Silver-catalyzed remote Csp3-H functionalization of aliphatic alcohols

Aliphatic alcohols are common and bulk chemicals in organic synthesis. The site-selective functionalization of non-activated aliphatic alcohols is attractive but challenging. Herein, we report a silver-catalyzed δ-selective Csp3-H bond functionalization of abundant and inexpensive aliphatic alcohols. Valuable oximonitrile substituted alcohols are easily obtained by using well-designed sulphonyl reagents under simple and mild conditions. This protocol realizes the challenging δ-selective C–C bond formation of simple alkanols.


Synthesis and Characterizations of Sulphonyl Reagents
Reagents A-E were Prepared According to Literature Methods. [9] A solution of NaOEt in EtOH (prepared fresh from 790 mg, 34.33 mmol, 1.2 equiv. of Na and 18 ml of EtOH) was added to the suspension of (Phenylsulfonyl)acetonitrile (98 % Aldrich, 5.29 g, 28.61 mmol, 1 equiv.) in EtOH (7 ml) at RT. To the resulting clear solution isoamyl nitrite (96 % Sigma-Aldrich, 4.8 ml, 34.33 mmol, 1.2 equiv.) was added. The mixture was stirred at RT for 2 hours upon which a yellow solid precipitated. The mixture was cooled in an ice-bath, the solid was filtered and washed with cold EtOH and then with Et2O. Drying under high vacuum afforded the sodium salt of N-hydroxy-1-(phenylsulfonyl)methanimidoyl cyanide as yellow powder (5.893 g, 89 %) that was used directly in the next step. N-(benzyloxy)-1-tosylmethanimidoyl cyanide (B) [10] A solution of NaOEt in EtOH (prepared fresh from 790 mg, 34.33 mmol, 1.2 equiv. of Na and 18 ml of EtOH) was added to the suspension of 2-tosylacetonitrile (5.85 g, 30 mmol, 1 equiv.) in EtOH (7 ml) at RT. To the resulting clear solution isoamyl nitrite (4.8 ml, 34.33 mmol, 1.2 equiv.) was added. The mixture was stirred at RT for 2 hours upon which a yellow solid precipitated. The mixture was cooled in an S9 ice-bath, the solid was filtered and washed with cold EtOH and then with Et2O.
The mixture was stirred at RT for 2 hours upon which a yellow solid precipitated. The mixture was cooled in an ice-bath, the solid was filtered and washed with cold EtOH and then with Et2O. Drying under high vacuum afforded the sodium salt of 1-((4-S10 fluorophenyl)sulfonyl)-N-hydroxymethanimidoyl cyanide as yellow powder (1.12 g, 45 %) that was used directly in the next step.

The Byproduct Analysis:
The results and the identified byproducts in some cases were summarized in Supplementary Table 3. For some substrates such as 1l, 1q, the alcohols were recovered. For some substrates such as 1t, 1u, the main identified byproducts were βscission products.

The Identification of Activation Site
As the alkanols have multiple carbon-hydrogen bonds with similar reactivity, the functionalized position is supposed to be determined. Thus, we designed an experiment to clarify the substitution position of alcohol substrates. With the developed radical decarboxylative strategy, [11] the 2-(3-methoxypropyl)hexanoic acid S1 could transformed to the corresponding oxime ether 8 in 66% yield ( Supplementary Fig. 1a). Meanwhile, the functionalized alcohol 2a modified from 1a in our conditions was reacted with iodomethane to afford S-8 in 14% yield (Supplementary Fig. 1b). We were glad to see that the 1 H NMR of 8 and S-8 are almost identical except for different E/Z isomer ratio of the oxime ether groups (Supplementary Fig. 2). This is a powerful evidence that the functionalized position is the δ carbon-hydrogen bond of the alkanols.

Radical clock experiment:
We synthesized the cyclopropane substrate 1x. Under standard conditions, 1x was consumed. However, we didn't obtain the ring-opening product. In contrast, a mixture product of complex products was detected. We infer this is because the following three reasons: 1) the addition of ring-opening primary alkyl radical to reagent A is difficult. 2) the generated alkene product S2 was broken by K2S2O8. 3) the produced product S2 might undergo further radical addition or cyclization reaction with the internal alkene group.

The reactivity of reagent E:
We tested the reactivity of reagent E [PhSO2CH2(H)NOBn] with 1a as the substrate under standard conditions, and the corresponding product was isolated in 24% yield, which is lower compared to reagent A [PhSO2CH2(CN)NOBn]. In our previous report, [12] the reactivity of reagent A and E was tested and the reaction rate of reagent A was three times faster compared to reagent E.