Pushing up the Size Limit of Metal Chalcogenide Supertetrahedral


Pushing up the Size Limit of Metal Chalcogenide Supertetrahedral...

0 downloads 97 Views 2MB Size

Communication pubs.acs.org/JACS

Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Pushing up the Size Limit of Metal Chalcogenide Supertetrahedral Nanocluster Xiaofan Xu,†,⊥ Wei Wang,†,⊥ Dongliang Liu,† Dandan Hu,† Tao Wu,*,† Xianhui Bu,‡ and Pingyun Feng*,§ †

College of Chemistry, Chemical Engineering and Materials Science, Soochow University, Suzhou, Jiangsu 215123, China Department of Chemistry and Biochemistry, California State University, Long Beach, California 90840, United States § Department of Chemistry, University of California, Riverside, Riverside, California 92521, United States ‡

S Supporting Information *

MX4 cluster (T1) to the infinite zinc-blende-type framework (T∞), where is the limit to the size of Tn cluster that can be made? Here, we report a breakthrough in this pursuit. Specifically, we have synthesized supertetrahedral T6 cluster and pushed up the size limit from 91 atomic sites in T5-CuInS to 140 in T6-ZnInS. An examination of charge-density matching at both global and local levels contributed to this synthetic advance. Global charge (the charge of the whole cluster) has been recognized as a key factor in the stability of Tn and its propensity for crystallization.3a,b For the hypothetical isolated II-VI Tn with the formula of [Mn(n+1)(n+2)/6S(n+1)(n+2)(n+3)/6](n+1)(n+2)−, its global charge increases with the increase of cluster size. For example, T4-ZnS has the formula of [Zn20S35]30−; however, the global charge of T5-ZnS ([Zn35S56]42−) quickly goes up. Such highly negative charge stems from low-coordinated μ2- and μ3-S sites on edges and faces of the cluster. To address this charge issue, the use of surface-capping ligands5 has proved effective to reduce global charge, and sometimes results in the formation of other cluster types such as invert-Tn2b,f and Cn.6 The surfaceligand-capping method is, however, unsuitable for constructing open frameworks due to low stability and pore-blocking by organic groups. The method that avoids the use of surface organic groups, but can deal with global charge issue is to replace divalent metal (M2+) with higher-valent one (M3+ or M4+).7 However, this method creates a different issue: local charge balance (the distribution of total cationic charge around each anion), especially for large Tn, because μ4-S sites do not favor highvalent (>+2) metal ions. According to Pauling’s electrostatic valence rule, coordination modes such as (M3+)4(μ4-S) or (M4+)4(μ4-S) are unstable due to the excessive cationic charge around S2−. Furthermore, such instability becomes exacerbated for larger Tn because there will be more μ4-S sites (Table S1). To reconcile the needs for both global and local charge balance, a mixed-metal strategy was previously adopted, i.e., replacing some M2+ sites with high-valent ions while keeping metal sites surrounding μ4-S divalent or a mix of monovalent and trivalent. It is from this consideration that T4-ZnInS and T5-CuInS were prepared.3e By replacing M2+ ions at edges and corners with M3+ ions, the global charge and local charge

ABSTRACT: The cubic ZnS structure type and the sizedependent properties of related nanoparticles are of both fundamental and technological importance. Yet, it remains a challenge to synthesize large atom-precise clusters of this structure type. Currently, only supertetrahedral clusters with 4, 10, 20, and 35 metal sites (denoted as T2, T3, T4, and T5, respectively) are known. Because the synthesis of T5 in 2002, numerous synthetic efforts targeting larger clusters only resulted in T2−T5 clusters in various compositions and intercluster connectivity, with T6 (56 metal and 84 anion sites) being elusive. Here, we report the so-far largest supertetrahedral cluster (T6, [Zn25In31S84]25−). New T6 clusters can serve as the host matrix for optically active centers. Mn-doped variants of T4 and T6 have also been made, allowing the investigation of site-dependent Mn emission. The results lead to the elucidation of the mechanism regulating Mn emission via size-dependent crystal lattice strain and provide new insight into Mn-doping chemistry in cluster-based chalcogenides at the atomic level.

G

roup II-VI semiconductor nanoparticles (also called quantum dots or QDs) have attracted great interest because their quantum confinement effect gives rise to size- and composition-dependent properties.1 However, unraveling their composition−structure−property correlation remains challenging due to the difficulty in obtaining atomic-level features. Molecular nanoclusters (NCs) with well-defined structure are desirable to address this challenge because high-quality single crystals formed from NCs allow the structure characterization via single crystal X-ray diffraction.2 The series of NCs that bear the closest resemblance to zincblende-type semiconductor are called supertetrahedral Tn NCs (n is the number of metal sites along tetrahedron edge).3 Over the past two decades, chalcogenide supertetrahedral NCs have undergone rapid development. Yet, despite efforts targeting synthesis of new Tn, prior to this work, the largest one was still T5-[Cu5In30S56]17− first reported in 2002.3e Since then, T5 clusters with a range of compositions have been synthesized as either molecular crystals or extended frameworks,4 but T6 clusters remained elusive and out of reach. There has been a fundamental and longstanding question: from an individual © XXXX American Chemical Society

Received: November 15, 2017

A

DOI: 10.1021/jacs.7b12092 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Communication

Journal of the American Chemical Society balance can be satisfied. For example, in the reported T4ZnInS, its edge and corner metal sites are occupied by In3+, and four Zn2+ ions are located on four faces (Figure 1a). As a result,

Figure 1. Ideal Tn-ZnInS and the corresponding Zn−S core.

the global charge is reduced from −30 to −14. For T5-ZnInS, the reported one was [Zn13In22S56],7b in which there are 22 In3+ sites at edges and corners, 12 Zn2+ ions on faces, and one Zn2+ surrounded by four μ4-S sites at core (Figure 1b). This reduces the charge to −20, much less than −42 in the T5-ZnS. In principle, the mixed-metal strategy should be applicable for building larger Tn. However, no success was reported prior to this work. We realize that among reported Tn, the ratio of M2+/M3+ is much less than 1. Presumably, the choice of such ratios during the synthesis is not random, but reflects an emphasis on the reduction of global charge. Based on the above analysis, to synthesize larger Tn, the ratio of M2+/M3+ needs to be varied to be in sync with the increased number of core μ4-S sites, and also with the charge density of structure-directing agents. Quite different from T4 and T5 with the Zn2+/In3+ ratios much less than 1, the ideal T6-ZnInS would have the formula of [Zn28In28S84]28− with the Zn/In ratio equal to 1, in which there are 24 Zn2+ on faces and 4 Zn2+ at core, and 10 μ4S sites (Figure 1c and Table S2). Such configuration reduces the cluster charge from −56 to −28. Indeed, OCF-100-ZnInS (OCF = organically directed chalcogenide frameworks) composed of T6-ZnInS was successfully obtained. In addition, we further prepared OCF-99-ZnInS made of T4-ZnInS. The phase purity and thermal stability of OCF-99 and OCF100 were investigated by PXRD and TGA (Figure S1, S2). The framework compositions were determined as [Zn4In16S35]14− and [Zn25In31S84]25− via ICP-MS and EDS data (Figure S3, S4). Single-crystal X-ray diffraction revealed that layered OCF-99 and OCF-100 crystallize in monoclinic space group P21/n and C2/c, respectively (Figure 2a and Table S3). Notably, the Zn/ In ratio in OCF-100 is 25:31. It is important to study the distribution of Zn2+/In3+ in T6 and its impact on crystal lattice strain because it is relevant to the mechanistic aspects of T6 formation and to the use of T6 clusters as the host matrix for optical applications. With the help of Pauling’s electrostatic valence rule and analysis based on M−S bond lengths (Figure 2b and Table S4), we concluded that all Zn ions are located at core and on the faces of NCs. The M−(μ4-S) bond lengths at core area in T6-ZnInS are in the range of 2.300−2.369 Å, very close to that of Zn−S in bulk ZnS (2.345−2.349 Å) and much less than that of In−S (2.453− 2.526 Å) in T5-InS from UCR-15, which supports the conclusion that four core sites are occupied by Zn2+. In addition, M−(μ4-S) and M−(μ3-S) bond lengths at core and face area are approximately 2.284−2.430 Å, being slightly

Figure 2. (a) 2D structure of OCF-99 (left) and OCF-100 (right). (b) M−S bond length in T6-ZnInS. (c) Proposed polyhedron of Mndoped NCs with one dopant.

greater than that of Zn−(μ4-S), but slightly less than that of In−(μ3-S) (2.475−2.526 Å) in T5-InS from UCR-15, which indicates that there are some In3+ ions residing on faces. Therefore, for the real T6-ZnInS, we postulate there are three extra In3+ ions located on faces. The M−(μ3-S) and M−(μ2-S) bond lengths (2.398−2.537 Å) at the edge and corner area are slightly greater than that of In−(μ4-S) in In2S3 (2.439−2.481 Å) and similar to that in T5-InS (2.391−2.526 Å), which indicates that the metal sites at the edge of T6 are completely occupied by In3+ ions. Notably, the big mismatch between In−S and Zn−S bond lengths implies the outer “In−S” surface layer suffers from tensile strain, and the “Zn−S” core in turn receives compressive strain (Figure 2c). The T6 reported here provides a new platform for probing Mn2+-related photoluminescence (PL) emission. The emission by Mn-doped semiconductors has attracted great interest due to the large Stokes shift relative to band-edge or defect-induced emission from the hosts.2a,8 Such dopant emission is generally attributed to the spin-forbidden relaxation between 4T1 and 6A1 state of Mn2+. The 4T1 and 6A1 energy states are mainly affected by the coordination environment of Mn2+, such as the nature of ligand (L), Mn−L bond length and symmetry of ligand field.8a Different from Mn-doped QDs displaying orange emission in the range of 580−600 nm, the Mn-doped T5-CdInS shows an obvious red-shifted emission, which is caused by large “crystal lattice strain” naturally formed in “Mn@CdS@InS” “core− shell” nanostructure with one Mn dopant at core center of supertetrahedron.2a,8c To further probe the size-dependent Mn emission in Tn, we sought to prepare Mn-doped T4-ZnInS and T6-ZnInS. We observed that complete replacement of Zn2+ with Mn2+ in T4 and T6 did not occur. In addition, to avoid the formation of difficult-to-separate byproducts and the possible big distortion in Tn structure caused by more Mn dopants, we purposely controlled the amount of Mn dopant to achieve the lowest possible levels. We succeeded in obtaining two pure Mn-doped B

DOI: 10.1021/jacs.7b12092 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Communication

Journal of the American Chemical Society

bond length, Mn dopant is postulated to reside on the face of the tetrahedron, i.e., at the interface between “Zn−S” core and “In−S” shell, where a serious “crystal lattice strain” is usually created due to big mismatch in bond length among In−S, Mn− S, and Zn−S. The “Zn−S” core seems to serve as a “cushion” layer to tune lattice strain (Figure 4a). If the “Zn−S” core is

samples (denoted as T4-MnZnInS and T6-MnZnInS), corresponding to [Mn0.54Zn3.46In16S35] and [Mn1.06Zn23.84In31S84]. On the basis of such Mn/Zn ratio determined by EDS, we postulate that there is no more than one Mn dopant in T4 and T6. Moreover, according to the described analysis on M−S bond lengths, the Mn ion will most likely replace the Zn site on the face of the supertetrahedron because Mn−S bond length (2.421−2.438 Å) in bulk MnS is greater than that of the Zn−S bond and similar to that of (Zn/ In)−S on faces of T6-ZnInS. The proposed structure of Tn with one Mn dopant is shown in Figure 2c. OCF-99s and OCF-100s exhibit the characteristic semiconducting property. The optical band gaps were determined to be 3.36, 3.23, 3.25, and 2.99 eV for T4-ZnInS, T4-MnZnInS, T6-ZnInS and T6-MnZnInS, respectively, by UV−vis diffuse reflectance spectra (Figure 3a and Figure S5). We can find two

Figure 4. (a) Proposed mechanism on Mn emission tuned by the thickness of “buffer cushion”. (b,c) Temperature-dependent PL spectra of T4-MnZnInS and T6-MnZnInS.

thick enough, like the one in T6, there may be weak torsion or distortion found in Mn−S bond and coordination geometry of Mn dopant, which is caused by the compressive strain from the outer “In−S” shell. However, when the “Zn−S” core is small, as observed one in T4, the “buffering effect” may be very weak. Finally, strong torsion or big distortion in Mn−S bond and coordination geometry of Mn ion is formed. Therefore, the “Zn−S” core behaves like “buffer cushion” to release the “crystal lattice strain”. Relative to T4-MnZnInS, the crystal field splitting of Mn2+ d-orbitals in T6-MnZnInS is decreased, and it in turn leads to the lifted 4T1 level and the blue shift in PL emission.9 On the other hand, a blue shift in Mn PL emission is also an indirect proof implying that Mn dopant is located on face, instead of core site of T6. On the basis of the described analysis on the “buffering effect”, we can make two rational predictions: (1) Mn-doped sample may display further blueshifted emission for large-sized Tn, compared with small-sized T4 NC; (2) Mn emission peak position is more sensitive to temperature for small-sized Tn. Although no direct proof can be obtained so far for the former prediction, our lowtemperature PL measurement confirms the latter one. T4MnZnInS displays a larger red shift (∼27 nm) in comparison with T6-MnZnInS (∼15 nm) when measurement temperature is changed from room temperature to 63 K (Figure 4b,c). Tn NC usually undergoes contraction under frozen condition and has shortened M−S bond length, which correspondingly leads to PL red shift. It is not difficult to understand that a large “Zn−S” core has a strong “buffering effect” to resist the big change of Mn−S bond length during shrinking. In summary, we have narrowed down the size gap between molecular NCs and II-VI semiconductor NPs by pushing up the size limit of supertetrahedral Tn into the unprecedented T6 member, which represents the largest one in Tn family so far. The synthetic strategy provided here may be useful for building

Figure 3. (a) UV−vis absorption spectra. (b,c) PLE and PL spectra. (d) PL decay curve.

facts that OCF-100s have smaller band gap than OCF-99s, and Mn-doped samples have a smaller band gap than the undoped ones. The former phenomenon can be explained by quantum confinement effect, and different electronic structure tuned by Mn dopants could account for the latter. T4-ZnInS has negligible PL emission, but its Mn-doped sample displays an obvious broad PL emission at 621 nm under the excitation of 362 nm (Figure 3b). However, T6-MnZnInS shows strong PL emission at 612 nm when excited at 361 nm (Figure 3c), shorter than that of T4-MnZnInS. On the basis of PL excitation (PLE) spectra, we can conclude that Mn emission comes from a rapid energy transfer from host NCs’ lattice into the split Mn2+ d-orbital and subsequent 4T1→6A1 transition. Such energy transfer from photoexcited NCs to Mn ion is also implied due to the fact that PLE spectrum is consistent with the absorption one. In addition, PL quantum yield (PLQY) of T6MnZnInS (12.11%) is much higher than that of T4-MnZnInS (0.68%) (Figure S6), and T6-MnZnInS also displays longer lifetime (707.57 μs) than T4-MnZnInS (151.03 μs) (Figure 3d). Generally, increasing crystal field splitting of Mn2+ d-orbitals can lower 4T1 level and correspondingly lead to red shift in PL emission.8b To understand the PL variation trend between Mndoped T4 and T6, we carefully studied their structure. Mndoped Tn NC can be structurally regarded as “MnZnS@InS” “core−shell” configuration. While considering ionic radius and C

DOI: 10.1021/jacs.7b12092 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Communication

Journal of the American Chemical Society

(4) (a) Xiong, W.-W.; Li, J.-R.; Hu, B.; Tan, B.; Li, R.-F.; Huang, X.Y. Chem. Sci. 2012, 3, 1200. (b) Su, W.; Huang, X.; Li, J.; Fu, H. J. Am. Chem. Soc. 2002, 124, 12944. (5) (a) Vaqueiro, P.; Romero, M. L. J. Am. Chem. Soc. 2008, 130, 9630. (b) Vaqueiro, P.; Romero, M. L. Chem. Commun. 2007, 3282. (6) Herron, N.; Calabrese, J. C.; Farneth, W. E.; Wang, Y. Science 1993, 259, 1426. (7) (a) Zheng, N.; Bu, X.; Wang, B.; Feng, P. Science 2002, 298, 2366. (b) Zheng, N.; Bu, X.; Feng, P. Nature 2003, 426, 428. (c) Zhang, X.M.; Sarma, D.; Wu, Y.-Q.; Wang, L.; Ning, Z.-X.; Zhang, F.-Q.; Kanatzidis, M. G. J. Am. Chem. Soc. 2016, 138, 5543. (8) (a) Pradhan, N. ChemPhysChem 2016, 17, 1087. (b) Beaulac, R.; Archer, P. I.; Ochsenbein, S. T.; Gamelin, D. R. Adv. Funct. Mater. 2008, 18, 3873. (c) Lin, J.; Hu, D.-D.; Zhang, Q.; Li, D.-S.; Wu, T.; Bu, X.; Feng, P. J. Phys. Chem. C 2016, 120, 29390. (9) Zuo, T.; Sun, Z.; Zhao, Y.; Jiang, X.; Gao, X. J. Am. Chem. Soc. 2010, 132, 6618.

larger-sized Tn members. Importantly, investigation on the sizedependent Mn emission from Mn-doped Tn becomes feasible to probe the relative site of Mn dopant in such ternary Zn−In− S NC. Compared to small-sized T4-MnZnInS, large-sized T6MnZnInS exhibits shorter emission wavelength, higher PLQY, longer lifetime, and greater resistance to temperature in the change of emission peak position. The present study on Mn emission also provides a fundamental understanding of Mndoping chemistry in NC-based chalcogenides at the atomic level.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b12092. Experimental section, additional figures and tables on structure descriptions, powder XRD, PLQY (PDF) Crystallographic data for OCF-99 and OCF-100 (CIF)



AUTHOR INFORMATION

Corresponding Authors

*[email protected] *[email protected] ORCID

Tao Wu: 0000-0003-4443-1227 Xianhui Bu: 0000-0002-2994-4051 Pingyun Feng: 0000-0003-3684-577X Author Contributions ⊥

These authors contributed equally.

Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We acknowledge National Natural Science Foundation of China (T.W., 21671142), Jiangsu Province Natural Science Fund for Distinguished Young Scholars (T.W., BK20160006), and NSF (P.F., DMR-1506661).



REFERENCES

(1) (a) Talapin, D. V.; Lee, J.-S.; Kovalenko, M. V.; Shevchenko, E. V. Chem. Rev. 2010, 110, 389. (b) Kovalenko, M. V.; Scheele, M.; Talapin, D. V. Science 2009, 324, 1417. (c) Hodes, G. Adv. Mater. 2007, 19, 639. (2) (a) Lin, J.; Zhang, Q.; Wang, L.; Liu, X.; Yan, W.; Wu, T.; Bu, X.; Feng, P. J. Am. Chem. Soc. 2014, 136, 4769. (b) Beecher, A. N.; Dziatko, R. A.; Steigerwald, M. L.; Owen, J. S.; Crowther, A. C. J. Am. Chem. Soc. 2016, 138, 16754. (c) Wu, T.; Zhang, Q.; Hou, Y.; Wang, L.; Mao, C.; Zheng, S.-T.; Bu, X.; Feng, P. J. Am. Chem. Soc. 2013, 135, 10250. (d) Rosemann, N. W.; Eußner, J. P.; Dornsiepen, E.; Chatterjee, S.; Dehnen, S. J. Am. Chem. Soc. 2016, 138, 16224. (e) Rosemann, N. W.; Eußner, J. P.; Beyer, A.; Koch, S. W.; Volz, K.; Dehnen, S.; Chatterjee, S. Science 2016, 352, 1301. (f) Beecher, A. N.; Yang, X.; Palmer, J. H.; LaGrassa, A. L.; Juhas, P.; Billinge, S. J. L.; Owen, J. S. J. Am. Chem. Soc. 2014, 136, 10645. (g) Cahill, C. L.; Parise, J. B. J. Chem. Soc., Dalton Trans. 2000, 1475. (3) (a) Feng, P.; Bu, X.; Zheng, N. Acc. Chem. Res. 2005, 38, 293. (b) Bu, X.; Zheng, N.; Feng, P. Chem. - Eur. J. 2004, 10, 3356. (c) Li, H.; Laine, A.; O’Keeffe, M.; Yaghi, O. M. Science 1999, 283, 1145. (d) Li, H.; Kim, J.; Groy, T. L.; O’Keeffe, M.; Yaghi, O. M. J. Am. Chem. Soc. 2001, 123, 4867. (e) Bu, X.; Zheng, N.; Li, Y.; Feng, P. J. Am. Chem. Soc. 2002, 124, 12646. D

DOI: 10.1021/jacs.7b12092 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX