Table 2.

Outcome of ball-milling reactions of 1 eq of Ac-dA(Bz) 2 (40 mg) with 1.5 eq of DMT-dC(Bz) phosphoramidite 1 and 2.5 eq of activator. Ball-milling frequency used was 25 Hz. Ionic liquids, NMI.Tf and Py.TFA were dried for 2 h under high vacuum before use. If a pre-milling step was used (entries 8–11), this involved milling 1 and 2 with the stated solvent for 5 min, followed by adding the activator, an additional 5 eq of solvent (total 10 eq) and milling for the stated length of time.

entry	liquid phase	milling time (h)	activator	pre-milling	1a	4a	3a
1	—	0.5	DCI	—	83	17	0
2	3 eq [C6mim][FAP]	0.5	DCI	—	80	17	3
3	3 eq [C6mim][NTf2]	0.5	DCI	—	52	38	10
4	5 eq DCM	0.5	DCI	—	24	56	20
5	5 eq MeCN	0.5	DCI	—	15	58	27
6	5 eq DCM	1	DCI	—	15	45	40
7b	5 eq MeCN	1	DCI	—	0	37	43
8	5 eq DCM	1	DCI	5 eq DCM	7	43	50
9	5 eq MeCN	1	DCI	5 eq MeCN	0	36	64
10	5 eq MeCN	1	NMI.Tf	5 eq MeCN	0	42	58
11	5 eq MeCN	1	Py.TFA	5 eq MeCN	0	40	60

^aPercentages of 1, 3 and 4 present were determined by ³¹P NMR analysis of the impure product in CDCl_3. 1 ppm, 149.1 and 148.9, 3 ppm 139.2 and 138.7, 4 ppm 7 and 6.9. The presence of two peaks by ³¹P NMR was due to the presence of diastereoisomers in 1.

^bPercentages of 1, 3, and 4 do not add up to 100 because of substantial formation of another by-product, proposed to be the pyrophosphite formed by a reaction between 1 and 4. See electronic supplementary material, figure S2.