In Chapter 5 “Sequential Cross Coupling Reactions of Polyhalogenated Heterocycles”, we discussed the use of QM calculated LUMO/LUMO map and IR carbon-halogen bond stretching wavenumbers to predict the order of haloselectivity for a sequence of palladium-catalyzed cross coupling reactions on monocyclic 2,4-dichloro-5-bromopyrimidine. We left a related question on predicting haloselectivity of palladium catalyzed cross coupling reaction for the bicyclic 3-bromo-5,7-dichloropyrazolopyrimidine (1). What will be the order of haloselectivity (Figure 1)?
Figure 1. Suzuki Cross Coupling reaction of compound 1
Shown in Figure 2 are the calculated LUMO and IR stretching vibration wavenumbers of the three carbon-halogen bonds of compound 1.
Figure 2. LUMO of compound 1 and IR stretching vibration wavenumbers of the C-X bonds
There is little LUMO lobe on the C3-Br carbon relative to the C5-Cl and C7-Cl carbons on compound 1. IR calculation shows that the stretching vibration wavenumbers of the C7-Cl and C5-Cl bonds are 1098 and 1143 cm-1, respectively, suggesting that the C7-Cl bond is weaker and will be more reactive toward Pd(0) oxidative addition. Our experimental result is consistent with the above prospective analysis, Suzuki cross coupling reaction of compound 1 with potassium methyl-trifluoroborate provides compound 2 selectively . An obvious follow-up question is haloselectivity of intermediate 2.
Figure 3. Suzuki Cross Coupling reaction of intermediate 1
With intermediate 2, there is little LUMO lobe on the C3-Br carbon relative to the C5-Cl one. IR calculation shows that the stretching vibration wavenumber of the C3-Br and C5-Cl bonds are 1168 and 1147 cm-1, respectively, suggesting that the C5-Cl bond is more reactive toward oxidative addition. Both LUMO and IR analyses are suggestive of selective reaction at C5-Cl, consistent with experimental results. Here we can see that the haloselectivity of metal-catalyzed coupling reactions of compounds 1 and 2 diverges from the general reactivity we learned from general organic chemistry for C-X bonds, i.e., X = I > Br ∼ OTf >> Cl >> F, yet fully accountable with relevant QM calculated parameters.
Figure 4. LUMO of intermediate 2 and IR stretching vibration wavenumbers of the C-X bonds
From the previous example, we learned that for certain bicyclic heterocyclic systems, C-Cl bonds could undergo oxidative addition with Pd(0) selectively over C-Br bonds. What have we learned about iodo containing polyhalogenated bicyclic systems? Below are two examples.
Figure 5. Buchwald reaction of aromatic heterocyclic compound 4
Shown in Figure 5 is 4-chloro-7-iodopyrrolotriazine 4, prepared by NIS iodination of commercially available chloropyrrolotriazine 3. Intuitively, we anticipated the C7-iodide bond to be more susceptible to oxidative addition to give selectively a C-7 amination product. However, only C-4 amination product was obtained under various Buchwald amination conditions. With compound 4, there is little LUMO lobe on the C7-I carbon relative to the C4-Cl one. IR calculation shows that the stretching vibration wavenumber of the C7-I and C4-Cl bonds are 873 and 823 cm-1, respectively, suggesting that the C4-Cl bond to be more reactive toward oxidative addition. Both LUMO and IR analyses are suggestive of selective reaction at C4-Cl, consistent with experimental observations. We learned to integrate these analyses to our retrosynthetic planning.
Figure 6. LUMO of compound 4 and IR stretching vibration wavenumbers of the C-X bonds
Second example comes from an OPRD paper published by Pfizer (Figure 7). For nucleophilic aromatic substitution of 2,4-dichloro-5-iodopyrrolopyrimidine 5 with a secondary alcohol compound, we also calculated for its LUMO and C-X bond stretching wavenumbers. As shown in Figure 8, there are LUMO lobes on all three carbon-halogen bonds, yet with the lobe on C4-Cl carbon significantly larger than the ones on C5-I and C2-Cl carbons. IR calculation shows that the stretching vibration wavenumber of the C2-Cl, C4-Cl, and C5-I bonds are 981, 837, and 924 cm-1, respectively, suggesting that the C4-Cl bond is weakest and most reactive toward nucleophilic substitution, accounting for the high regioselectivity reported.
Figure 7. SnAr and Negishi reactions of polyhalogenated pyrrolopyrimidine 5
Figure 8. LUMO of compound 5 and IR stretching vibration wavenumbers of the C-X bonds
With polyhalogenated heterocycles, the C-X bonds’ relative reactivity towards metal-catalyzed coupling reactions does not necessarily follow the general reactivity order of X = I > Br ∼ OTf >> Cl >> F. In the case of bicyclic heteroaromatic systems, if the C-Cl bond is on the electron-deficient part of the system, while the C-I or C-Br bonds are on the electron-rich part, we shall calculate for LUMO and IR C-X bonds’ stretching vibration wavenumbers to sequence properly the order of reactions. These intrinsic changes in halo reactivity are predictable.
How about polyhalogenated tricyclic heteroaromatic substrates? Let’s analyze 5-bromo-9-chloropyridopyrropyrimidine (8). Based on the calculated LUMO and C-X bond stretching wavenumbers shown in Figure 9, will palladium-catalyzed reactions be selective on the C-Br or C-Cl bond? Which factor controls the selectivity observed for this substrate ?
Figure 9. LUMO of compound 8 and IR stretching vibration wavenumbers of the C-X bonds
Zhong Zheng, Qiuyue Wang, Yongsheng Chen, John S. Wai
 Z. Xu, Y. Lou, Fused heterocyclic compound, preparation method therefor, pharmaceutical composition, and uses therefore. US 20160244432 A1, August 25, 2016, page 28 on related Suzuki reaction with Compound 1.
 Y. Tao, N.F. Keene, K.E. Wiglesworth, B. Sitter, J.C. McWilliams, Org. Process Res. Dev. 2019, 23, 382.
 A. Baeza, J. Mendiola, C. Burgos, J. Alvarez-Builla, J. J. Vaquero, Eur. J. Org. Chem. 2010, 29, 5607.