English

• Phase-transition materials revealing changeable properties under external stimulus have great application potentials in nonlinear-optical switch [1-4], caloric storage [5-7], and ferroic (ferromagnetic/ferroelectric/ferroelastic) materials [8-11]. Recent years have witnessed development of organic-inorganic hybrid crystals with adjustable multiple components in designing phase-transition materials [12-15]. Understanding the microscopic origin of phase transitions in hybrid crystals is of importance to regulate desired physical properties, however, meanwhile is a huge challenge because of the significantly increased complicity arising from the multiple molecular components.

  As an essential physical property for designing actuators, and mechanical switches, ferroelasticity, endowing switchable orientation states of spontaneous strain in crystals [16, 17], has attracted more and more attentions. According to Aizu [18], ferroelasticity emerges through phase transition from a paraelastic phase in higher-symmetry crystal system to a ferroelastic phase in lower-symmetry crystal system. For the design of ferroelastic phase transitions based on multi-component hybrid crystals, it is essentially important to deal with the delicate intermolecular interactions and crystal packing via rationally choosing molecular components. In this sense, judiciously substituting functional groups on the organic parts and changing metal ion species in the inorganic parts are important and effective methods to rationally adjust ferroelastic phase transitions for hybrid crystals [19-25]. For instance, halogen substitution on dimethyl-isopropyl-ethyl-ammonium in cadmium thiocyanate hexagonal perovskites brings increased ferroelastic transition temperature from 249 K to 368 K, benefiting from enhanced energy barrier for dynamic motion of organic cations [26]. Alternately substituting divalent metal ions by mono- and tri-valent metal ions in halide hexagonal perovskites could reduce the symmetry of inorganic chains, and thus endow the resulting hexagonal double perovskite, (piperidinium)₂[K^ⅠBi^ⅢCl₆], with a ferroelastic phase transition accompanied with a relatively large spontaneous strain [27]. Nevertheless, it is still a challenge to rationally regulate multi-component molecular ferroelastics, especially those with complicate multi-step phase transitions.

  Very recently, we have reported a unique hybrid crystal, (Me₃NCH₂CH₂OH)₄[Ni(NCS)₆] (1), revealing unusual low- and high-temperature dual ferroelasticity arising from a conventional and an inverse symmetry-breaking ferroelastic phase transition at 269 and 360 K, respectively [28]. Such complicate multi-step ferroelastic phase transitions were found highly associated with the hydroxy‑participated intermolecular interactions. In this sense, replacing –OH by a halogen group such as –Cl and –Br could be a simple and feasible method to bring different halogen-participated intermolecular interactions hence to tune ferroelastic phase transitions for such a hybrid-crystal family consisting of discrete inorganic [Ni(NCS)₆]⁴⁻ anion and organic quaternary ammonium. Our attempts have successfully yielded two new hybrid crystals, (Me₃NCH₂CH₂X)₄[Ni(NCS)₆] (X = Cl for 2, and Br for 3). We found that, 2 undergoes multi-step phase transitions with space-group changes of P1–P2₁/n–A2/a–Cmce from room temperature to 400 K, including an above-room-temperature ferroelastic transition, whereas 3 does not undergo ferroelastic transition but a solid-liquid phase transition accompanying with a slow decomposition at above 419 K. Herein, such distinct phase transitions in 2 and 3, as well as the hidden role of different intermolecular interactions, were comprehensively discussed via differential scanning calorimetry (DSC) measurements, in-situ variable-temperature single-crystal/powder X-ray diffraction, and Hirshfeld surface analyses.

  Thermal gravimetric analysis with a heating rate of 10 K/min showed that 2 could be stable up to 450 K under N₂ atmosphere, which is 40 K below that of 1. DSC curve of 2 showed three endothermic peaks at 323 K, 343 K, and 378 K, respectively, in the initial heating run up to 420 K, but only one exothermic peak at 351 K in the following cooling run, owing to a slow decomposition at high temperature close to 420 K (Fig. S3 in Supporting information). Therefore, to avoid sample decomposition, a DSC measurement on 2 was carried out at below 370 K (Fig. 1a), and revealed two pairs of endothermic/exothermic peaks at 323/318 K with an enthalpy change of 1.1 kJ/mol and 343/340 K with an enthalpy change of 0.8 kJ/mol, respectively. Such two-step reversible phase transitions were further confirmed by variable-temperature powder X-ray diffraction (PXRD) patterns (Fig. 1b), which revealed mergences or disappearances of several peaks when heating from 300 K to 350 K and a recovery when cooling back to 300 K. If the sample was heated to 400 K for a short period, a PXRD pattern with fewer peaks could be recorded. These facts implies that the structural symmetry increases upon heating. For convenience, we label the crystalline phases of 2 as 2α (below 323 K), 2β (between 323 K and 343 K), 2γ (between 343 K and 378 K), and 2δ (above 378 K), respectively.

  Figure 1

  

  Figure 1.  (a) DSC curves in temperature range of 300–360 K for 2. (b) Variable-temperature PXRD patterns of 2.

  DownLoad: Full-Size Img PowerPoint

  To understand the mechanisms of these structural phase transitions in 2, in-situ variable-temperature single-crystal X-ray diffraction were performed at 300, 330 and 355 K for 2α and 2β and 2γ, respectively. As shown in Fig. 2, 2α crystallizes in the triclinic space group P1 with the asymmetric unit containing two half [Ni(NCS)₆]⁴⁻ anions and four Me₃NCH₂CH₂Cl⁺ cations. The crystal packing mode of 2 is the same as 1, i.e., the inorganic coordination anions pack in a body-centered mode while the organic cations occupy the vacancies. Consequently, similar to the case in 1, eight Ni…Ni distances between adjacent inorganic units in 2α fall in range of 10.861–12.127 Å (Fig. S5 in Supporting information).

  Figure 2

  

  Figure 2.  Crystal structures of (a) 2α, (b) 2β, and (c) 2γ. For clarity, H atoms are omitted. Symmetry codes for 2α: (A) 1 − x, 1 − y, 1 − z; (B) x, −1 + y, z; for 2β: (A) −1/2 + x, 3/2 − y, −1/2 + z; (B) 1/2 + x, 3/2 − y, −1/2 + z; (C) 1/2 − x, −1/2 + y, 1/2 − z; (D) 3/2 − x, −1/2 + y, 1/2 − z; for 2γ: (A) 1/2 + x, 1 − y, z; (B) 1 − x, 1/2 − y, 1/2 − z; (C) 3/2 − x, −1/2 + y, 1/2 − z; (D) 3/2 − x, y, −z; (E): 1/2 − x, y, −z.

  DownLoad: Full-Size Img PowerPoint

  Upon heating to 330 K, 2β belongs to the space group P2₁/n, with the asymmetric unit consisting of one [Ni(NCS)₆]⁴⁻ anion and four two-fold disordered Me₃NCH₂CH₂Cl⁺ cations. For 2γ (355 K), it belongs to the space group A2/a (a nonstandard setting of C2/c for convenience in structural comparison) with the asymmetric unit containing half [Ni(NCS)₆]⁴⁻ anion, one Me₃NCH₂CH₂Cl⁺ cation being two-fold disordered, and two half Me₃NCH₂CH₂Cl⁺ cations being ordered. The Ni–N bonds fall in range of 2.071(4)–2.101(4) Å, 2.082(4)–2.092(4) Å, and 2.090(4)–2.109(6) Å for 2α and 2β and 2γ, respectively, while Ni–N–C angles vary in a range of 170.4(4)–179.0(4)^o, 168.3(5)–174.3(4)^o, and 171.3(5)–172.9(4)° for 2α and 2β and 2γ, respectively. Namely, the coordination geometries in 2α, 2β, and 2γ did not reveal significant differences. Symmetry element analyses for 2β and 2γ (Fig. S6 in Supporting information) showed that the appearance of additional 2-fold axes and glide planes in 2γ could be attributed to the slightly enhanced dynamic motion of organic cations and the synchronous displacement of inorganic units. The slight change in dynamic motion of polar organic cations during these two-step phase transitions was also reflected in the slowly increased dielectric constant (Fig. S4 in Supporting information) upon heating. It should be noted that no significant dielectric anomaly accompanies with phase transitions in 2, as the revealed motions of Me₃NCH₂CH₂Cl (such as the rotations of C–CH₃ bonds, the swaying of –CH₂–CH₂– group [30], and the slight swaying of entire cation) have little influence on dipole moments for entire crystal.

  The temperature-dependent lattice parameters were monitored for 2α, 2β and 2γ via Pawley refinements on PXRD patterns collected in a temperature range of 300–355 K. Consequently, the thermal expansion coefficients were calculated by the PASCal program (Fig. 3, Tables S3–S5 in Supporting information) [31], giving volume expansion coefficients of +155(8) × 10⁻⁶ K⁻¹, +163(22) × 10⁻⁶ K⁻¹, and +172(8) × 10⁻⁶ K⁻¹ for 2α, 2β and 2γ, respectively, showing little differences between these phases and reflecting slight change in molecular dynamic motion upon heating. For the transition from 2α to 2β, the b axis increase by 3.2% and cell volume expanse by 1.8%. Such significant changes are highly associated with the order-disorder transition of organic cations N1 and N4 (Fig. 2a) and the synchronous displacement of [Ni(NCS)₆]⁴⁻ octahedra reflecting in the changes on the Ni…Ni distances from 12.762 Å (Ni2…Ni2B) in 2α to 13.171 Å (Ni1B…Ni1D) in 2β. By contrast, the cell volume only changes by about 0.1% during 2β→2γ transition, as all organic cations in both 2β and 2γ are being in similar disordered states. Except the cation N2 in 2γ reveals a larger swing amplitude of –CH₂CH₂Cl group than that in 2β, leading to a relatively significant increase of adjacent Ni1…Ni distance along the a axis from 13.188 Å (Ni1B…Ni1A) in 2β to 13.297 Å (Ni1D…Ni1E) in 2γ.

  Figure 3

  

  Figure 3.  Normalized temperature-dependent cell parameters of 2.

  DownLoad: Full-Size Img PowerPoint

  During the phase transition from P2₁/n (2β) to P1 (2α), the number of symmetry elements varies from 4 to 2, and such symmetry breaking belongs to a ferroelastic phase transition with an Aizu notation of 2/mF1. Based on the cell parameters at 323 K deduced by linear fitting variable-temperature cell parameters (Table S4), the total spontaneous strain for this ferroelastic phase transition was estimated as 0.0475. For further inspection of the ferroelastic phase transition, a variable-temperature polarization microscopy was performed on a single crystal of 2. As shown in Fig. 4, no observable multi-domain structure was present in paraelastic phase at 330 K. After cooling to 300 K, some stripes corresponding to ferroelastic domains were observed, and these ferroelastic domains disappeared when heating back to 330 K, well confirming the ferroelastic transition at 323 K.

  Figure 4

  

  Figure 4.  Evolution of domain structures of the ferroelastic phase and single-domain paraelastic phase in the cooling-heating run of 2.

  DownLoad: Full-Size Img PowerPoint

  As disclosed by DSC measurement (Fig. S3), a high-temperature phase, i.e., 2δ, exists when heating to above 378 K. However, the attempts to collect the single-crystal X-ray diffraction data for 2δ at 400 K were failed, as the single crystallinity could not maintain at high temperature. Fortunately, a PXRD pattern for 2δ was successfully obtained at 400 K in a short period, and a Pawley refinement on this PXRD pattern (Fig. S7 in Supporting information) suggested that 2δ belongs to the orthorhombic space group Cmce with cell parameters of a = 19.541(5) Å, b = 17.867(4) Å, c = 13.235(3) Å, and V = 4620.9(1) ų, which are similar to those in the high-temperature orthorhombic phase of 1 (Table S2 in Supporting information). In other words, 1 and 2 have isostructural high-temperature phases (Cmce) with organic cations being "melt-like" disordered [29]. These facts indicated that both 1 and 2 reveal multi-step solid-solid phase transitions upon heating, in spite of their detailed changes in space group are different, i.e., P2₁/n–P4/mnc–Cmce for 1, but P1–P2₁/n–A2/a–Cmce for 2.

  In contrast, their Br-substituted analogue, i.e., 3, reveals totally different phase transitions upon heating. The single-crystal X-ray diffraction at 298 K showed that 3 crystallizes in the monoclinic space group P2₁/c with a crystallographically ordered asymmetric unit containing half [Ni(NCS)₆]⁴⁻ anion and two Me₃NCH₂CH₂Br⁺ cations. Similar to the case in 2, the Ni–N bonds and Ni–N–C angles in 3 vary in the range of 2.066(3)–2.097(3) Å and 168.9(4)–175.9(3)^o, respectively. However, the packing mode of 3 is different from 2 that the equal amounts of [Ni(NCS)₆]⁴⁻ anions and Me₃NCH₂CH₂Br⁺ cations alternately arrange into a layer parallel to the bc planes while the rest organic cations fill between the layers (Fig. 5 and Fig. S8 in Supporting information). As shown in Fig. S9 (Supporting information), no solid-solid structural phase transition was detected in 3 but a solid-liquid phase transition occurs at 419 K and followed by a slow decomposition. There facts implied that the Br-substitution bring about a much-enhanced energy barrier for order-disorder phase transition and a reduced thermal stability.

  Figure 5

  

  Figure 5.  Crystal packing diagrams, the Hirshfeld surfaces, and fingerprint plots for [Ni(NCS)₆]⁴⁻ in (a) 1 (b) 2 and (c) 3.

  DownLoad: Full-Size Img PowerPoint

  To understand the influence of functional group substitution of organic cations in view of the intermolecular interactions, the effective volumes and the Hirshfeld surfaces were calculated for [Ni(NCS)₆]⁴⁻ anions and organic cations in different phases of 1–3. With increasing sizes of functional groups from –OH to –Cl and –Br, the average cation effective volumes (in different phases) are 158 ų (P2₁/n), 165 ų (P4/mnc) and 177 ų (Cmce) in 1 [28], 203 ų (PT), 203 ų (P2₁/n) and 207 ų (A2/a) in 2, and 207 ų (P2₁/c) in 3, respectively (Fig. S10 in Supporting information), well reflecting an enhanced steric effect with employment of ever-larger functional groups. As shown in Fig. 5 and Fig. S11 (Supporting information), besides the inter-ionic coulomb interactions, the main intermolecular interactions are associating with the short contacts between S atoms from inorganic part and H atoms from organic part. In particular, all the smallest (d_i, d_e) pairs in fingerprints come from S…H short contacts, i.e., (1.43 Å, 0.82 Å) representing O–H…S hydrogen bond in 1, (1.58 Å, 1.02 Å) in 2 and (1.50 Å, 1.00 Å) in 3 both representing C–H…S interactions. Namely, the intermolecular interactions between halogen-substituted organic cations and inorganic [Ni(NCS)₆]⁴⁻ parts are weakened in 2 and 3, when comparing with the case in 1.

  As disclosed by Hirshfeld surface analyses for the organic cations, the functional groups in 1–3 participate in quite different intermolecular interactions to play an essential role in the crystal packing and phase transitions. In detail, the hydroxyl group with a smaller size features in forming O–H…S weak hydrogen bonds to inorganic part (Fig. S11a), eventually endowing 1 with multi-step phase transitions with an unusual symmetry breaking thanks to the entangling dynamics of inorganic and organic parts [28]. The chloro group with a medium size participates in much weaker C–H…Cl hydrogen bonds to adjacent organic parts (Fig. S11b), such that 2 has relatively weaker interactions between inorganic and organic parts, hence reveals multi-step phase transitions mainly arising from the step-wise dynamic changes of flexible organic parts. In contrast, the heavier and larger bromo group forms additional Br…Br interactions (Figs. S8 and S11c), eventually leading to a very different packing mode for 3 (Fig. 5c). Such weak but abundant intermolecular interactions, together with significant steric effect of bromo group, prevent the organic parts to freely rotate/sway to trigger order-disorder transition in crystalline state for 3 until melting at a high-enough temperature.

  In conclusion, through replacing –OH by –Cl and –Br groups on organic cations in 1, we obtained two new hybrid crystals, i.e., 2 and 3, and investigated the influences of such delicately changed groups on the crystal structures, phase transitions, and thermal stabilities. Both 2 and 3 have relatively weakened inorganic-organic intermolecular interactions, but they undergo distinct phase transitions. The medium-size chloro group only participates very weak C–H…Cl interactions, as a result, 2 possess a same packing mode with 1 and undergoes P1–P2₁/n–A2/a–Cmce multi-step phase transitions from room-temperature to 400 K, including a 2/mF1 ferroelastic transition. Whereas the larger bromo group forms additional Br…Br interactions to give a different crystal packing mode for 3, and eventually preventing order-disorder transition in crystalline state but leading to a solid-liquid phase transition at above 419 K. Such drastic influences on crystal packing and phase transition of halogen substitution well reflect the delicate balances established by abundant weak intermolecular interactions in multi-component crystals. To systemically understand these delicate balances for such kind of multi-component crystals based on discrete coordination anions (e.g., the present [Ni(NCS)₆]⁴⁻) and various organic cations [32, 33] may expand new material systems, similar to well-known molecular perovskites, for designing novel phase-transition materials not only based on their multi-step solid-solid phase transitions but also unique solid-liquid transitions [3].

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (NSFC, Nos. 22071273 and 21821003), and Local Innovative and Research Teams Project of Guangdong Pearl River Talents Program (No. 2017BT01C161).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2022.03.033.

  
  
• 1. [1]

     W.J. Xu, C.T. He, C.M. Ji, et al., Adv. Mater. 28 (2016) 5886–5890. doi: 10.1002/adma.201600895

  2. [2]

     S. Liu, L. He, Y. Wang, et al., Chin. Chem. Lett. 33 (2022) 1032–1036. doi: 10.1016/j.cclet.2021.07.039

  3. [3]

     S.Y. Zhang, X. Shu, Y. Zeng, et al., Nat. Commun. 11 (2020) 2752. doi: 10.1038/s41467-020-15518-z

  4. [4]

     Y. Li, B. Zhao, J.P. Xue, et al., Nat. Commun. 12 (2021) 6908. doi: 10.1038/s41467-021-27282-9

  5. [5]

     S.L. Chen, Z.R. Yang, B.J. Wang, et al., Sci. China Mater. 61 (2018) 1123–1128. doi: 10.1007/s40843-017-9219-9

  6. [6]

     H. Xu, W. Guo, J. Wang, et al., J. Am. Chem. Soc. 143 (2021) 14379–14385. doi: 10.1021/jacs.1c07521

  7. [7]

     C.R. Huang, X. Luo, X.G. Chen, et al., Natl. Sci. Rev. 8 (2021) nwaa232. doi: 10.1093/nsr/nwaa232

  8. [8]

     W.J. Xu, P.F. Li, Y.Y. Tang, et al., J. Am. Chem. Soc. 139 (2017) 6369–6375. doi: 10.1021/jacs.7b01334

  9. [9]

     Z.X. Zhang, C.Y. Su, J.X. Gao, et al., Sci. China Mater. 64 (2020) 706–716.

  10. [10]

      Y. Li, T. Yang, X. Liu, et al., Inorg. Chem. Front. 7 (2020) 2770–2777.

  11. [11]

      K. Li, Z.G. Li, J. Xu, et al., J. Am. Chem. Soc. 144 (2022) 816–823. doi: 10.1021/jacs.1c10188

  12. [12]

      X.X. Dong, N. Song, B. Huang, et al., Inorg. Chim. Acta 515 (2021) 120051. doi: 10.1016/j.ica.2020.120051

  13. [13]

      C. Shi, M.M. Hua, Z.X. Gong, et al., Eur. J. Inorg. Chem. 2019 (2019) 4601–4604. doi: 10.1002/ejic.201900994

  14. [14]

      W.J. Xu, Z.Y. Du, W.X. Zhang, et al., CrystEngComm 18 (2016) 7915–7928. doi: 10.1039/C6CE01485B

  15. [15]

      S.S. Wang, X.X. Chen, B. Huang, et al., CCS Chem. 1 (2019) 448–454. doi: 10.31635/ccschem.019.20190012

  16. [16]

      E.K.H. Salje, Annu. Rev. Mater. Res. 42 (2012) 265–283. doi: 10.1146/annurev-matsci-070511-155022

  17. [17]

      J. Sapriel, Phys. Rev. B 12 (1975) 5128–5140. doi: 10.1103/PhysRevB.12.5128

  18. [18]

      K. Aizu, J. Phys. Soc. Jpn. 28 (1970) 706–716. doi: 10.1143/JPSJ.28.706

  19. [19]

      Y. Yu, P. Huang, Y. Wang, et al., Chin. Chem. Lett. 32 (2021) 3558–3561. doi: 10.1016/j.cclet.2021.02.040

  20. [20]

      H. Ye, W.H. Hu, W.J. Xu, et al., APL Mater. 9 (2021) 031102. doi: 10.1063/5.0035793

  21. [21]

      L. Zhou, P.P. Shi, X. Zheng, et al., Chem. Commun. 54 (2018) 13111–13114. doi: 10.1039/C8CC07311B

  22. [22]

      L.H. Kong, D.W. Fu, Q. Ye, et al., Chin. Chem. Lett. 25 (2014) 844–848. doi: 10.1016/j.cclet.2014.05.028

  23. [23]

      O. Sato, Nat. Chem. 8 (2016) 644–656. doi: 10.1038/nchem.2547

  24. [24]

      Y. Ai, H.P. Lv, Z.X. Wang, et al., Trends Chem. 3 (2021) 1088–1099. doi: 10.1016/j.trechm.2021.09.010

  25. [25]

      B.Y. Wang, C.T. He, B. Huang, et al., Sci. China Chem. 58 (2015) 1137–1143. doi: 10.1007/s11426-015-5325-x

  26. [26]

      L. He, L. Zhou, P.P. Shi, et al., Chem. Mater. 31 (2019) 10236–10242. doi: 10.1021/acs.chemmater.9b04232

  27. [27]

      Q.R. Meng, W.J. Xu, W.H. Hu, et al., Chem. Commun. 57 (2021) 6292–6295. doi: 10.1039/D1CC02085D

  28. [28]

      D.X. Liu, X.X. Chen, Z.M. Ye, et al., Sci. China Mater. 65 (2021) 263–267.

  29. [29]

      E.L. Smith, A.P. Abbott, K.S. Ryder, Chem. Rev. 114 (2014) 11060–11082. doi: 10.1021/cr300162p

  30. [30]

      J.Z. Ge, X.Q. Fu, T. Huang, et al., Cryst. Growth Des. 8 (2010) 3632–3637.

  31. [31]

      M.J. Cliffe, A.L. Goodwin, J. Appl. Crystallogr. 45 (2012) 1321–1329. doi: 10.1107/S0021889812043026

  32. [32]

      J.Y. Liu, S.Y. Zhang, Y. Zeng, et al., Angew. Chem. Int. Ed. 57 (2018) 8032–8036. doi: 10.1002/anie.201802580

  33. [33]

      Z.H. Jia, J.Y. Liu, D.X. Liu, et al., J. Mater. Chem. C 9 (2021) 8076–8082. doi: 10.1039/D1TC01568K

• 

  Figure 1  (a) DSC curves in temperature range of 300–360 K for 2. (b) Variable-temperature PXRD patterns of 2.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Crystal structures of (a) 2α, (b) 2β, and (c) 2γ. For clarity, H atoms are omitted. Symmetry codes for 2α: (A) 1 − x, 1 − y, 1 − z; (B) x, −1 + y, z; for 2β: (A) −1/2 + x, 3/2 − y, −1/2 + z; (B) 1/2 + x, 3/2 − y, −1/2 + z; (C) 1/2 − x, −1/2 + y, 1/2 − z; (D) 3/2 − x, −1/2 + y, 1/2 − z; for 2γ: (A) 1/2 + x, 1 − y, z; (B) 1 − x, 1/2 − y, 1/2 − z; (C) 3/2 − x, −1/2 + y, 1/2 − z; (D) 3/2 − x, y, −z; (E): 1/2 − x, y, −z.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Normalized temperature-dependent cell parameters of 2.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Evolution of domain structures of the ferroelastic phase and single-domain paraelastic phase in the cooling-heating run of 2.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Crystal packing diagrams, the Hirshfeld surfaces, and fingerprint plots for [Ni(NCS)₆]⁴⁻ in (a) 1 (b) 2 and (c) 3.

  下载: 全尺寸图片 幻灯片
•