A luminescent view of the clickable assembly of LnF3 nanoclusters

Nanoclusters (NCs) bridge the gap between atoms and nanomaterials in not only dimension but also physicochemical properties. Precise chemical and structural control, as well as clear understanding of formation mechanisms, have been important to fabricate NCs with high performance in optoelectronics, catalysis, nanoalloys, and energy conversion and harvesting. Herein, taking advantage of the close chemical properties of Ln3+ (Ln = Eu, Nd, Sm, Gd, etc.) and Gd3+–Eu3+ energy transfer ion-pair, we report a clickable LnF3 nanoparticle assembly strategy allowing reliable fabrication of diversely structured NCs, including single-component, dimeric, core-shelled/core-shell-shelled, and reversely core-shelled/core-shell-shelled, particularly with synergized optical functionalities. Moreover, the purposely-embedded dual luminescent probes offer great superiority for in situ and precise tracking of tiny structural variations and energy transfer pathways within complex nanoarchitectures.

. where KET is the energy transfer rate from donor (GdF3 NP) to acceptor (EuF3 NP), c represents the "energy transfer strength" and is a constant in our system, r is the donoracceptor distance [4,5] . We would like to emphasize that the energy transfer rate here can also be understood as the energy acceptance rate of EuF3 NPs from adjacent GdF3 NPs.
However, those so far reported strategies are applied to calculate the energy transfer processes occurred between adjacent ions [5][6][7] . In our case, the donor (GdF3) and acceptor (EuF3) are separated NPs, and the connection among those individual NPs during selfassembly process makes it possible to realize efficient GdF3-EuF3 energy transfer. So, the modelling of our NC system is a bit complicated and the energy flux process including energy transfer, energy migration, and energy loss induced by concentration or solvent quenching must be considered at the same time. Even though we couldn't give a comprehensive calculation right now, we would like endeavor to give a precise-possible calculation based on rationally simplified models, in which the GdF3-EuF3 distances and the GdF3/EuF3 ratios are mainly concerned.
For dimeric NCs, the co-assembly of EuF3 NPs and GdF3 NPs undergoes completely.
As a typical example, the calculated Eu/Gd atomic ratio of 33.3:66.7 from elemental mapping result confirmed the EuF3&2GdF3 composition of the dimeric structure, which perfectly matches with our experimentally designs of respective NCs and suggests the perfect and complete self-assembly of EuF3 and GdF3 NPs. So, every single EuF3 NPs in EuF3&2GdF3 structure will accept more energy from surrounding GdF3 NPs than in the case of EuF3&GdF3 due to the increased GdF3/EuF3 ratio.
As shown in Supplementary Fig. 15a Supplementary Fig. 15a), c as a constant can be reduced. The calculated Ksum2/Ksum1 = 106%, suggesting the increased energy transfer rate in EuF3&2GdF3 compared with that in EuF3&GdF3. However, this value is smaller than the calculated increase of QYs: QYs2/QYs1 = 132%, which is reasonable because EuF3&GdF3 may suffer from more severe concentration quenching caused by cross relaxation between EuF3 NPs, due to the relatively low content of GdF3 compared with that in EuF3&2GdF3. So, the increased QYs benefited from not only the enhanced energy transfer efficiency but also the inhibition of concentration quenching among EuF3 NPs due to the spatial separation of GdF3 NPs. Thus, it is necessary to introduce a correction factor (α) to refine the final energy transfer rate and QYs: For core-shelled EuF3@GdF3 and EuF3@GdF3@GdF3 structures, the inner EuF3 NCs core and outer GdF3 NCs shell can be treated as an integral whole acceptor and donor respectively, and r is determined by the distance from geometric center of EuF3 core and GdF3 shell. As shown in Supplementary Fig. 15b, the energy transfer rate for EuF3 NCs core in EuF3@GdF3 (Ksum3) and EuF3@GdF3@GdF3 (Ksum4) can be calculated as follows: in which r3 = 23.9 nm refers to the sum of the radius of EuF3 NCs and half of the thickness of the first GdF3 shell, r4 refers to the sum of the radius of EuF3@GdF3 NCs and half thickness of the second GdF3 shell, which is 33.0 nm, and similarly the calculated value of R0, R1, and R2 is 20.1, 27.7, and 38.3 nm respectively, c as a constant can be reduced. The calculated Ksum4/Ksum3 = 128%, which shows obviously increased energy transfer rate in EuF3@GdF3@GdF3 than in EuF3@GdF3. However, this value is smaller than the calculated increase of QYs: QYs4/QYs3 = 162%, for that the increased QYs benefited from not only the enhanced energy transfer efficiency but also the inhibition of solvent quenching provided by GdF3 shell. It is also necessary to introduce a correction factor (β) here to refine the final energy transfer rate: where similarly as above, the calculated value of r5, r6 is 19.9 and 27.0 nm, respectively, while those of R3, R4, and R5 are 16.9, 22.9, and 31.1 nm, respectively, c as a constant can be reduced. The calculated Ksum6/Ksum5 = 130%, which shows obviously increased energy transfer rate in GdF3@EuF3@EuF3 than in GdF3@EuF3. However, this value is larger than the increased QYs: QYs6/QYs5 = 112%. This is also reasonable because although more EuF3 NPs in GdF3@EuF3@EuF3 receive energy from GdF3 core, which enhances the energy transfer efficiency, the energy transfer efficiency is also suffered from enhanced concentration quenching among EuF3 component. Moreover, since the energy transferred from Gd 3+ is fixed, the second layer of Eu 3+ can only absorb the residual energy that passes through the first Eu 3+ layer, and the relatively long distance from GdF3 core to EuF3 shell also hinders the energy transfer. So, a correction factor (γ) is introduced here to refine the final energy transfer rate: According to the above calculations, we can see that the increase of energy transfer rate occurs in all the dimeric, core-shelled, and reversely core-shelled NCs, which is consistent with our experimental data. However, it's worth pointing out that due to the contributions from the variation of energy loss processes including concentration and solvent quenching, the correction factor, α, β, and γ mentioned above, shall be considered to refine the model system.