Screening of a Sulfonamides Library by Supercritical Fluid Chromatography
description
Transcript of Screening of a Sulfonamides Library by Supercritical Fluid Chromatography
Screening of a Sulfonamides Library
by
Supercritical Fluid Chromatography
Coupled to Mass Spectrometry (SFC-MS).
Preliminary properties-retention study.
SFC User Meeting Cambridge
31st May – 2nd June 2006
Amaury Cazenave GassiotDr. G. John Langley
SFC User Meeting - Cambridge - 2006
2
OUTLINE
Introduction
SOTLIB Library
Isocratic study
Retention models in chromatography
Polycratic study
- Looking for linear relationships
- Correlating data with physico-chemical properties
Conclusion and future work
SFC User Meeting - Cambridge - 2006
3
INTRODUCTION
SFC-MS appears more and more as a technique complementary to HPLC for high throughput analysis
Importance of knowing which technique is more suitable for a specific type of analytes
Possibility of outlining a set of properties-based rules allowing prediction of the retention of a given compound by SFC ?
To investigate those questions: design of a small library of sulfonamides compounds to be screened by SFC
SFC User Meeting - Cambridge - 2006
4
SOTLIB LIBRARY: NEUTRAL SET Coupling of 4 sulfonyl chlorides with 3 amines
All compounds synthesized and analysed by SFC
S
O
O
N 1
S
O
O
N
F
F
F4
S
O
O
N 7
S
O
O
NNH2 10
S
O
O
N2
S
O
O
NF
F
F 5
S
O
O
N8
S
O
O
NNH2
11
S
O
O
NN
3
S
O
O
NN
F
FF
6
S
O
O
NN 9
S
O
O
NN
NH2 12
SFC User Meeting - Cambridge - 2006
5
SOTLIB LIBRARY: BASIC SET Coupling of 4 sulfonyl chlorides with 5 amines
All compounds synthesized and analysed by SFC
S
O
O
N N
13
NNS
O
O
14
NN
S
O
O
15
S
O
O
N
NEt2
NEt216
S
O
O
N
N
17
S
O
O
N
N
F
F
F
22
S
O
O
N
NEt2
NEt2
F
F
F
21
NN
S
O
O
F
F
F
20
NNS
O
O
F
F
F
19
S
O
O
N N
F
F
F
18
S
O
O
N N
23
NNS
O
O
24
NN
S
O
O
25
S
O
O
N
NEt2
NEt226
S
O
O
N
N
27
S
O
O
N
N
NH2
32
S
O
O
NNH2
NEt2
NEt231
NN
S
O
O
NH2
30
NNS
O
O
NH2
29
S
O
O
N NNH2
28
SFC User Meeting - Cambridge - 2006
6
ISOCRATIC STUDY: NEUTRAL SET (I)
Screening of the library using isocratic elution
Neutral set studied on:
- 2-ethyl-pyridyl column 4.6x250mm- cyano column 4.6x250mm- diol column 4.6x250mm- bare silica column 4.6x250mm
Chromatographic Conditions:
- mobile phase: 20% v/v MeOH in CO2 - flow: 4mL.min-1
- temperature: 35°C- outlet pressure: 100 bars
- injection volume: 4mL (~0.5mg.mL-1)
SFC User Meeting - Cambridge - 2006
7
S
O
O
N
F
F
F S
O
O
N
S
O
O
N
S
O
O
NNH2
S
O
O
NN
F
FF
S
O
O
NN
S
O
O
NN S
O
O
NN
NH2
S
O
O
NF
F
F
S
O
O
N
S
O
O
N
S
O
O
NNH2
ISOCRATIC STUDY: NEUTRAL SET (II)
Retention time related to structure of compounds
2-ethyl-pyridyl (2-EP) and cyano columns:
tR
tR
On 2-EP phase: co-elution of 2 compounds, despite distinct physico-chemical properties.
SFC User Meeting - Cambridge - 2006
8
S
O
O
N
F
F
F S
O
O
N
S
O
O
N
S
O
O
NNH2
S
O
O
NN
F
FF
S
O
O
NN
S
O
O
NN S
O
O
NN
NH2
S
O
O
NF
F
F
S
O
O
N
S
O
O
N
S
O
O
NNH2
Different trend observed on diol and silica phases: benzimidazole derivatives eluted the latest
On bare silica column: compounds 3 and 9 co-eluting
tR
tR
ISOCRATIC STUDY: NEUTRAL SET (III)
SFC User Meeting - Cambridge - 2006
9
COLUMN COMPARISON
Compounds not more retained by a particular column
Same interactions between compounds and all four columns ?
Retention mechanism mainly due to free silanol groups ?
Method: isocratic 20% MeOH, 4mL/min, 35°C, 100 bars
Retention time vs compound number
Compound Number0 2 4 6 8 10 12
Re
ten
tion
Tim
e (
min
)
0
1
2
3
4
2-EP 60A 6u 4.6x250mmCyano 60A 6u 4.6x250mmDIol 60A 6u 4.6x250mmSilica 60A 6u 4.6x250mm
SFC User Meeting - Cambridge - 2006
10
Screening of the library using isocratic elution
Basic set studied on:
- 2-EP column 4.6x250mm- cyano column 4.6x250mm- diol column 4.6x250mm
Chromatographic Conditions:
- modifier: MeOH + 0.1% v/v DEA- mobile phase: 20% v/v modifier in CO2 - flow: 4mL.min-1
- temperature: 35°C- outlet pressure: 100 bars
- injection volume: 4mL (~0.5mg.mL-1)
ISOCRATIC STUDY: BASIC SET (I)
SFC User Meeting - Cambridge - 2006
11
Retention time related to structure of compounds
BUT: only with regards to sulfonyl part
Same trend as observed for neutrals:
tR
ISOCRATIC STUDY: BASIC SET (II)
F
F
F
S
O
O
N
R2
R1S
O
O
N
R2
R1S
O
O
N
R2
R1S
O
O
N
R2
R1
NH2
Diol columns: 4 compounds not eluted:
S
O
O
NNH2
NEt2
NEt2
31S
O
O
N
NEt2
NEt2
16 S
O
O
N
NEt2
NEt2
26 S
O
O
N
N
NH2
32
SFC User Meeting - Cambridge - 2006
12
COLUMN COMPARISON
More differences than observed with neutral set
However retention close on all columns for most of the compounds
Retention mechanism mainly due to free silanol groups ?
Isocratic elution, 20% MeOH+0.1%DEA, 4mL/min, 35°C, 100 bars
Retention time versus basic compounds number
Compound Number12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
t R (
min
)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
2-ethyl-pyridyl 60A 6u 4.6x250mm cyano 60A 6u 4.6x250mm diol 60A 6u 4.6x250mm
SFC User Meeting - Cambridge - 2006
13
LIMITS OF THE ISOCRATIC APPROACH
No simple correlation highlighted between physico-chemical properties and tR on the silica column
Is the isocratic approach right?
Indeed: methanol composition (20%) chosen arbitrarily
So, why 20% and not 30% ?
Retention time: the most obvious parameter, but is it the most appropriate ?
SFC User Meeting - Cambridge - 2006
14
0
0
ttt
k R
More than the retention time, the capacity ratio k is a convenient way to normalize the retention of an analyte:
Most successful models of retention developed by Snyder1
and Soczewinski2, in reverse phase systems like ODS/H2O+MeOH:
n log log wkk
In normal phase system, strong deviation observed, but linear correlation can be found in log-log coordinate system:
mlogconst log k
1. Snyder, L. R., Principles of Adsorption chromatography. First ed.; M. Dekker: New York, 19682. Soczewinski, E., Mechanistic molecular model of liquid-solid chromatography - Retention-eluent composition
relationships. Journal of Chromatography A 2002, 965, (1-2), 109-116
RETENTION MODEL IN CHROMATOGRAPHY (I)
SFC User Meeting - Cambridge - 2006
15
% Modifier vs. log k
% Modifier
0 5 10 15 20 25
log k
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
log kw
Regression analysis: calculation of intercept log kw and slope
log kw: standardised parameter more reliable than arbitrary isocratic log k
Start of “polycratic” study, aim: calculation of intercept “log k0” and slope S
RETENTION MODEL IN CHROMATOGRAPHY (II)
log k = S x + log kw
SFC User Meeting - Cambridge - 2006
16
Requirement: keep log k within a range of “practical interest” as defined by Schoenmaker et al.1:
0 < log k < 1
Getting the retention factor within the range:
- slow eluting compounds: no difficulties since increasing allows to elute them soon enough
- fast eluting compounds: decreasing doesn’tsystematically allow sufficient increase in retentionto comply with requirement
Consequence: some compounds to be removed from study
Analyses carried out three times, plotting of mean values
« POLYCRATIC » STUDY
1. Schoenmaker et al. Journal of Chromatography 1979, 185, 179-195
SFC User Meeting - Cambridge - 2006
17
2-EP COLUMN
10 compounds within k range
plot vs. log k: R2 > 0.98
log k0 and S calculated for the 10 compounds
S
O
O
N
F
F
F S
O
O
N
S
O
O
N
S
O
O
NNH2
S
O
O
NN
F
FF
S
O
O
NN
S
O
O
NN S
O
O
NN
NH2
S
O
O
NF
F
F
S
O
O
N
S
O
O
N
S
O
O
NNH2
SFC User Meeting - Cambridge - 2006
18
7 compounds within k range
plot vs. log k: R2 > 0.99
log k0 and S calculated for the 7 compounds
S
O
O
N
F
F
F S
O
O
N
S
O
O
N
S
O
O
NNH2
S
O
O
NN
F
FF
S
O
O
NN
S
O
O
NN S
O
O
NN
NH2
S
O
O
NF
F
F
S
O
O
N
S
O
O
N
S
O
O
NNH2
SILICA COLUMN
SFC User Meeting - Cambridge - 2006
19
11 compounds within k range
plot vs. log k: R2 > 0.99 (except compound 6 at R2=0.92)
log k0 and S calculated for the 11 compounds
S
O
O
N
F
F
F S
O
O
N
S
O
O
N
S
O
O
NNH2
S
O
O
NN
F
FF
S
O
O
NN
S
O
O
NN S
O
O
NN
NH2
S
O
O
NF
F
F
S
O
O
N
S
O
O
N
S
O
O
NNH2
CYANO COLUMN
SFC User Meeting - Cambridge - 2006
20
8 compounds within k range
plot vs. log k: R2 > 0.97
log k0 and S calculated for the 8 compounds
S
O
O
N
F
F
F S
O
O
N
S
O
O
N
S
O
O
NNH2
S
O
O
NN
F
FF
S
O
O
NN
S
O
O
NN S
O
O
NN
NH2
S
O
O
NF
F
F
S
O
O
N
S
O
O
N
S
O
O
NNH2
DIOL COLUMN
SFC User Meeting - Cambridge - 2006
21
CORRELATING DATA WITH COMPOUNDS PROPERTIES (I)
On none of the columns, log k0 and S could be correlated with lipophilicity constants of the analytes (log P [ACD and Spartan], log D at pH 7.4) nor with pKa values.
However, other properties of the neutral set compounds calculated with Spartan, e.g.:
- total dipole moment - surface area A- volume V- electronic charges on single atoms
Multiple regression analysis performed to correlate properties with retention characteristics log k0 and S
SFC User Meeting - Cambridge - 2006
22
On 2-EP column:
• when regressing log k0 vs. , A and min:
log k0 = 0.232 + 0.001 A – 1.325 min – 2.200 R=0.95
• when regressing S vs. , A and min:
S = -0.003 - 1.6x10-5 A + 0.27 min – 0.023 R=0.90
0.000
0.200
0.400
0.600
0.800
1.000
1.200
1.400
1.600
0.000 0.200 0.400 0.600 0.800 1.000 1.200 1.400 1.600
exp log k0 vs predicted log k0
R=0.95
-0.040
-0.035
-0.030
-0.025
-0.020
-0.015
-0.010
-0.005
0.000
-0.0400 -0.0350 -0.0300 -0.0250 -0.0200 -0.0150 -0.0100 -0.0050 0.0000
exp S vs predicted S
R=0.90
CORRELATING DATA WITH COMPOUNDS PROPERTIES (II)
SFC User Meeting - Cambridge - 2006
23
0.000
0.200
0.400
0.600
0.800
1.000
1.200
1.400
0.000 0.200 0.400 0.600 0.800 1.000 1.200 1.400 -0.080
-0.070
-0.060
-0.050
-0.040
-0.030
-0.020
-0.010
0.000
-0.100 -0.050 0.000 0.050 0.100 0.150 0.200
On cyano column:
• when regressing log k0 vs. , A and min:
log k0 = 0.063 + 0.001 A – 1.579 min – 1.25 R=0.94
• when regressing S vs. , A and min:
S = -0.011 - 2.2x10-5 A + 0.071 min – 0.040 R=0.84exp log k0 vs predicted log k0
R=0.93
exp S vs predicted S
R=0.80
CORRELATING DATA WITH COMPOUNDS PROPERTIES (III)
SFC User Meeting - Cambridge - 2006
24
-0.035
-0.030
-0.025
-0.020
-0.015
-0.010
-0.005
0.000-0.035 -0.030 -0.025 -0.020 -0.015 -0.010 -0.005 0.000
0.000
0.200
0.400
0.600
0.800
1.000
1.200
1.400
0.000 0.200 0.400 0.600 0.800 1.000 1.200 1.400
On diol column:
• when regressing log k0 vs. , A and min:
log k0 = 0.124 + 0.0002 A – 1.409 min – 1.173 R=0.97
• when regressing S vs. , A and min:
S = -0.0004 - 1.1x10-5 A + 0.030 min – 0.008 R=0.95exp log k0 vs predicted log k0
R=0.97
exp S vs predicted S
R=0.95
CORRELATING DATA WITH COMPOUNDS PROPERTIES (IV)
SFC User Meeting - Cambridge - 2006
25
0.000
0.200
0.400
0.600
0.800
1.000
1.200
1.400
1.600
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 -0.040
-0.035
-0.030
-0.025
-0.020
-0.015
-0.010
-0.005
0.000
-0.04 -0.035 -0.03 -0.025 -0.02 -0.015 -0.01 -0.005 0
On silica column:
• when regressing log k0 vs. , A and min:
log k0 = 0.119 - 0.001 A – 1.714 min – 1.178 R=0.99
• when regressing S vs. , A and min:
S = -0.0008 - 7.7x10-5 A + 0.040min – 0.012 R=0.84exp log k0 vs predicted log k0
R=0.99
exp S vs predicted S
R=0.84
CORRELATING DATA WITH COMPOUNDS PROPERTIES (V)
SFC User Meeting - Cambridge - 2006
26
Isocratic approach: highlighting of trends in retention
Polycratic study:
CONCLUSION (I)
- supposed better characterization of the retention
- no correlation of log k0 or S with lipophilicity constants or pKa
- study of 12 neutral sulfonamides on 2-EP, cyano, diol and bare silica stationary phases
- linearity found for log k = log k0 + S x ϕ R2>0.97
SFC User Meeting - Cambridge - 2006
27
CONCLUSION (II)
Correlating retention characteristics and molecular descriptors
- Retention characteristics log k0 and S correlated with total dipole moment μ, surface area A and atomic charge on the most negatively charged atom δmin
- Encouraging results for log k0 which are correlated with molecular descriptors with R > 0.94 on all four stationary phases
- Less promising results for S, especially on cyano and silica columns (R = 0.84).
- Results to be considered with care, since obtained on small set of structurally similar compounds.
SFC User Meeting - Cambridge - 2006
28
FUTURE WORK
Analysis of a set of 20 basic sulfonamides currently under way
Study to be extended to compounds of different structures
Recalculation of molecular descriptors by more powerful algorithms under way
Regression analysis using other descriptors has to be undertaken in an attempt to highlight better correlation (experimental log P to be measured)
SFC User Meeting - Cambridge - 2006
29
ACKNOWLEDGEMENTS