“…Thus and as shown in Scheme 4 , the Nozaki–Hiyama–Kishi (NHK) reaction ( 4 ) between 33 and 8 (NiCl 2 /CrCl 2 , 47% yield, unoptimized conditions) followed by cycloetherification [1,8-diazabicyclo[5.4.0]undec-7-ene (DBU)] afforded sulfone methyl ester 35 (via transient intermediate 34 ), which underwent smoothly the expected intramolecular displacement at the mesylate position under the basic conditions employed (81% yield), as shown in Scheme 4 . Unfortunately, however, attempts to induce the macrocyclization of p -tolylsulfone methyl ester 35 under various basic conditions [e.g., lithium diisopropylamide (LDA) ( 42 ), sodium hexamethyldisilazide (NaHMDS) ( 43 ), potassium hexamethyldisilazide (KHMDS) ( 44 ), potassium tert -butoxide (KO t -Bu) ( 45 )] failed to form the desired eribulin precursor 36 , leading us to search for an alternative approach.…”