DoE in Organic Chem Slides PDF
DoE in Organic Chem Slides PDF
DoE in Organic Chem Slides PDF
Experiments Applied to
Organic Synthesis
Luis Sanchez
Michigan State University
October 11th, 2006
• Statistical Design of Experiments
Strategy
• Appropriate statistical analysis before any
DoE experimental data are obtained
Objective
• To get as much information as possible
from a minimum number of experiments
Bayne, C. K.; Rubin, I. B., Practical experimental designs and optimization methods for chemists. VCH
Publishers, USA, 1986.
Tranter, R., Design and analysis in chemical research. Sheffield Academic; CRC Press: Sheffield, England, 2000.
• Experimentation in Organic synthesis
In any synthetical procedure there are factors
temperature, time, pressure, reagents, rate of
addition, catalyst, solvent, concentration, pH
that will have an influence on the result
yield, purity, selectivity
Carlson, R., Design and optimization in organic synthesis. Elsevier: Amsterdam ; New York, 1992.
• Conventional approach to optimization
T °C
X+Y Z
t minutes
Yield vs. Temperature (t=130 min) Yield vs. Reaction time (T=125°C)
80 80
75 75
Yield (%)
Yield (%)
70 70
65 65
60 60
55 55
105 115 125 135 145 155 40 70 100 130 160 190
Temperature (°C) Time (min)
Tranter, R., Design and analysis in chemical research. Sheffield Academic; CRC Press: Sheffield, England, 2000.
• How yield actually behaves
actual
Yield vs (Time and Temperature)
maximum
155
Temperature (°C)
145
94 local
maximum
135 90 maximum
yield
80
125 70
60
115
105
55 80 105 130 155 180
“response surface” Time (min)
Carlson, R., Design and optimization in organic synthesis. Elsevier: Amsterdam ; New York, 1992.
Tranter, R., Design and analysis in chemical research. Sheffield Academic; CRC Press: Sheffield, England, 2000.
• The conventional approach
• Analysis of the effect of one particular reaction condition
by keeping all the other ones constant
Amount of
catalyst Catalyst
A+ B C
T °C
t minutes
Temperature
Concentration
of substrate
The problem:
• The optimum conditions obtained depend on the starting point
Owen, M. R.; Luscombe, C.; Lai, L. W.; Godbert, S.; Crookes, D. L.; Emiabata-Smith, D.
Org. Proc. Res. Dev. 2001, 5, 308-323.
• The DoE approach
• To rationally choose points throughout the cube to fully
represent the entire space.
Amount of
catalyst Catalyst
A+ B C
T °C
t minutes
Temperature
Concentration
of substrate
Owen, M. R.; Luscombe, C.; Lai, L. W.; Godbert, S.; Crookes, D. L.; Emiabata-Smith, D.
Org. Proc. Res. Dev. 2001, 5, 308-323.
• Outline
Determining important reaction conditions
• Fractional factorial design
number of
number of conditions
values
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• 23 factorial design
OH COOR
COOR T °C
ROOC ROOC
H 2O COOR
COOR
acid catalyst
(H 2SO 4/H3 PO4 )
• 8 experimental runs:
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
run T C K label yield (%)
1 - - - 1 60
2 + - - t 72
3 - + - c 54
4 + + - tc 68
5 - - + k 52
6 + - + tk 83
7 - + + ck 45
8 + + + tck 80
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Measuring the effect: Temperature
run T C K label yield (%)
Effect of T
1 - - - 1 60
12 (-1,1,1) (1,1,1)
2 + - - t 72
(-1,1,-1) (1,1,-1)
3 - + - c 54
14
4 + + - tc 68 (-1,-1,1) (1,-1,1)
(-1,-1,-1) (1,-1,-1)
5 - - + k 52
31
6 + - + tk 83
One half of the average of
7 - + + ck 45 the differences of each pair
35
8 + + + tck 80
⎡ ( t − 1) + ( tc − c ) + ( tk − k ) + ( tck − ck ) ⎤ ⎡12 + 14 + 31 + 35 ⎤
⎢ ⎥ ⎢⎣ ⎥
= ⎣ 4 ⎦ =
4 ⎦ = 11.5
2 2
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Measuring the effect: Concentration
run T C K label yield (%)
Effect of C
1 - - - 1 60
-6
2 + - - t 72
3 - + - c 54 -4
4 + + - tc 68
5 - - + k 52
-7
6 + - + tk 83
One half of the average of
7 - + + ck 45 -3 the differences of each pair
8 + + + tck 80
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Measuring the effect: Catalyst
run T C K label yield (%)
Effect of K
1 - - - 1 60
-8
2 + - - t 72
11
3 - + - c 54
-9
4 + + - tc 68
5 - - + k 52
6 + - + tk 83 12
One half of the average of
7 - + + ck 45 the differences of each pair
8 + + + tck 80
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Concentration-temperature interaction
run T C K label yield (%)
Effect of C on the
1 - - - 1 60 effect of T
12 6
2 + - - t 72
1
3 - + - c 54
14 7
4 + + - tc 68
5 - - + k 52
31 15.5
6 + - + tk 83
2 One half of the average
7 - + + ck 45
35 17.5 of the differences of
8 + + + tck 80 each pair of effects
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Temperature-concentration interaction
run T C K label yield (%)
Effect of T on the
1 - - - 1 60 effect of C
-6 -3
2 + - - t 72
1
3 - + - c 54
-4 -2
4 + + - tc 68
5 - - + k 52
-7 -3.5
6 + - + tk 83
2 One half of the average
7 - + + ck 45
-3 -1.5 of the differences of
8 + + + tck 80 each pair of effects
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Concentration-temperature interaction
run T C K label yield (%)
Effect of C on the
1 - - - 1 60 effect of T
12 6
2 + - - t 72
1
3 - + - c 54
14 7
4 + + - tc 68
5 - - + k 52
31 15.5
6 + - + tk 83
2 One half of the average
7 - + + ck 45
35 17.5 of the differences of
8 + + + tck 80 each pair of effects
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Temperature-catalyst interaction
run T C K label yield (%)
1 - - - 1 60
12 6
2 + - - t 72
9.5
3 - + - c 54
14 7
4 + + - tc 68
5 - - + k 52 10.5
31 15.5
6 + - + tk 83
One half of the average
7 - + + ck 45
35 17.5 of the differences of
8 + + + tck 80 each pair of effects
⎧⎡ ( tk − k ) ( t − 1) ⎤ ⎡ ( tck − ck ) ( tc − c ) ⎤ ⎫ ⎧⎡ 31 12 ⎤ ⎡ 35 14 ⎤ ⎫
⎨⎢ − + − ⎬ 2 ⎨⎢ − ⎥ + ⎢ − ⎥ ⎬ 2
⎩ ⎣ 2 2 ⎥⎦ ⎢⎣ 2 2 ⎥⎦ ⎭ ⎩⎣ 2 2⎦ ⎣2 2 ⎦⎭
on = = =5
2 2
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• TCK interaction
run T C K label yield (%)
1 - - - 1 60
12 6
2 + - - t 72
9.5 4.75
3 - + - c 54
14 7 0.5
4 + + - tc 68
5 - - + k 52 10.5 5.25
31 15.5
6 + - + tk 83
7 - + + ck 45
35 17.5
8 + + + tck 80
⎧⎡ ( tck − ck ) ( tc − c ) ⎤ ⎡ ( tk − k ) ( t − 1) ⎤ ⎫ ⎧⎡ 35 14 ⎤ ⎡ 31 12 ⎤ ⎫
⎨⎢ − − − ⎬ 2 ⎨⎢ − ⎥ − ⎢ − ⎥ ⎬ 2
⎩ ⎣ 2 2 ⎥⎦ ⎢⎣ 2 2 ⎥⎦ ⎭ ⎩⎣ 2 2⎦ ⎣2 2 ⎦⎭
on = = = 0.25
2 2
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Measuring the effect and interactions
Yates’s algorithm: works for any 2n factorial design
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• What do those numbers mean?
• First we need to evaluate if they are significant
Effect
6
4
2
0
-2
3x -4
(when there is T C TC K TK CK TCK
no central point) Factor
T C K
Temperature Concentration Catalyst
(°C) (M)
120 160 1.5 2.5 H3PO4 H2SO4
-1 +1 -1 +1 -1 +1
OH COOR
COOR H 2SO 4(aq)
ROOC ROOC
heat COOR
COOR
number of “ignored”
reaction conditions
Tranter, R., Design and analysis in chemical research. Sheffield Academic; CRC Press: Sheffield, England, 2000.
• Effects vs. interactions
• This is what we
Important?
got before:
Tranter, R., Design and analysis in chemical research. Sheffield Academic; CRC Press: Sheffield, England, 2000
• 24-1 Fractional factorial design
Yates’s algorithm:
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Fractional factorial designs
6
4
2
0
-2
-4
T C TC K TK CK TCK
Factor
% error
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Application example
O
O Br
O
PivO OPiv
O NO 2
NO2 OPiv H
H O O N
HO N 2 O N
N
N CF3
N CF3 Ag 2O PivO OPiv
1 mol.sieves
OPiv
18h 3
(Koenigs-Knorr 3%
glucuronidation)
N
TMEDA =
N
Stazi, F.; Palmisano, G.; Turconi, M.; Clini, S.; Santagostino, M. J. Org. Chem. 2004, 69, 1097-1103.
• Application example
O
O Br
O
PivO OPiv
O NO 2
NO2 OPiv H
H O O N
HO N 2 O N
N
N CF3
N CF3 Ag 2O PivO OPiv
1 OPiv
10 equiv. TMEDA 3
mol.sieves 27 %
18h
HMTTA = N
N
N
• Application example
O
O Br
O
PivO OPiv
O NO 2
NO2 OPiv H
H O O N
HO N 2 O N
N
N CF3
N CF3 Ag2 CO3 PivO OPiv
1 OPiv
10 equiv. HMTTA 3
mol.sieves 42 %
18h
Reaction condition -1 +1
A pre-complex time (min) 0 60
B reaction time (h) 2 6
C Ag2CO3 (equiv) 1.5 3.8
D HMTTA (equiv) 1.5 12.6
E sugar derivative (equiv) 1.5 3
F 4 Å mol sieves (mg) 0 100
G solvent (mL) 0.5 1.5
• Application example
A pre-complex time (min)
B reaction time (h)
• 27-4 factorial design results: C Ag2CO3 (equiv)
D HMTTA (equiv)
E sugar derivative (equiv)
F 4 Å mol sieves (mg)
1 - - - + + + - 14.7
2 + - - - - + + 19.5
3 - + - - + - + 24.4
4 + + - + - - - 11.2
5 - - + + - - + 34.2
6 + - + - + - - 83.2
7 - + + - - + - 56.5
8 + + + + + + + 55.4
Stazi, F.; Palmisano, G.; Turconi, M.; Clini, S.; Santagostino, M. J. Org. Chem. 2004, 69, 1097-1103.
• Application example
O
O Br
O
Stazi, F.; Palmisano, G.; Turconi, M.; Clini, S.; Santagostino, M. J. Org. Chem. 2004, 69, 1097-1103.
• Outline
Determining important reaction conditions
• Fractional factorial design
yield yield
Carlson, R., Design and optimization in organic synthesis. Elsevier: Amsterdam; New York, 1992.
• Response surface analysis
OH COOR
COOR H2 SO4(aq) 1.0M
ROOC ROOC
T °C
COOR COOR
t min
run t T
t T 1 - -
time Temperature 2 + -
(min) (°C)
3 - +
70 80 127.5 132.5 4 + +
-1 +1 -1 +1 5 0 0 Central point:
6 0 0 three times to
calculate the
7 0 0 experimental error
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Response surface analysis
OH COOR
COOR H2 SO4(aq) 1.0M
ROOC ROOC
T °C
COOR COOR
t min
Temperature (°C)
2 + - 60.3 64.6 68.0
133
3 - + 64.6
131 62.3
4 + + 68.0
5 0 0 60.3 129
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Response surface analysis
Yield vs. (Time and Temperature)
160
58.2
150
Temperature (°C)
87.4
140
69.1
130
120
60 70 80 90 100 110
time (min)
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Response surface analysis
Yield vs. (Time and Temperature) • Equation for the 22 factorial
design:
155
91.1
yield = 82 . 09 − 2 . 69 t + 6 . 97 T ±
91.9 85.9
150
87.4
surface:
145 86.8 79.3
yield = 87 . 36 − 2 . 69 t + 6 . 97 T
140 − 2 .15 t 2 − 3 .12 T 2 − 0 . 58 Tt ±
77.2 73.01
71.2
135
70 80 90 100 110
time (min)
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Response surface analysis
yield = 87 . 36 − 2 . 69 t + 6 . 97 T − 2 . 15 t − 3 .12 T − 0 . 58 Tt ±
2 2
160
155 93
Temperature (°C) 90 85
150
88
Optimum conditions:
T = 157 °C 145
t = 73 min
140
yield: 93%
70 80 90 100 110
time (min)
Box, G. E. P.; Hunter, W. G.; Hunter, J. S., Statistics for experimenters : an introduction to design, data analysis,
and model building. Wiley: New York, 1978.
• Sequential nature of experimentation
Plan
Response
surface analysis
Tranter, R., Design and analysis in chemical research. Sheffield Academic; CRC Press: Sheffield, England, 2000.
• Application of response surface analysis
TBSO H R'' HO H R'' H
H TEA•3HF H R H
O O
+
R N R N
O NMP O O N
O
O R' O R' O R'
O O O
1 2 3
24 central composite
reaction condition range units
temperature 10 30 °C
time 19 31 hours
volume of NMP 3 7 mL/g of substrate
equivalents of TEA.3HF 1 1.67 Equivalents
Monitored results:
• % yield of alcohol
• % lactone
• % remaining silyl ether
Owen, M. R.; Luscombe, C.; Lai, L. W.; Godbert, S.; Crookes, D. L.; Emiabata-Smith, D.
Org. Proc. Res. Dev. 2001, 5, 308-323.
• Application of response surface analysis
TBSO H R'' HO H R'' H
H TEA•3HF H R H
O O
+
R N R N
O NMP O O N
O
O R' O R' O R'
O O O
1 2 3
Owen, M. R.; Luscombe, C.; Lai, L. W.; Godbert, S.; Crookes, D. L.; Emiabata-Smith, D.
Org. Proc. Res. Dev. 2001, 5, 308-323.
• Application
TBSO H R'' HO H R'' H
H TEA•3HF H R H
O O
+
R N R N
O NMP O O N
O
O R' O R' O R'
O O O
1 2 3
target/constraints T (°C) Time (h) solvent Et3N·3HF predicted actual predicted actual
lactone < 2%, solvent < 3.5 mL/g 14 31 3.45 1.58 93.9 94.2 1.8 2.0
lactone < 2% Et3N.3HF < 1.18eq. 28 19.5 7 1.17 93.7 93.4 1.9 2.0
lactone < 2%, time < 23 h 24 23 6.3 1.41 94.2 94.2 2.0 1.9
Owen, M. R.; Luscombe, C.; Lai, L. W.; Godbert, S.; Crookes, D. L.; Emiabata-Smith, D.
Org. Proc. Res. Dev. 2001, 5, 308-323.
• When DoE “fails”
O O
N N
H O H O
1) AcBr, Ac 2 O, CH 2Cl2
H
2) KOH, MeOH
H 3) HCl, CH 2Cl2 H
O HO
1 2
Larkin, J. P.; Wehrey, C.; Boffelli, P.; Lagraulet, H.; Lemaitre, G.; Nedelec, A.
Org. Proc. Res. Dev. 2002, 6, 20-27.
• Outline
Determining important reaction conditions
• Fractional factorial design
Recent advances
• Software
• Automation
• “DoE involves a lot of math, it’s rather
complicated”
Lendrem, D.; Owen, M.; Godbert, S. Org. Proc. Res. Dev. 2001, 5, 324-327.
• Software
Most commonly used:
• Umetrics MODDE ®
(http://www.umetrics.com)
On-line HPLC
Harre, M.; Tilstam, U.; Weinmann, H. Org. Proc. Res. Dev. 1999, 3, 304-318.
• Example of the use of automation
OH
Ar
• System: HO O
PPh3 , DIAD Ar
Automated liquid handler
R toluene R
On-line HPLC 50 - 70 %
• Reaction conditions:
A equivalents of alcohol
B equivalents of DIAD
C volume of toluene
D temperature
E addition rate of DIAD
• 20 experimental runs
Important factors: ratio DIAD/alcohol,
• Total research time: 5 days alcohol, temperature
Emiabata-Smith, D. F.; Crookes, D. L.; Owen, M. R. Org. Proc. Res. Dev. 1999, 3, 281-288.
• Why DoE methods are ideal for us
Further exploration
would lead us to
obtain > 94% yield
Emiabata-Smith, D. F.; Crookes, D. L.; Owen, M. R. Org. Proc. Res. Dev. 1999, 3, 281-288.
• Some final comments
• DoE offers powerful mathematical models that are
applicable to the behavior of organic reactions
Lendrem, D.; Owen, M.; Godbert, S. Org. Proc. Res. Dev. 2001, 5, 324-327.
• Acknowledgements
Prof. Maleczka
Prof. Walker