measuring pkas, logp and solubility by automated titration
DESCRIPTION
Presentation by Sirius Analytical covering measurement of pKa, LogP, LogD, Solubility, Supersaturation and precipitation kinetics.For more details visit www.sirius-analytical.comTRANSCRIPT
© 2008
Sirius Analytical
Measuring pKas, logP and Solubility by Automated titration
Jon MoleTechnical Sales Manager
www.sirius-analytical.com
© 2008
Contents Introduction to Sirius
Overview of instrumentation & assays for pKa, logP/D
Validation studies
Principle of our “CheqSol” Solubility method
Our early theories
Four classes of solubility behaviour - implications
Modelling gastrointestinal precipitation/dissolution
Conclusions
2/ 70
© 2008
An introduction to Sirius
Sirius was founded in 1990 in the UK. We are a manufacturer and vendor of instrumentation for measurement of physicochemical parameters.
We also run an Analytical Service, and measure thousands of samples for hundreds of customers, worldwide, each year.
3/ 70
© 2008
Sirius locations
Sirius Analytical Ltd. Company headquarters in UK
– Manufacturing
– Engineering
– Software
– Chemistry R & D
– Administration
Located in Forest Row, East Sussex
30 minutes from London Gatwick Airport
Direct sales in some countries, distributors in others
Sirius Analytical Inc. Support for North American
customers– Instrument service
– Installation
– Training
– Sales
– Stock of parts
Located in Lakewood, NJ 60 minutes from Newark Airport
4/ 70
© 2008
What we do
We make instruments for measuring physicochemical properties of ionizable compounds
– pKa
– logP/D– Solubility– Dissolution
Widely accepted assays regarded as “gold standard”
Instruments installed at most major pharmaceutical companies
We also offer an analytical service for these parameters (and many more)
5/ 70
© 2008/ 706
Propranolol (a base): pKa = 9.53
O
NH2+
HO
H3C
H3C
O
NH
HO
H3C
H3C
BH+ B
BH+ B
Flumequine (an acid): pKa = 6.27
N
O OH
O
CH3
F
N
O O-
O
CH3
F
HA A-HA A-
pKa is the pH at which an ionisable group is “half-ionised”
© 2008/ 707
Most drugs ionize in solution Other properties (lipophilicity, solubility, permeability) are
pKa-dependent
In general:– neutral molecules are more easily absorbed by membranes– ionized molecules remain in plasma and are predominantly cleared by
renal excretion
Other reasons:Useful in formulation and salt selection. With knowledge of
pKa values, the species which is stable over the largest pH range can be selected.
Why is pKa important?
© 2008/ 708
2 or more acidic groups, no basic ~ 3%
1 basic group, no acidic ~ 42%
1 acidic group, no basic ~ 12%
Others ~ 3%1 basic group + 2 or more acidic ~ 3%
1 acidic group + 2 or more basic ~ 4%
1 acidic group + 1 basic ~ 8%
2 or more basic groups, no acidic ~ 25%
With thanks to Tim Mitchell and Ryszard Koblecki, Millennium Pharmaceuticals Ltd.
32,437 Ionizable drugs in World Drug Index (63% of total)
© 2008/ 709
Human Gastrointestinal (GI) Tract
STOMACH 0.1 m2
DUODENUM 0.1 m2
JEJUNUM 60 m2
ILEUM 60 m2
COLON 0.3 m2
pH (fasted)
4.6 (2.4 - 6.8)
6.1 (5.8 - 6.2)
1.7 (1.4 -2.1)
6.5 (6.0 - 7.0)
6.5
8.05.0 - 8.0
pH (fed)
5.0 (0.1 hr)
4.5 - 5.5 (1 hr)
4.7 (2 hr)
6.5
8.0
3-4 h small Intestinetransit time
© 2008/ 7010
P = partition coefficient.
The ratio of concentrations of
unionised species dissolved in
two immiscible solvents (e.g.
water + octanol) which are in
equilibrium.
D = Distribution Coefficient.
The ratio of all species
dissolved in two immiscible
solvents which are in
equilibrium.
water
octanol
species ionised unionised
species ionised unionised D
LogP and logD describe lipophilicity
P is constant D is pH-dependent
water
octanol
speciedunionised
speciesunionisedP
© 2008/ 7011
-2
-1
0
1
2
3
4
5
0 2 4 6 8 10 12 14
pH
log
D
DesipraminepKa = 10.14
DiphenhydraminepKa = 8.26
TriamterenepKa = 3.92
These molecules all have similar value for log D at pH 7.4.
Their lipophilicity profiles are quite different
Flat part of curve: log D = log P of neutral species
Big changes in lipophilicity occur over physiological pH range
Physiological pH range
DiclofenacpKa = 3.99
PhenobarbitalpKa = 7.43
NifuroximepKa = 10.56
Lipophilicity profiles are pKa and pH dependent
© 2008/ 7012
LogD at pH 7.4 Implications for drug development
Below 0 Intestinal and CNS permeability problems.
Susceptible to renal clearance.
0 to 1 May show a good balance between permeability and solubility.
At lower values, CNS permeability may suffer
1 to 3 Optimum range for CNS and non-CNS orally active drugs.
Low metabolic liabilities, generally good CNS penetration
3 to 5 Solubility tends to become lower. Metabolic liabilities increase
Above 5 Low solubility and poor oral bioavailability. Erratic absorption.
High metabolic liability, although potency may still be high.
Why is logP (and logD) important?
LogP and logD (lipophilicity) provide a rough guide to pharmacokinetic behavior.
© 2008/ 7013
Why is solubility important?
Poorly soluble molecules rarely make successful drugs– They are difficult to absorb, to formulate and to analyze
Discovery and lead optimization– it helps in identification of potential screening and bioavailability
issues– it is valuable in planning chemistry changes
Biopharmaceutical evaluation– it is important for the confirmation of bioavailability issues– during early trials of drugs, it is used in the design of animal
formulations, as well as for human formulation design Development
– solubility knowledge is needed for biopharmaceutical classification, biowaivers and bioequivalence
– it is also required for formulation optimization and salt selection Manufacturing
– solubility affects the optimization of manufacturing processes
© 2008/ 7014
Solubility is also pKa and pH dependent
NH
OH
O Cl
Cl
Compounds are more soluble when ionized
Lower plateau is the “intrinsic solubility”
Propranolol
Base
pKa = 9.54
S0 = 81μg/mL (314μM)
Diclofenac
Acid
pKa = 3.99
S0 = 0.9μg/mL (4.1μM)
O
NH
OH
CH3
CH3
© 2008
SiriusT3
Our “next generation” system
pKa
logP/DSolubilityBuilt in UV/VisSub-mg sample
requirementAutoloader for 192
samplesMore automation,
easier to use
www.sirius-analytical.com/products/SiriusT3.shtml15/ 70
© 2008
SiriusT3 - Dispenser Module
5 Mini dispensersCosolvent 6 way valveReagents storageArgon spargingUV/Vis diode array &
detector
16/ 70
© 2008
SiriusT3 – Titrator Module
Arm moves probes into calibration, wash and assay positions
Built in turbidity sensorPeltier temperature
controlArgon flowFlowing water wash for
optimum cleaning
17/ 70
© 2008
SiriusT3 – Autoloader Module
4 x 48 vial racksStandard footprintRobotic gripper arm to
automatically move samples
Built in ultra-sonic bath
18/ 70
© 2008
Titration cell for SiriusT3
Capillaries, for adding reagents
pH electrode, diameter 3mm
Glass vial, 4 ml total capacity
Electronic thermometer
Automatic overhead
Stirrer
Probes require a minimum of 0.5mL of solution contained in glass vial. Typical assay volume = 1ml.
UV Dip Probe.
19/ 70
© 2008
pH-metric pKa – experimental process
Set up assay
– SiriusT3 software includes templates for all assays
Data Collection
– weigh sample into vial (or dispense stock solution)
– instrument adds water (or water-CoSolvent)
– instrument adjusts pH, then titrates with acid or base
Calculation of results (in software)
– pKa result obtained by analyzing the shape of the titration curve
– Calculation fully automated in SiriusT3 software
Titration of flumequine in 39.8% methanol
20/ 70
© 2008
A refined solution
21/ 70
© 2008
½Á
(b
ou
nd
H¤
per
Van
)
Example: vancomycinMeasured pKas:
Van4- + H+ VanH3- pKa6 11.88 ± 0.01
VanH3- + H+ VanH22- pKa5 10.15 ± 0.01
VanH22- + H+ VanH3
- pKa4 9.28 ± 0.01
VanH3- + H+ VanH4 pKa3 8.62 ± 0.01
VanH4 + H+ VanH5- pKa2 7.48 ± 0.01
VanH5+ + H+ VanH62+ pKa1 2.64 ± 0.01
0
1
2
3
4
5
6
7
1 2 3 4 5 6 7 8 9 10 11 12 13 pH (concentration scale)
Difference Curve
0
50
100
% S
pe
cie
s
1 2 3 4 5 6 7 8 9 10 11 12 13
pH
Distribution of species
VanH5
Difference Curve can handle 1, 2 or several pKas
OHHOOH
HN
HOOC H
ONH
O O
Cl
NH
O
ClH
HHO
HHN
H HN
O
OHH
NH
NHCH3
HC H
O
OO
O
O
O
O
H
OH
HO
CH2OH
CH3
NH2
H3C
HO
O
H2N
H3C
CH3
22/ 70
© 2008 * D-PAS = Dip-Probe Absorption Spectroscopy* D-PAS = Dip-Probe Absorption Spectroscopy
pKa measurement by UV
Multi-wavelength UV technique
220 to 750nm, diode array
Fibre optic dip probe allows spectral
measurement during titration
Less sample required (down to 10-6 M)
3uL of 10mM Stock
equally sensitive over entire pH range
pKas measured below 1, above 13
Allows fast pKa measurement (just 4 mins)23/ 70
© 2008
Set up assay
– SiriusT3 software has built-in templates for UV pKa assays
Data Collection
– Prepare 10mM stock solution of sample in DMSO
– Pipette 3μL of stock solution into vial
– instrument adds water (or water + CoSolvent) & buffer
– instrument adjusts pH, then titrates with acid or base
Calculation of results (in software)
– Target Factor Analysis (TFA) method finds pKas, even when they are
overlapping or spectral change as a function of ionization is small
– Calculation fully automated in SiriusT3 software
UV pKa - experimental
24/ 70
© 2008
3-D spectrum: pH vs. absorbance vs.
wavelength
pKa from UV spectra
25/ 70
© 2008
pKa from UV spectra
26/ 70
© 2008
pKa result calculated by Target Factor Analysis (TFA)
pKa Result
27 / 70
© 2008
Using CoSolvents
Dissolve sample in water-miscible CoSolvent + water
– Water-CoSolvent should be ionic-strength-adjusted to 0.15M with KCl
– Solvents supported fully:
methanol (80%) 1,4 dioxane (60%)
DMSO (60%) ethanol (60%)
ethylene glycol (60%) DMF (60%)
THF (60%) Acetonitrile
(50%)
MDM-mix (20% Methanol, 20% dioxane, 20% acetonitrile) (60%)
Iso-propyl alcohol
28/ 70
© 2008
Using Sirius: Quinine hydrochloride pKa
Sample weighed into vial
SiriusT3 adds CoSolvent + water (45 wt%
methanol), lowers pH and titrates with KOH (green
curve)
SiriusT3 adds more water (now 35 wt%
methanol), lowers pH and titrates with KOH (red
curve)
SiriusT3 adds more water (now 25 wt%
methanol), lowers pH and titrates with KOH (blue
curve)
Results from Yasuda-Shedlovsky extrapolation:
– pKa1 = 8.49 ± 0.02
– pKa2 = 4.22 ± 0.01
N
O
CH3
HO
H
N+
H
H
H
Cl-
Example of a CoSolvent titration
29/ 70
© 2008
3-aminobenzoic acid
4
5
6
7
12 13 14 15 16 17 181/e
p¿K¾+log[H
2O]
...using CoSolvent psKa titrations
COOH
NH3+
COOH
NH2
COO-
NH2
Yasuda-Shedlovsky slope direction:
up (red) = acidic group
down (blue) = basic group
3-aminobenzoic acid: an example of an ordinary ampholyte
Assigning pKas to ionizable groups
30/ 70
© 2008
Detecting precipitation
SiriusT3 has built in turbidity detection .
Shaded area shows pH where sample precipitated.
This is a warning, do not use this data to determine pKa of miconazole!
Repeat in cosolvent to avoid precipitation and get reliable pKa data.
N
N
O
Cl
Cl Cl
Cl
31/ 70
© 2008
Set up assay
– SiriusT3 software has templates for logP assays
Data Collection
– weigh sample into vial (or use stock solution).
– instrument adds water and octanol
– instrument adjusts pH, then titrates with acid or base
Calculation of results (in software)
– logP result obtained by analyzing the shape of the titration curve
(procedure requires pKa value)
– Calculation fully automated in SiriusT3 software
pH-metric logP - experimental
32/ 70
© 2008
Principles of pH-metric logP measurement
]][X[H
][XHlog pK
0a
aq0
oct0
][X
][Xlog logP
A solution of the sample is titrated in a two-phase system (water + octanol)
The sample can ionise in water (pKa), or it can partition into octanol (logP)
The presence of the octanol disturbs the pKa equilibrium.
The pKa shifts to a new value (poKa) to minimise this disturbance. We calculate the logP from this shift in pKa.
33/ 70
© 2008
Flumequine (acid)pKa = 6.27, poKa = 7.99 log P = 1.72
Lipophilicity profiles: these
profiles are correct for high logD, but
do not show partitioning of ionic species
Diacetylmorphine (base)pKa = 7.95, poKa = 6.37 logP = 1.58 O
O
O
H
N
CH3
O
CH3
H3C O
Titrations with equal volumes of water and octanol
Aqueous pKa poKa
N
O OH
O
CH3
F
34/ 70
© 2008
Shake flask vs. pH-metric
line calculated using log D
equation for monoprotic base,
using pKa = 9.54, log P0 = 1.83,
log P1 = -1.32 (pH-metric data,
0.15M KCl, 25°C [8])
[8] Caron, G., Steyaert, G., Pagliara, A., Reymond, F., Crivori, P., Gaillard, P., Carrupt, P.A., Avdeef, A., Comer, J., Box, K.J., Girault, H.H., Testa, B. Helv Chim Acta. 82, 1211-1222 (1999)
[9] Barbato, F., Caliendo, G., Larotonda, M.I., Morrica, P., Silipo, C., Vittoria, A. Farmaco. 45, 647-663 (1990)
HN
OHN
OH
Pindolol
points from shake-flask experiments, various
buffers, 0.1M [9]
35/ 70
© 2008
The world’s most powerful system for measuring pKa
SiriusT3 is our third generation instrument.
We have spent over 15 years of continuous research and development on improving our assays and calculations for pKa, logP and solubility measurement.
If a sample has a pKa between 2 and 12, we can always measure it on the SiriusT3 system.
Every year we measure pKa of hundreds of samples that customers send us, and we never fail (provided the sample has a pKa and is chemically stable and pure).
36/ 70
© 2008
R.I. Allen, K.J. Box, J.E.A. Comer, C. Peake, K.Y. Tam, J. Pharm. Biomed. Anal., 17, 699-712, 1998.K.Y. Tam, K. Takács-Novák, Pharm. Research,. 1999, 16, 374-381R.C. Mitchell, C.J. Salter, K.Y. Tam, J. Pharm. Biomed. Anal., 1999, 20, 289-295 K.Y. Tam, M. Hadley, W. Patterson, Talanta, 1999, 49, 539-546
Validation of UV pKa method
0
2
4
6
8
10
12
0 2 4 6 8 10 12pH-metric pKa
UV
pK
a
pKa (spec) = 1.006 x pKa (pH-metric) n = 31 R2 = 0.999 RMSD = 0.098pKa (spec) = 1.006 x pKa (pH-metric) n = 31 R2 = 0.999 RMSD = 0.098
Benzoic acid (3.98)Icotidine (3.29, 5.39, 6.22, 9.97)Lupitidine (2.79, 5.96, 8.25, 9.66)Nicotinic acid (2.10, 4.63)Nitrazepam (2.90, 10.39)Niflumic acid (2.28, 4.86)m-aminobenzoic acid (3.17, 4.54)p-aminosalicylic acid (1.79, 3.58)Phthalic acid (2.70, 4.86)Phenol (9.73)Phenolphthalein (8.87, 9.35)Pyridoxine (4.90, 8.91)Quinine (4.33, 8.59)SB-221789 (2.74)SKF-75250 (1.48, 6.59)
(measured at 25ºC and an ionic strength of 0.15 M)
37/ 70
© 2008
B. Slater, A. McCormack, A. Avdeef and J.E.A. Comer, J. Pharm. Sci. 1994, 83,1280-1283
J.E.A. Comer, K. Chamberlain and A. Evans in J. Devillers (Ed.), SAR QSAR Environ. Res., Vol.3 Issue 4; Molecular Descriptors, Gordon and Breach, Philadelphia 1995, pp. 307-313.
K. Takács-Novák and A. Avdeef, J. Pharm. Biomed. Anal. 1996, 14,1405-141;
Validation of pH-metric logP method
61 samples over eight logP units
amino acids, peptides, ampholytes, barbiturates, ß-blockers, herbicides, phenols, various others
61 samples over eight logP units
amino acids, peptides, ampholytes, barbiturates, ß-blockers, herbicides, phenols, various others
Graph plotted using Polyfit program from RefinementPro 2
38/ 70
© 2008
CheqSol Technology – for Solubility measurement
In 2004, Sirius introduced a new patented technology for measuring
solubility of ionizable drugs.
Recent investigations at Sirius using our CheqSol solubility assay has given
us some interesting insights into supersaturation effects and the
relationships between dissolution and precipitation rates for a range of
drugs.
Our latest research indicates that molecules can be placed into one of four
classes: Chasers, Non-Chasers, Super-Dissolvers and Ghosts.
39/ 70
© 2008
Introduction to CheqSol
Unique method for solubility measurement
Runs on Sirius GLpKa, PCA200 & SiriusT3 instruments
Requires pKa value
Uses “Chasing Equilibrium” process to determine intrinsic solubility
40/ 70
© 2008
Sirius definitions of solubility
Kinetic Solubility is the concentration of a compound in solution at the time when an induced precipitate first appears
Equilibrium Solubility* is the concentration of compound in a saturated solution when excess solid is present, and solution and solid are at equilibrium
Intrinsic Solubility ** is the equilibrium solubility of the free acid or base form of an ionizable compound at a pH where it is fully un-ionized
* also called Thermodynamic Solubility** Hörter, D.; Dressman, J. B. Adv. Drug Deliv. Rev., 1997, 25, 3-14
41/ 70
© 2008
Solid added to vial. (5 to 20mg on GLpKa)(0.5 to 2mg on SiriusT3)
Instrument adds water (or water-cosolvent), then adjusts pH to dissolve sample.
Solution titrated towards the pH where the sample becomes neutral.Eventually it precipitates
Precipitation causes light scattering, and system detects this as an increase in the light absorbed.
Kinetic solubility determined at point of precipitation.
Starting the CheqSol Assay - Seeking precipitation
Before precipitation, no light is absorbed by the solution
Ab
sorb
an
ce
0.0
0.8
1.6
2.4
3.2
200 300 400 500 600 700
After precipitation, most light is scattered
Wavelength (nm)
0.0
0.8
1.6
2.4
3.2
200 300 400 500 600 700
42/ 70
© 2008/ 7043
The Bjerrum Graph: a graphical view of solubility
Bjerrum Graphs provide a graphical view of solubility
They are theoretical curves plotted using pH, pKa, solubility and concentration of sample
They are related to the Distribution of Species graph
– Graph below shows distribution of species of Pindolol in aqueous solution
The next slide shows the Bjerrum function Bj vs. pH
BBH+
pH
2 4 6 8 10 12
% S
peci
es
0
50
100
pKa = 9.54
© 2008/ 7044
Sample = base with one pKa
Understanding the Bjerrum Graph
Precipitate is present
pH
0.0
0.5
1.0
2 4 6 8 10 12
pH = pKa
Sample is unionised at
this pH
Sample is ionised at this pH
Bj =
Mo
les
of
boun
d H
+
ions
per
mol
e of
sam
ple
Precipitation Bjerrum Graph. For a base with one pKa,
atotal
0
]K[X][HS
Bj
Solution Bjerrum Graph. For a base with one pKa ,
aK][H
][HBj
1.0
BBH+
pH
2 4 6 8 10 12
% S
peci
es
0
50
100
© 2008/ 7045
Sample = base with one pKa
Understanding the Bjerrum Graph
This distance depends on:
Solubility (for a given concentration, distance increases as solubility decreases)
Concentration (for a given solubility, distance increases as concentration increases)
Precipitate is present
pH
0.0
0.5
1.0
2 4 6 8 10 12
pH = pKa
Sample is unionised at
this pH
Sample is ionised at this pH
Bj =
Mo
les
of
boun
d H
+
ions
per
mol
e of
sam
ple
Precipitation Bjerrum Graph. For a base with one pKa,
atotal
0
]K[X][HS
Bj
Solution Bjerrum Graph. For a base with one pKa ,
aK][H
][HBj
1.0
pH
0.0
1.0
2 4 6 8 10
0.5
Sample = acid with one pKa
Sample is ionised at
this pH
Sample is unionised at this pH
© 2008
CheqSol example – solubility of Pindolol (a chaser)
Pindolol is a beta-blocker, used to reduce hypertension
It’s a secondary amine with pKa of 9.54 (25°C, 0.15M ionic strength)
The neutral form B is poorly soluble in water at high pH
The ionized form BH+ is soluble at low pH
BH+
solubleBinsoluble
NH
ONH
OH
CH3
CH3
NH
ONH2+
OH
CH3
CH3
46/ 70
© 2008
Pindolol - Chasing method
For Pindolol, the kinetic solubility falls on a different Precipitation Bjerrum Graph to the rest of the data
Intrinsic solubility is determined from the data points on the Equilibrium Precipitation Graph, as explained in the following slides
Kinetic solubility is higher than Intrinsic solubility Assay took 37 minutes to measure kinetic and Intrinsic solubility
pH
0.0
0.5
1.0
0 2 4 6 8 10 12 14
Mole
s of
bound H
+
ions
per
mole
of
sam
ple
Precipitation detected at pH 9.07 (Kinetic point)
While chasing equilibrium, all data
points fall on the Equilibrium
Precipitation Graph
Equilibrium Precipitation Graph
Kinetic Precipitation
Graph
47/ 70
© 2008
Overview of Chasing Equilibrium
CheqSol adds HCl or KOH solution while precipitate is present, and records the rate of pH change*
*after waiting until the onset of sustained response
This forces the neutral species to cycle between two states:
Between these states, a point will be crossed where the concentration of neutral species would be at equilibrium
This technique is called Chasing Equilibrium
NOTE: CheqSol is short for Chasing equilibrium Solubility
CheqSol was invented in April 2004 at Sirius. Sirius have a patent for the CheqSol method.
supersaturated (excess neutral species in solution)subsaturated (excess undissolved neutral species)
48/ 70
© 2008
When KOH is added to raise the pH, the solution of Pindolol becomes supersaturated before it precipitates
When precipitate first appears there is a good deal of dissolved unionized material B, but it could take hours before it all falls out of solution
KOH addition pauses when precipitation has been detected Molecules of B interact to form particles of precipitate, and
BH+ ions convert to B to replace some the B that was lost. This releases H+ ions, and the pH goes down
Supersaturated Pindolol
B(aq) + H+
B(s)
BH+ (aq)
Solid State
Solution
CheqSol reports the gradient of this line
0.183dt
dpH
pH
Time (s)
8.6
8.7
8.8
30 40 50 60
49/ 70
© 2008
Subsaturated Pindolol
After recording the linear fit to the gradient, CheqSol adds HCl to lower the pH.
Some of the dissolved B converts to BH+ in solution The solution becomes subsaturated, i.e. there is an excess of
precipitate that could dissolveB dissolves. Some of it converts to BH+ in solution. This
consumes H+ ions, and the pH goes up
pH
Time (s)
8.5832
8.5840
8.5848
8.5856
0 25
00570dt
dpH.
50/ 70
© 2008
Chasing Equilibrium continues until a graph like this can be drawn. In the graph above, there are eight crossing points.
After collecting a specified number of crossing points, the instrument adjusts pH back to the starting pH to re-dissolve the sample, then cleans the probes for the next experiment.
The Crossing Point Graph for Pindolol
Black lines and circles - nothing addedBlue lines and triangles - KOH addedRed lines and triangles - HCl added
The system would be at equilibrium at the crossing points
dp
H/d
t
-0.04
0
0.04
36 56Time (minutes)
Subsaturated basic sample is dissolving
Supersaturated basic sample is precipitating
Crossing points
51/ 70
© 2008
dp
H/d
t
Concentration (µg/mL)
-0.04
0
0.04
0 10 20 80 90
41.32 µg/mL
dp
H/d
t
-0.04
0
0.04
6 36Time (minutes)
Subsaturated basic sample is dissolving
Supersaturated basic sample is precipitating
Crossing points
The concentration of unionized species at each point in the Crossing Point Graph is calculated. This requires
– weight of sample– total volume of solution– concentrations and volumes of
acid and base dispensed– pH at each point
– pKa(s) of sample, both value and type (acid, base)*
Gradient vs. concentration is plotted in the graph below
The average value of all crossing points is the concentration of the unionized species at equilibrium.
This is the Intrinsic Solubility CV shows the quality of assay
Calculating the result for Pindolol
CV
* The procedure is sensitive to errors in pKa – an error of 1 pKa causes an error of 1 logS unit52/ 70
© 2008
Pindolol is a chaser
Pindolol is a chaser because its kinetic solubility is significantly higher than its Intrinsic solubility
The neutral species of Pindolol forms a supersaturated aqueous solution
It precipitates slowly, and would take a long time for all the substance to precipitate at a given pH
If Pindolol were to pass from the stomach to the upper intestine, it may be expected to form a supersaturated solution before precipitation – this may help to drive absorption.
The ratio between the kinetic solubility and intrinsic solubility provides an indication of the supersaturation factor – a useful number for modelling absorption?
53/ 70
© 2008
Everything a “Chaser”?
When we initially discovered CheqSol – we thought every drug would supersaturate to some degree, and therefore “Chase Equilibrium”
Our first paper presents 6 chasers:
Stuart, M. Box, K. Chasing equilibrium: measuring the intrinsic solubility of weak acids and bases. Anal. Chem. 2005 (77(4)) pp 983-990
54/ 70
© 2008
Solubility of six compounds from our first paperpKa Sample
weight (mg)
Time taken (min)
Kinetic solubility (µg/mL)
Intrinsic solubility (µg/mL)
33 45 ± 6 0.9 ± 0.1 0.8 ± 0.2 [1]
43 180 ± 10 50 ± 4 49 ± 2 [1]
79 4600 ± 900 3500 ± 100 3810 ± 20 [3]Lidocaine 7.95 96-280
3.4-24
Ibuprofen 4.35 6.2-51
Diclofenac 3.99
CheqSolIntrinsic solubility (µg/mL)
Literature
All measurements at 25ºC in aqueous 0.15M KCl solution Results for Propranolol and Famotidine are the mean of 6 (others are mean of 10) Time taken includes dissolution time
61 5900 ± 650 740 ± 40 1100 ± 200 [1]
60 120 ± 1 5.3 ± 0.2 5.6 ± 0.3 [2]
60 340 ± 20 81 ± 6 70 ± 20 [1] Propranolol 9.54 10-19
102-123
Warfarin 4.94 10-12
Famotidine 6.77, 11.01
[1] Avdeef, A. Berger, C M. Brownell, C. Pharm. Res. 2000, 17 (10, 85-89[2] Bergström, C A S. Strafford, M. Lazorova, L. Avdeef, A. Luthman, K. Artusson, P. J. Med. Chem. 2003, 46, 558-570[3] Powell, M F. in Analytical Profiles of Drug Substances: Florey, K (ed); Academic \Press, San Diego 1986, 15, 761-779
55/ 70
© 2008
Everything a “Chaser”?
We then discovered some compounds that did not follow the “Chasing Equilibrium” process.
We named these compounds “Non-Chasers”
Example – Verapamil (tertiary amine – a base with pKa of 8.72 @25°C, 0.15M ionic strength)
N
NH+
O
O
O
O
CH3
CH3
CH3
CH3
CH3
CH3
CH3N
N
O
O
O
O
CH3
CH3
CH3
CH3
CH3
CH3
CH3
BH+
solubleBinsoluble
56/ 70
© 2008
Verapamil Bjerrum Curve
For Verapamil, all points collected, including the kinetic solubility, fall on the Precipitation Bjerrum Graph
After titrating with base, the instrument adds acid to check that the points are still on the Precipitation Bjerrum Graph. If they are, then …
Kinetic solubility = Intrinsic solubility = the solubility value required to fit a precipitation curve to the data points
Assay took just 19 minutes to measure kinetic and Intrinsic solubility
Precipitation detected at pH 7.82 (Kinetic point)
Mole
s of
bound H
+
ions
per
mole
of
sam
ple
pH
0.0
0.5
1.0
0 2 4 6 8 10 12 14
All data points collected after precipitation fall on
the same Precipitation Bjerrum Graph as the
Kinetic Point
57/ 70
© 2008
Other Non-Chasers
Chlorpromazine, Imipramine, Quinine, Amitryptyline, Diphenhydramine, Nortriptyline, Desipramine, Diltiazem, Deprenyl.
These compounds would not supersaturate and therefore would precipitate as soon as they exceed their solubility limit.
Questions raised:– What implications does this have for oral absorption?– What determines the degree to which a compound will
supersaturate or not?– Can we predict supersaturation behaviour from structure?– Do chasers precipitate and dissolve at equal rates?
58/ 70
© 2008
Can we predict whether a sample is a non-chaser?
N
N
CH3
CH3
Imipramine
Non-chaser
N
CH3
CH3
Amitryptyline
Non-chaser
Chlorpromazine
N
S
N
Cl
CH3
CH3Non-chaser
N
NHCH3
Desipramine
Non-chaser
NHCH3
Nortriptyline
Non-chaser
NHCH3
Maprotiline
Chaser
Secondary and tertiary amines with logP > 4.
S
N
Cl
CH3
CH3
Chlorprothixene
Non-chaser converts to
chaser
N
N
CH3
CH3
CH3
Similar structures, but maprotiline contains a -CH2-CH2- bridge.
Non-chaser
Trimipramine
59/ 70
© 2008
N
N
O
O
O
O
CH3
CH3
CH3
CH3
CH3
CH3
CH3
More non-chasers.…..
Amiodarone
Verapamil
O
O
I I
ON
CH3
CH3
CH3
N
N
CH2
OH
OCH3
H
H
H
Quinine
N
S
O
O O
N
O
CH3
CH3
CH3
CH3
H
H
Diltiazem
CHN
CH3
CH3
Deprenyl
60/ 70
© 2008
N
N
O
OH
Cl
CH3
CH3
O NH
OH
OH
OH
CH3
CH3
CH3
..….. and some chasers
Terfenadine
Nadolol
LoperamideNOH
OH
CH3
CH3CH3 ONH
NH2
Cl
O
N CH3
CH3
CH3
Metoclopramide
NH
N
N
Cl
OH
CH3
CH3
Amodiaquin
N
N
NH2
NH2
Cl
CH3
Pyrimethamine
61/ 70
© 2008
Investigating precipitation and dissolution behaviour
Piroxicam
SN
NH
O O
O
NOH
CH3
SNHN
O
O
N
NH2
CH3
SulfamerazineSupersaturated acidic sample
Subsaturated acidic sample
Subsaturated acidic sample
Supersaturated acidic sample
Most of the early compounds we investigated show a tight symmetry.
62/ 70
© 2008
Investigating precipitation and dissolution behaviour
NO
O
O
O
CH3
CH3
CH3
CH3
Papaverine
Furosemide
S
NH
OO
O
NH2
O
Cl
OHSupersaturated acidic sample
Subsaturated acidic sample
Subsaturated basic sample
Supersaturated basic sample
Some compounds show a clear offset!
63/ 70
© 2008
Using the Precipitation Rate graph to investigate ~100 ionisable drugs, we have found that there appears to be four classes of behaviour.
Slow Precipitator Fast Precipitator
Slow Dissolver
“Chasers”
Slow rate for bothprecipitation and dissolution
Examples:Ibuprofen, Benzocaine, Benzthiazide.
“Non-Chasers”
Fast rate of precipitation, slow rate of dissolving
Examples:Nortriptyline, Amitryptyline, Imipramine.
Fast Dissolver
“Super Dissolvers”
Slow rate of precipitation,Fast rate of dissolving
Examples:Tolmetin, Papaverine, Chlorzoxazone.
“Ghosts”
Fast rate for precipitation and dissolution
Examples:None yet discovered.
The Four Class Model
64/ 70
© 2008
CheqSol Technology A weak base might dissolve fully in the stomach but precipitate on entering the high pH environment of the upper intestinal tract. Can the patterns we observe in CheqSol be used to identify which formulation/delivery methods can be used to improve bioavailability?
Does the supersaturation exhibited by “chasers” mean that the bioavailability is already enhanced over what the thermodynamic properties imply, and thus further formulation/delivery work is unwarranted?
Do non-chasers fall out of solution as amorphous material whereas chasers produce crystalline precipitate?
Amorphous materials are amenable to solid state dispersion nanoparticle delivery methods.
Alternatively, could a formulation technique be used to keep a supersaturated sample in a supersaturated state for longer than expected?
Are the properties we observe inherent to the compound, or can they be changed by the use of excipients, milling techniques etc.?
CheqSol is a unique tool for investigating the precipitation characteristics of a drug.65/ 70
© 2008
Kinetic Solubility
Kinetic Solubility
CheqSol Shake-Flask Literature Chaser non-chaser
1 Phthalic Acid 5330 5950 8462
2 Quinine 363 201 491 391
3 Trazodone 134.6 138.0 435
4 Nitrofurantoin 112.5 109.5 78.9 319
5 Nortriptyline 27.0 49.3 20.0 27.3
6 Verapamil 48.5 48.5 9.7 47.8
7 Niflumic Acid 9.53 29.5 59
8 Imipramine 17.2 21.7 18.1 17.3
9 Flumequine 34.2 20.7 121
10 Furosemide 19.7 20.4 5.9 96
11 Maprotiline 5.80 8.05 3.49 77
12 Piroxicam 5.92 5.95 3.16 233
13 Warfarin 5.30 5.25 5.60 120
14 Chlorpromazine 2.70 2.41 1.71 2.70
15 Lidocaine 3500 3810 4600
16 Famotidine 740 1100 5900
17 Hydrochlorothiazide 630 700 2400
18 Chlorpheniramine 608.3 615.2 668
19 Sulfamerazine 200.3 203.0 701
20 Ketoprofen 130.6 178.0 336
21 Propranolol 81.0 70.0 340
22 Ibuprofen 50.0 49.0 180
23 Pindolol 41.7 32.7 1424
24 Miconazole 1.00 0.67
25 Diclofenac 0.90 0.80 45
26 Amodiaquin 0.41 8.8
27 Pamoic acid 0.0003 0.019
All results in µg/mL
Name Equilibrium solubility
Validation of CheqSol solubility method
19 compounds in this group – see next slide
CheqSol vs. Shake-flask results for compounds 1– 14
Six replicate shake-flask experiments compared with six CheqSol experiments for each compound
Box, K J. Völgyi, G. Baka, E. Stuart, M. Takács-Novák, K. Comer, J E A. J. Pharm. Sci. 2006, in press
66/ 70
© 2008
CheqSol validation
19 compounds with low solubility Good correlation between CheqSol and Shake-flask results
67/ 70
© 2008
Conclusion
Sirius are experts in physicochemical measurement, serving a global market.
Sirius provide several turn-key tools for measuring important physicochemical parameters - pKa, LogP/D and Solubility.
Our unique CheqSol assay measures solubility AND supersaturation
Our unique CheqSol assay provides information on dissolution AND precipitation rates
Highly automated instrumentation is available
We are always working to improve and update our instruments and software through innovative research, publications and and collaboration.
68/ 70
© 2008
Sources of Further Information
Website:
www.sirius-analytical.com• Overview of Sirius, our products and our technologies
• Detailed list of all our literature publications
• Brochure Downloads
Our recent publications:
• Stuart, M. Box, K. Chasing equilibrium: measuring the intrinsic solubility of weak acids and bases. Anal. Chem. 2005, 77(4), 983-990
• Box, K J. Völgyi, G. Baka, E. Stuart, M. Takács-Novák, K. Comer, J E A. Equilibrium vs. kinetic measurements of aqueous solubility, and the ability of compounds to supersaturate in solution - a validation study. J. Pharm. Sci. 2006, 95, 1298-1307.
• Sköld, C. Winiwarter, S. Johan Wernevik, J. Bergström, F. Engström, L. Allen, R. Box, K. Comer, J. Mole, J. Hallberg, A. Lennernäs, H. Lundstedt, T. Ungell, A-L. Karlén, A. Presentation of a Structurally Diverse and Commercially Available Drug Data Set for Correlation and Benchmarking Studies. J. Med. Chem. 2006, 49(23), 6660-6671
• Llinàs, A. Burley, J C. Box, K J. Glen, R C. Goodman, J M. J. Diclofenac Solubility: Independent Determination of the Intrinsic Solubility of Three Crystal Forms. Med. Chem.; 2007, 50 (5), 979-983 Collaborative research with the University of Cambridge
• Llinàs, A. Box, K J. Burley, J C.Glen, R C. Goodman, J M.J. A new method for the reproducible generation of polymorphs: two forms of Sulindac with very different solubilities. J. Applied Crystallography, 2007, 40(2), 379-381. Collaboration with the University of Cambridge.
69/ 70