“…[3] The crystal engineering of M/S/Ag clusters [5, 6] (M = Mo, W), in particular, has proven significantly more challenging for both control of complexity and diversity of structures than that of the Cu-containing counterparts. [2][3][4] 7] Most of the known 1D polymeric M/S/Ag clusters have been synthesised by following a single protocol, namely, the reaction of [NH 4 ] 2 This methodology is not broadly applicable because it has several clear disadvantages: 1) AgNO 3 reacts readily with [MS 4 ] 2À , resulting in a Ag 2 S precipitate that is very difficult to react with further [MS 4 ] 2À moieties; 2) only S atoms from [MS 4 ] 2À moieties serve as the bridge to build M/S/Ag 1D chain clusters-no other linking atoms or units participate; 3) the solvent-coordinated rare-earth cations are trivalent and usually induce trivalent repeat units in the anionic chains, reducing the diversity and novelty of polymeric M/S/Ag clusters; 4) the known 1D M/S/Ag clusters synthesised by this method are formed in very low yields, which is a significant disadvantage with respect to applications as precursors for functional materials. [6] Herein, we report a new synthetic method for constructing M/S/Ag clusters in high yields, structural studies that reveal an unusual discrete octanuclear planar "open" square-like skeleton for [SrA ][W 2 S 8 Ag 4 I 2 ]} n (3), NLO studies that demonstrate strong third-order non-linear refraction, absorption and a large optical limiting (OL) capability, tuning of non-linearity by conceptual replacement of Mo by W, and time-dependent density functional theory (TD-DFT) studies that provide insight into the electronic transitions and spectral characterisation of these functionalised NLO molecular materials.In our new procedure, Sr 2 + cations and AgI take the roles of the directing cations and Ag + source, respectively, for synthesising M/S/Ag clusters.…”