Quadruple C-H activation coupled to hydrofunctionalization and C-H silylation/borylation enabled by weakly coordinated palladium catalyst

Unlike the well-reported 1,2-difunctionalization of alkenes that is directed by classic pyridine and imine-containing directing groups, oxo-palladacycle intermediates featuring weak Pd-O coordination have been less demonstrated in C-H activated cascade transformations. Here we report a quadruple C-H activation cascade as well as hydro-functionalization, C-H silylation/borylation sequence based on weakly coordinated palladium catalyst. The hydroxyl group modulates the intrinsic direction of the Heck reaction, and then acts as an interrupter that biases the reaction away from the classic β-H elimination and toward C-H functionalization. Mechanistically, density functional theory calculation provides important insights into the key six-membered oxo-palladacycle intermediates, and indicates that the β-H elimination is unfavorable both thermodynamically and kinetically. In this article, we focus on the versatility of this approach, which is a strategic expansion of the Heck reaction.

Response: Thanks for your comments. We are really sorry for our inappropriate presentation of the manuscript. Now we have tried our best to revise the presentation in a more scholarly manner. After careful consideration, we move the discussion of reaction condition screening from main article to SI, to simplify the main article. Besides, we invited Prof. Lyle Isaacs (University of Maryland) to polish our language in the paper, to further improve the quality of manuscript. Response: Thank you very much for your earnest and careful advice! We apologize for our scholarly presentation errors. After careful consideration, we hope to use Arabic numerals to label all of the structures, because some variations on a central core are more than 26. Now the numbering of structures have been revised.

IVS
But theoretically, functionalization on the original aryl is quite possibile. Thus we used GC-MS to detect the possible isomer. We take the silylation as the example because of the relatively low boil point of the corresponding compounds. As shown in Scheme R4, when we increased the reaction temperature to 150 o C, we detected a possible trace-amount of isomer /)R (note: this compound can not be seen on TLC, thus we could not separate and verify its exact structure), the ratio was 82 : /)R= 97 : 3. However, in the standard conditions (70 o C, Scheme R5), the ratio was 82 : /)R= 99.1 : 0.9 (the exact structure of 82 was verified by X-ray). This experiment indicates that higher temperature may promote the C-H functionalization on original aryl. However, this process is extremely unfavorable compared with the formation of main product.

:?CAGA 9,&
To further explain this selectivity, two competing pathways have been investigated by DFT calculations (Scheme R6). We found that when complex <R is formed, one phenyl group coming from iodobenzene coordinates onto Pd leading to the discrimination of another one. After a ligand exchange with acetate, a more stable complex <4R can be formed, where the coordination of phenyl group is remained. We found that the dissociation of phenyl group by ligand exchange with another oxygen atom in coordinated acetate need to bear a free energy barrier of 19.1 kcal mol -1 via transition state TS<4R-<44R? (Path A). As a contrast, the activation free energy for the alternative proton abstraction via a six-membered transition state (TS<4R-<44R=) to afford complex <44R= is only 15.4 kcal mol -1 (path B). Therefore, the exchange of two phenyl groups cannot be achieved. We have revised this part in new version of draft and highlighted it.
Finally, according to your advice, the structure <4R and the corresponding energy in the main article, as well as the corresponding data of DFT calculation in SI, have been revised. Furthermore, we added some discussions about this notable selectivity in the main article (in the discussion of Scheme 11).

IVS aWZgZObW]\ O\R P]`gZObW]\ ]\Zg c\RS`U] []\]*Tc\QbW]\OZWhObW]\) eVS`SOa bVS O`gZ Q]\aWabS\bZg c\RS`U]^]ZgTc\QbW]\OZWhObW]\+ LVOb Wa bVS`SOa]\ T]`bVWa R`OabWQ RWTTS`S\QS9
Response: Thanks for your comments. Now we have different understanding of the silylation and borylation process by following your valuable advice (The reaction in Scheme 4 is a hydroarylation followed by a silylation. What is the source of the added hydrogen? Is this the one that was originally on the C-H activated position of the aryl?).
:?CAGA 9.& Take the silylation process as an example, after our isotopic labeling experiments (Scheme R7), we found that the added hydrogen was probably come from water, rather than DMF or ArI. Besides, the 80% deuteration of D 2 W\RWQObS O`SdS`aWPZS t-H elimination might be involved in the catalytic cycle. Based on these findings, we came up with a new mechanism of the silylation process (Scheme R8). If the silylation process undergoes a di-functionalization (path B), then a TMS anion species would be formed, which is unstable and make this process unfavorable (Scheme R9). Therefore, we considered that the non-reacted part of TMS should be combined with Pd(II) untill the reductive elimination of Pd(II), to afford TMSX (X=anion) and Pd(0). Based on this understanding, we proposed that intermediate G undergoes a consecutive protonation to give TMS-Pd(II)-X and final mono-silylation product, and this pathway would be a kinetically favored process.
:?CAGA 9/& 8ILLD>FA 5A?C=HDLG IB :DFQF=MDIH :?CAGA 90& On the other hand, as for poly-functionalization of aryl, after the first o-position functionalization completed, the t-H elimination or protodepalladation process is still unfavorable for both thermodynamic and kinetic aspects (Scheme R10). Comparatively, due to the good leaving ability of iodine anion, the corresponding intermediate would undergoes the second, a very fast another o-position functionalization, which is a kinetically favorable step. Therefore, we considered that the drving force of this domino-type process is very strong.
The added isotopic labeling experiments, corresponding 1 H NMR data and new mechanism of silylation (Page S13-S15 in SI) have been added to the revised SI. Response: Thank you for your comments. There are indeed some gaps of the exact mechanism in dehydrogenation process. Because (1) we could still observed the dehydrogenation product in middle yield without the addition of PhI; (2) we could not detect the corresponding amount of benzene by GC-MS. Despite this, according to the fact that no hydrogen gas was detected during the reaction (SI, Page 7), then the possibility of a oxidative addition process from -O-H to Pd(0) could be excluded. So we considered that the dehydrogenation process was initiated by Pd(II) rather than Pd(0). Therefore, the key issue is to find the real oxidant that oxidize Pd(0) to Pd(II).
In this regard, we think it is reasonable to take the excess PhI as the potential oxidant. The reasons are: (1) In scheme 9e, the yield of dehydrogenation product obviously reduced when we removed the PhI. (2) we do have detected a certain amount of benzene that possibly is a reduction product of PhI.
Actually, PhI is not likely the only oxidant in this process. An exact and complete mechanistic process maybe more complicated than that we considered. But we believe that it is reasonable to consider the excess PhI as a potential oxidant. Finally, according to your advice, we have added some discussions in the main article (reference [ 22] ) about the detection of benzene by GC-MS, as well as the existence of possibility for other plausible oxidative pathways.