8.5.5.5 Case Study 2: Concluding Remarks. Extreme changes in selectivity
and reversal of elution order of phenolic isomeric compounds were obtained
after changing either the pH of the mobile phase, the temperature of the
system, or the type of organic eluent employed. Changes in the analyte ion-
ization state were observed upon increasing the acetonitrile composition as
well as the temperature. Method development for ionizable analytes requires
a judicious choice of the mobile-phase conditions and system parameters in
order to perform the analysis of the compounds in their desired ionization
state. Choosing the optimal parameters in the “chromatographer’s toolbox”
allows for the development of rugged and reproducible methods.
8.5.6 Case Study 3: Method Development for a Diprotic Basic Compound
A case study is presented for the method development of a diprotic base com-
pound. The first step in method development is to look at the chemical struc-
ture of the analyte and to determine if there are any ionizable sites on the
molecule. If there are ionizable sites then their respective pK
a
values should
be determined. The pK
a
values may have been already determined by the pre-
formulation group and close communication with that group would avoid
duplication of work. However, commercially available programs such as ACD
Labs (Advanced Chemistry Development,Toronto, Canada) are also available
to allow for in-silico prediction of the analyte pK
a
. Also, using selected frag-
ments of the molecule can also be helpful for pK
a
determination of the desired
molecule because the pK
a

5.2). If one does not have a program to predict pK
a
, the pK
a
of this analyte
could be estimated to be close to that of meta chloro pyridine,
w
w
pK
a
2.95. The
other basic functionality contains a phenyl group attached to a morpholine
group. The pK
a
of morpholine is 8.8; but because the phenyl ring is attached
to the nitrogen group, this leads to resonance stabilization and consequently
leads to a reduction of the analyte pK
a
.
Note that because this compound does have multiple basic functionalities,
two ionization equilibria could be written for this amphoteric species. At
mobile
w
w
pH values between 3 and 5 the existence of multiple species is
expected. Since the two pK
a
values are close to one another (~2 pK
a
units

for determination of the suitable chromatographic conditions to analyze the
analyte.
If the pK
a
is known, the chromatographer can go directly to step 4.
8.5.6.2 Step 2: Determine the Isocratic Conditions for pH Scouting
Experiments. If the chromatographer intends to determine the chromato-
graphic pK
a
and understand the influence of mobile-phase pH on the target
analyte retention, pH scouting studies need to be performed in isocratic mode.
In order to begin this process, the appropriate set of isocratic conditions to
adequately retain the analyte in its fully ionized form and to elute the analyte
in its fully neutral state needs to be determined. Usually a steep gradient run
is used to estimate the initial isocratic elution conditions. In an example
shown in Figure 8-38 a probe linear gradient from 5 to 95 v/v% acetonitrile
METHOD DEVELOPMENT APPROACHES 407
Figure 8-38. Mobile phase
A: 0.2 v/v% H
3
PO
4
. Mobile phase B: Acetonitrile, linear
gradient from 5% B to 95% B over 10 min. Column: Luna C8(2) 150 × 4.6 mm.
Injection volume, 10 µL; flow, 1.0 mL/min; wavelength, 300 nm, column temperature,
35°C.
from 0–10 min was run and the target basic analyte (in its predominately
ionized form) eluted at 6.0 min. The
w
w

D
/F). Note that the velocity of the analyte moving through
the column under gradient conditions is not constant and follows a pseudo-
exponential profile (see Chapter 2, Section 2.17). The estimation given in
Scheme 1 serves as an approximation to determine the starting isocratic
elution conditions from the probe gradient run. By taking into account the
gradient slope of 9% ACN/min and accounting for the dwell time (1.5 min)
and elution time (6 min from Figure 8-38) of the analyte from the probe gra-
dient run, the estimated isocratic composition to elute the analyte at the same
retention as in the gradient probe run can be calculated as shown. The esti-
mated isocratic composition in which the analyte would elute at 6 min (k ~3.5)
is estimated as 41 v/v% acetonitrile ±10% acetonitrile using 0.2 v/v% H
3
PO
4
(
w
w
pH 1.9) mobile phase. The isocratic conditions chosen to perform the pH
scouting study was 30 v/v% acetonitrile.
408 METHOD DEVELOPMENT
Gradient slope × (Elution time from probe gradient run − Dwell time)
= Isocratic % organic composition ± 10% organic composition
9% ACN/min × (6 min − 1.5 min) = 41% ACN ±10% ACN
Gradient Slope =
90% ACN
min
ACN
min
Dwell Time =

a
values of the two ionization centers were predicted by ACD, and the
s
s
pH
values chosen were at least 1 unit less than the lowest
s
s
pK
a
in the molecule
and at least 1 unit greater than the highest
s
s
pK
a
of the molecule. Generally,
the pK
a
of a basic compound decreases by about 0.2 pK
a
units per 10 v/v% ace-
tonitrile (see Chapter 4, Sections 4.5 and 4.6) and if 30 v/v% acetonitrile is
used, it is expected to lead to a reduction of 0.6 pK
a
units (0.2*3 = 0.6 pK
a
unit
basic analyte pK
a

s
s
pK
a
of the molecule). The lower
s
s
pH for the mobile phase (containing 30 v/v%
MeCN) that should be prepared for this study should be 1.7, but this would
mean that an aqueous mobile-phase
w
w
pH of 1.1 would have to be prepared to
obtain a
s
s
pH of 1.7 (see Chapter 4, Section 4.5 for pH shift). Remember that
the pH shift of the mobile phase for a phosphate buffer is approximately
0.2 pH units in the upward direction for every 10 v/v% acetonitrile. In this
case, not to compromise the stability of the packing material (column chosen
has recommended a lower pH limit of
w
w
pH 1.5), a pH of
w
w
pH 1.6 was
chosen to be prepared which correlates to a
s
s

each successive pH experiments.
The k (retention factor) values are then plotted versus the
s
s
pH values, and
the inflection point of this sigmoidal relationship could be taken as the
s
s
pK
a
of that particular compound at particular hydro-organic mixture. The
s
s
pK
a
determined at 30 v/v% MeCN was determined to be 3.9 (using nonlinear
regression analysis program MathCad 8). This corresponds well to our origi-
nal estimation of
s
s
pK
a
3.7.
The retention of this analyte leveled off between pH values
w
w
pH 4.6 (
s
s
pH

s
s
pH t
R
6.1 6.7 49.8
4.6 5.2 49.1
3.6 4.2 34.3
2.6 3.2 11.2
2.1 2.7 6.0
1.6 2.2 3.8