This is a move I thought long and hard about, and would probably not make again in the same position. It has the obvious immediate advantage that it wins the backwards d6 pawn, giving White a queen-side pawn majority with endgame chances. On the other hand, it has long-term strategic disadvantages. a) it changes plans in the middle of the game. b) it trades off White's best knight on the best square of the board for Black's worst bishop on the most-blocked square of the board. And c) it leaves White with an awkward e4 pawn to defend. This may be a case in which the threat is stronger than the execution.

Black had nothing better here.

There now begins a new phase of the game. Having abandoned a very promising king-side attack for a one-pawn advantage, White must now submit to a period of defensive retreat and consolidation before he can even consider trying to advance the queen-side pawn majority. Black immediately sets about harrying the weak White center pawn.

White is only just barely hanging onto his one-pawn advantage here.

This prevents the White knight from occupying c5.

White must now shuffle and scurry to secure his weak center and activate his buried bishop. The White knight will come to d2 to hold the e4 pawn and allow the White rook to dispute the Black rook's control of the open file.

Preventing the White rook from coming to d3. Black is having things pretty much his own way here, and White is unable to make any use of his material advantage.

Preparing to advance the queen-side, at last, and reclaim some of the white squares from Black's dominant bishop.

Anticipating White's next move and allowing an escape square for the bishop.

Not c4, as the White bishop would then be completely entombed behind his own pawns.