N 1
N 2
N 3
Intermediary
a
Mean assay (%)
RSD (%)
99.43
101.12
98.33
0.52
0.78
0.60
99.63
1.32
Standard error.
N 1: Day 1, Analyst 1, Instrument 1; N 2: Day 2, Analyst 2, Instrument 2;
N 3: Day 3, Analyst 1, Instrument 2.
b
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
2182
L. Z. Meneghini et al.
FIGURE 7 Chromatogram showing hydrogen peroxide (1), DF (2) and secondary peak from oxidation
stress (3) performed on validation chromatographic parameters.
SDLOQ and the experimental value, because the point in LOQ is not only the
lowest quantification level but a compromise with the precision[18] and this
was only possible with 400.00 ng mL1 level (RSD < 1.0%). For the precision
study (Table 7), although were used variations involving analysts, days, and
instruments; the method showed adequate values for the RSD.
The results for the robustness study are presented in Figure 8. The
method can be considered robust because the significant factors pointedout in a half-normality curve (not shown) and Paretto chart (F factor for
percentage assay and A factor for tail) have been statistically different from
nominal condition, but they do not exceed the limits from the model requirements (2% RSD for the repeatability and 2.0 as the maximum value for
tail). When one or more dummy factors are excluded from the model, this
situation is maintained (not shown). Also, the other chromatographic
parameters (theoretical plates, retention factor, and resolution) presented
values within appropriate specifications.
FIGURE 8 Pareto charts (a, b) for the effects analysis showing the effects on assay % response (a) and (b)
showing the effects on tail, where A is acetonitrile content (%, v=v) and F is column over temperature.
Chemometric Evaluation of Darifenacin Hydrobromide
2183
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
CONCLUSIONS
The study successfully developed and validated a stability-indicating
RP-LC=UV method. The chemometric methodologies employed were useful to improve information about the influence of the factors on responses
R and k0 without taking a lot of time and being cost-consuming (no buffers,
low organic modifier content). It was only possible to optimize the chromatographic method using Derringer’s desirability, because resolution and
retention factor presented distinct behaviors when the factors were
changed. A robustness study provided more complete evaluation about
deliberate variations in the method. Finally, it can be said that the method
developed is fully applicable to the routine assay of the drug product.
ACKNOWLEDGMENTS
The authors thank CAPES and CNPq of Brazil, LANAGRO (Brazil) and
Labex Inc. (USA).
REFERENCES
1. Harvey, D. Modern Analytical Chemistry, McGraw-Hill: New York, NY, 2000.
2. Jancic-Stojanovic, I. J.; Malenovic, A.; Markovic, S.; Ivanovic, D.; Medenica, M. Optimization
and Validation of an RP-HPLC Method for Analysis of Hydrocortisone Acetate and Lidocaine in
Suppositorie. J. AOAC Int. 2010, 93 (1), 102–107.
3. Barmpalexis, P.; Kanaze, F. I.; Georgarakis, E. Developing and Optimizing a Validated Isocratic
Reversed-Phase High-Performance Liquid Chromatography Separation of Nimodipine and
Impurities in Tablets Using Experimental Design Methodology. J. Pharm. Biomed. Anal. 2009, 49,
1192–1202.
4. Vignaduzzo, S. E.; Castellano, P. M.; Kaufman, T. S. Experimentally Designed, Validated HPLC
Simultaneous Determination of Pridinol and Diclofenac in Their Combined Pharmaceutical
Formulations, Which Allows Limiting Diclofenac Related Compound A. J. Liq. Chromatogr. R. T.
2010, 33 (19), 1720–1732.
5. Novara, G.; Galfano, A.; Secco, S.; D’Elia, C.; Cavalleri, S.; Ficarra, V.; Artibani, W. A Systematic
Review and Meta-Analysis of Randomized Controlled Trials with Antimuscarinic Drugs for Overactive Bladder. Eur. Urol. 2008, 54, 740–764.
6. Kaye, B.; Herron, W. J.; Macrae, P. V.; Robinson, S.; Stopher, D. A.; Venn, R. F.; Wild, W. Rapid, Solid
Phase Extraction Technique for the High-Throughput Assay of Darifenacin in Human Plasma. Anal.
Chem. 1996, 68, 1658–1660.
7. Venn, R. F.; Goody, R. J. Synthesis and Properties of Molecular Imprints of Darifenacin: The
Potential of Molecular Imprinting for Bioanalysis. Chromatographia 1999, 50, 407–414.
8. Radhakrishnanand, P.; Rao, D. V. S.; Himabindu, V. H. A Validated LC Method for Determination of
the Enantiomeric Purity of Darifenacin in Bulk Drug and Extended Release Tablets. Chromatographia
2008, 68, 1059–1062.
9. Montgomery, D. C. Design and Analysis of Experiments, Wiley: New York, NY, 2001.
10. Brereton, R. G. Chemometrics: Data Analysis for the Laboratory and Chemical Plant, Wiley: New York,
2003.
11. Kazakevich, Y.; Lobrutto, R. LC for Pharmaceutical Scientists, Wiley: New Jersey, 2007.
12. Myers, R. H. Response Surface Methodology: Process and Product Optimization Using Designed Experiments,
Wiley: New Jersey, 2009.
Downloaded by [Leonardo Meneghini] at 08:50 07 November 2011
2184
L. Z. Meneghini et al.
13. Jancic-Stojanovica, B.; Malenovica, A.; Ivanovica, D.; Rakica, T.; Medenica, M. Chemometrical
Evaluation of Ropinirole and Its Impurity’s Chromatographic Behavior. J. Chromatogr. A 2009,
1216, 1263–1269.
14. Lewis, G. A.; Roger, D. M.; Phan, T. L. Pharmaceutical Experimental Design, Marcel Dekker: New York,
1999.
15. Sivakumar, T.; Manavalan, R.; Muralidharan, C.; Valliappan, K. Multi-Criteria Decision Making
Approach and Experimental Design as Chemometric Tools to Optimize HPLC Separation of
Domperidone and Pantoprazole. J. Pharm. Biomed. 2007, 43, 1842–1848.
16. ICH. International Conference on Harmonisation-Harmonised Tripartite Guideline: Validation of Analytical
Procedures: Text and Methodology, Q2(R1), 2005.
17. Dejaegher, B.; Heyden, Y. V. Ruggedness and Robustness Testing. J. Chromatogr. A 2007, 1158,
138–157.
18. Shabir, G. A. Validation of High-Performance Liquid Chromatography Methods for Pharmaceutical
Analysis Understanding the Differences and Similarities Between Validation Requirements of the
US Food and Drug Administration, the US Pharmacopeia and the International Conference on
Harmonization. J. Chromatogr. A 2003, 987, 57–66.
19. Mason, R. L.; Gunst, R. F.; Hess, J. L. Statistical Design and Analysis of Experiments, Wiley: New Jersey,
2003.
20. Bakshi, M.; Singh, S. Development of validated stability-indicating assay methods—critical review.
J. Pharm. Biomed. Anal. 2002, 28, 1011–1040.
21. USP. 31th edition. The United States Pharmacopeia, Rockville, United States, 2008.
22. Glajch, J. L.; Kirkland, J. J.; Squire, K. M. Optimization of solvent strength and selectivity for
reversed-phase liquid chromatography using an interactive mixture-design statistical technique.
J. Chromatogr. 1980, 199, 57–59.
23. McCalley, D. V. Effect of Temperature and Flow-Rate on Analysis of Basic Compounds in HighPerformance Liquid Chromatography using a Reversed-Phase Column. J. Chromatogr. A 2000, 902,
311–321.