Examined variables in BO-based optimization of reaction conditions.
 | |||
|---|---|---|---|
| pKaH = P (Base) | T (°C) | Conc. = V (M) | Z/W (mL/min) |
| 5.20 (Pyridine) | 0.05 | 2.20 (Z2.2:W1.0) | |
| 7.00 (NMI) | −20 | 1.67 (Z2.2:W1.2) | |
| 7.70 (N-Ethylmorpholine) | −10 | 0.10 | 1.29 (Z1.8:W1.4) |
| 8.90 (Me2NBn) | 0 | 1.00 (Z1.6:W1.6) | |
| 9.50 (Et2NBn) | 10 | 0.15 | 0.78 (Z1.4:W1.8) |
| 9.70 (DMAP) | 20 | 0.60 (Z1.2:W2.0) | |
| 11.4 (DIEA) | 30 | 0.20 | 0.45 (Z1.0:W2.2) |
| 13.0 (DBU) | 40 | 0.25 | |
| |||
|---|---|---|---|
| pKaH = P (Base) | T (°C) | Conc. = V (M) | Z/W (mL/min) |
| 5.20 (Pyridine) | 0.05 | 2.20 (Z2.2:W1.0) | |
| 7.00 (NMI) | −20 | 1.67 (Z2.2:W1.2) | |
| 7.70 (N-Ethylmorpholine) | −10 | 0.10 | 1.29 (Z1.8:W1.4) |
| 8.90 (Me2NBn) | 0 | 1.00 (Z1.6:W1.6) | |
| 9.50 (Et2NBn) | 10 | 0.15 | 0.78 (Z1.4:W1.8) |
| 9.70 (DMAP) | 20 | 0.60 (Z1.2:W2.0) | |
| 11.4 (DIEA) | 30 | 0.20 | 0.45 (Z1.0:W2.2) |
| 13.0 (DBU) | 40 | 0.25 | |
The employed BO scheme uses Gaussian process (GP) models as surrogates and includes (i) initialization through 5 experiments with combinations of reaction conditions randomly defined using Latin hypercube sampling (LHS). (ii) Training of the GP model using the results of 5 experiments. (iii) Identifying the subsequent reaction conditions by maximizing the upper confidence bound (UCB) acquisition function determined by the GP model from step 2). (iv) Performing new experiments under the identified conditions. (v) Repeating steps (ii)–(iv) until the yield exceeds 90%.
Examined variables in BO-based optimization of reaction conditions.
| |||
|---|---|---|---|
| pKaH = P (Base) | T (°C) | Conc. = V (M) | Z/W (mL/min) |
| 5.20 (Pyridine) | 0.05 | 2.20 (Z2.2:W1.0) | |
| 7.00 (NMI) | −20 | 1.67 (Z2.2:W1.2) | |
| 7.70 (N-Ethylmorpholine) | −10 | 0.10 | 1.29 (Z1.8:W1.4) |
| 8.90 (Me2NBn) | 0 | 1.00 (Z1.6:W1.6) | |
| 9.50 (Et2NBn) | 10 | 0.15 | 0.78 (Z1.4:W1.8) |
| 9.70 (DMAP) | 20 | 0.60 (Z1.2:W2.0) | |
| 11.4 (DIEA) | 30 | 0.20 | 0.45 (Z1.0:W2.2) |
| 13.0 (DBU) | 40 | 0.25 | |
| |||
|---|---|---|---|
| pKaH = P (Base) | T (°C) | Conc. = V (M) | Z/W (mL/min) |
| 5.20 (Pyridine) | 0.05 | 2.20 (Z2.2:W1.0) | |
| 7.00 (NMI) | −20 | 1.67 (Z2.2:W1.2) | |
| 7.70 (N-Ethylmorpholine) | −10 | 0.10 | 1.29 (Z1.8:W1.4) |
| 8.90 (Me2NBn) | 0 | 1.00 (Z1.6:W1.6) | |
| 9.50 (Et2NBn) | 10 | 0.15 | 0.78 (Z1.4:W1.8) |
| 9.70 (DMAP) | 20 | 0.60 (Z1.2:W2.0) | |
| 11.4 (DIEA) | 30 | 0.20 | 0.45 (Z1.0:W2.2) |
| 13.0 (DBU) | 40 | 0.25 | |
The employed BO scheme uses Gaussian process (GP) models as surrogates and includes (i) initialization through 5 experiments with combinations of reaction conditions randomly defined using Latin hypercube sampling (LHS). (ii) Training of the GP model using the results of 5 experiments. (iii) Identifying the subsequent reaction conditions by maximizing the upper confidence bound (UCB) acquisition function determined by the GP model from step 2). (iv) Performing new experiments under the identified conditions. (v) Repeating steps (ii)–(iv) until the yield exceeds 90%.
