Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices

Coordination Chemistry Reviews xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Coordination Chemistry Reviews journal homepage: www.elsevier.com/locate/ccr

Review

Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices William P. Lustig, Jing Li ⇑ Department of Chemistry, Rutgers University, 610 Taylor Road, Piscataway, NJ 08854, USA

a r t i c l e

i n f o

Article history: Received 20 July 2017 Received in revised form 12 September 2017 Accepted 17 September 2017 Available online xxxx

a b s t r a c t The development of lower cost and higher performance phosphor materials for energy efficient lighting and other optoelectronic applications is both necessary and feasible. Luminescent metal–organic frameworks and coordination polymers (LMOFs and LCPs, respectively) are a class of materials that hold great promise for this application. Their luminescence is eminently tunable, and a myriad of structures with incredibly diverse properties have been reported, with emission wavelengths covering the entire visible spectrum, white light emission from a variety of mechanisms, and quantum yields approaching unity. This review will briefly describe the luminescence mechanisms commonly observed in these materials, discuss strategies for the rational design of LMOF/LCP phosphors, and present a number of representative examples of each mechanism and/or design strategy. Ó 2017 Elsevier B.V. All rights reserved.

Contents 1.

2.

3. 4.

5.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1. General lighting devices and solid state lighting technologies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2. Luminescent metal–organic frameworks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Luminescence mechanisms in LMOF/LCP phosphors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Localized luminescence processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Charge transfer luminescence processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . LMOF/LCP phosphor design strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rare earth metal based LMOFs/LCPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Lanthanide LMOFs/LCPs with colored emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Lanthanide LMOFs/LCPs with white emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Actinide LMOFs/LCPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rare earth metal free LMOFs/LCPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Rare-earth metal free LMOFs/LCPs with colored emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

00 00 00 00 00 00 00 00 00 00 00 00 00

Abbreviations: 3,5-dsb, 3,5-disulfobenzoate; Ac, acetone; ad, adenine; AF, acriflavine; bdc, 1,4-benzenedicarboxylate; bp4mo, N-oxide-4,40 -bipyridine; bpdc, 4,40 -biphe nyldicarboxylate; bpe, 4,40 -bis(4-pyridyl)ethane; bpee, 4,40 -bipyridyl-ethylene; bpp, 1,3-bis(4-pyridyl)propane; bpy, 4,40 -biphenyl; btb, 1,3,5-tris(4-carboxyphenyl)-benze ne; btc, 1,3,5-benzenetricarboxylate; btec, benzene-1,2,4,5-tetracarboxylate; btpca, 1,10 ,100 -(benzene-1,3,5- triyl)tripiperidine-4-carboxylate; DEF, diethylformamide; DMA, dimethylacetamide; DMA+, dimethylammonium; DMF, dimethylformamide; dppy, 4-pyridyldiphenylphosphane oxide; ettc, 40 , 4000 , 4000 00 , 4000 000 0 -(ethene-1,1,2,2-tetrayl) tetrakis (([1,10 -biphenyl]-4-carboxylate)); glu, glutarate; H2dcbppy, 2,20 -bipyridine-4,40 -dicarboxylic acid; H2ida, iminodiacetic acid; H2idpa, 5-(1-oxoisoindolin-2-yl)isophthalic acid; H3ttaa, N,N0 ,N00 -1,3,5-triazine-2,4,6-triyltris(4-aminomethylbenzoic acid); H4L6, 2-hydroxy-trimesic acid; H6tatpt, 2,4,6-tris(2,5-dicarboxylphenylamino)-1,3,5-triazine; hfa, hexafluoroacetylacetonate; HL1, 2-(2-sulfophenyl)-imidazo(4,5-f)(1,10)-phenanthroline; Hppy, 2-phenylpyridine; Htzib, 1-tetrazole-4-imidazole-benzene; im, imidazolate; imdc, 4,5-imidazoledicarboxylate; ina, isonicotinate; ip, isophthalate; L2, 3,5-bis(3-carboxyphenyl)-1,2,4-triazole; L3, p-terphenyl-2,200 ,2000 ,5,500 ,5000 -hexacarboxylate; L4, p-terphenyl-3,200 ,300 ,5,500 ,5000 -hexacarboxylate; L5, cyanobenzoate; m-bdc, 1,3-benzenetricarboxylate; ndc, 2,6-naphthalenedicarboxylate; oda, oxydiacetate; phda, phenylenediacetate; phen, 1,10-phenanthroline; ppmc, 2-phenylpyrimidine-4-carboxylate; py, pyridine; Q1, bis-8-hydroxyquinoline; sdc, (E)-4,40 -(ethene-1,2-diyl)dibenzoate; TBA+, tetrabutylammonium; tbapy, (1,3,6,8-tetrakis(p-benzoate)pyrene; tcbb, 1,3,5-tris(40 -carboxy[1,10 -biphenyl]-4-yl)benzene; tcbpe-F, 40 ,4000 ,4000 00 ,4000 000 0 -(ethene-1,1,2,2-tetrayl) tetrakis(3-fluoro-[1,10 -biphenyl]-4-carboxylate); tf-bdc, tetrafluorobenenedicarboxylate; thca, p-ter-phenyl-3,30 ,300 ,5,50 ,500 -hexacarboxylate; tib, 1,3,5-tris(imidazolyl) benzene; tppa, tri(4-pyridylphenyl)amine; tppe, 1,1,2,2-tetrakis(4-(pyridin-4-yl)phenyl)ethane; ttb, 4,40 ,400 -s-triazine-2,4,6-triyl-tribenzoate; ttca, triphenylene-2,6,10tricarboxylate. ⇑ Corresponding author. E-mail address: [email protected] (J. Li). https://doi.org/10.1016/j.ccr.2017.09.017 0010-8545/Ó 2017 Elsevier B.V. All rights reserved.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

2

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

6.

7.

8.

5.2. Rare-earth metal free LMOFs/LCPs with white emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Host–guest LMOF/LCP phosphors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1. Host–guest LMOFs/LCPs with colored emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. Host–guest LMOFs/LCPs with white emission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other types of LMOF/LCP materials for lighting devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1. Hybrid LMOF/LCP materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2. Electroluminescent LMOFs and LCPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction 1.1. General lighting devices and solid state lighting technologies Global efforts to improve energy efficiency are important for reducing energy cost and consumption, decreasing carbon dioxide emission and slowing down global warming. A significant portion of global energy is directed toward infrastructure – transportation of goods and people, lighting, and heating/cooling are responsible for the bulk of energy use [1,2]. Improving the energy efficiency of these processes will lead to significant payoffs in global energy usage. Lighting is an especially attractive target, as it accounts for a significant portion of energy use. Developing more efficient lighting technologies has already begun, and addressing its energy efficiency requires less alteration of global infrastructure. Currently, three main types of general lighting technologies exist. Conventional incandescent bulbs generate white light by heating a filament to incandescence. Fluorescent bulbs function by ionizing mercury vapor through the use of an electric current, which produces UV radiation. This UV radiation excites a phosphor material on the interior surface of the bulb, which emits white light. Solid-state lighting based on light-emitting diodes (LEDs) uses an electroluminescent diode to produce narrow emission peaks, which can be converted into white light in a variety of ways. In multi-chip LEDs, white light is produced by mixing emission from red, green, and blue LED chips. However, using three LED chips drastically increases the cost of these bulbs. In phosphorconverted white LEDs (pc-WLEDs), phosphors excited by a

00 00 00 00 00 00 00 00 00 00

single-chip LED produce white light, either directly or by combining the emission of the selected chip. There are three main varieties of pc-WLEDs. In the first, a UV-emitting LED chip is used to excite a mix of red, green, and blue phosphor materials to produce white light. The second is similar, with the UV-emitting LED chip exciting a phosphor which directly produces white light. The third common variety is a blue chip based pc-WLED, in which a blue-emitting LED chip is used to excite a yellow phosphor or multicomponent phosphors. The combined emissions from the blue chip and phosphor(s) give the white light. When qualifying the light produced by a lighting device, two important characterization metrics are the color temperature and chromaticity. The color temperature of an emissive material relates the color of light produced to the temperature at which an ideal black body radiator would produce light of the same color. As such, it is only of use when describing light colors produced by black body radiators, from red, through orange and yellow, and into white light. It is most commonly used to indicate whether a bulb produces ‘‘cold” blue-white light (higher color temperatures) or ‘‘warm” yellow-white light (lower color temperatures) and is provided for most commercial light bulbs. Chromaticity describes the color of light more completely. The international standard method of plotting chromaticity was developed by the International Commission on Illumination (CIE) in 1931, which uses a coordinate system to indicate a specific color (Fig. 1) [3]. The CIE coordinates of a given light source may be calculated using its spectral power distribution and three color matching functions, allowing the hue of light perceived by the human eye to be determined from spectral data. For pure white light, the CIE coordinates are (0.333, 0.333). While LED bulbs are the most energy efficient and longerlasting general lighting technology, their highest initial cost has slowed their adoption. This is unfortunate, as the US Dept. of Energy has estimated that if the United States switched to entirely LED lighting, over 300 TWh of energy would be saved annually, which is nearly double the amount expected to be generated by wind and solar power generation plants by 2030 [5]. As the phosphor materials currently used in WLEDs rely on rare-earth elements (REEs), which contribute significantly to their high cost, developing new, more efficient phosphors materials that have little or no dependence on REEs could reduce the cost of these devices, resulting in faster adoption of the technology and major global energy savings.

1.2. Luminescent metal–organic frameworks

Fig. 1. CIE chromatography plot, showing the colors corresponding to each region of the plot. Reproduced with permission from Ref. [4]. Copyright 2016, Elsevier B. V.

Metal-organic frameworks (MOFs) or coordination polymers (CPs) are crystalline solids composed of single metal ions (primary building units, PBUs) or metal ion clusters (secondary building units, SBUs) linked together by organic ligands with multiple binding sites to form extended network structures. As MOFs and CPs are crystalline materials, diffraction techniques can provide precise information about their structure, while their chemical and

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

3

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx Table 1 List of selected LMOFs and LCPs, organized by emission color, emission mechanism, and quantum yield, with excitation and emission wavelengths also given. Emission Color

Emission Mech.

QY (%)

Formula

kex (nm)

kEM (nm)

Ref.

Blue Blue Blue Blue Blue Blue Blue Blue Blue Blue Blue Blue

LC LC LC LC LC LC LC LC LC LC, LLCT LC, LLCT LLCT

– – – – – 22.5 35 40.3 99.9 28.1 39.8 30

[GdCd2(imdc)2(Ac)(H2O)2]H2O LaZn4(imdc)4(Him)4 [In3(btb)4]3DMA+, 12DMF, 22H2O In2(OH)2(tbapy) [Gd7(3,5-dsb)4(OH)9(H2O)15]4H2O Zn(idpa)(py) GdCl3(bpy)(py)2]py [Cd(ttaa)]DMA+, 2H2O Zr6(OH)6(ettc)3 [Cd(idpa)(bpp)]nH2O [Zn(idpa)(phen)(H2O)]nH2O Zr6(OH)6(ndc)6

350 354 365 405 270 365 365 313 350 365 340 371

425 430 420 460 450 420 530 430 470 435 423 400

[21] [22] [23] [24] [25] [26] [27] [28] [15] [29] [29] [30]

Green Green Green Green Green Green Green Green Green Green Green Green Green Green Green Green Green

LLCT LC, MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S

50.9 1.7 2.6 25 31.3 31.6 32 47.4 65.6 67 67 74.8 78 78.9 86 90.0 95.2

[Zn2(tcbpe-F)(bpy)]xDMA Tb[Zn8(ad)4(bpdc)6O]2DMA+, 8DMF,11H2O [TbCd2(imdc)2(Ac)(H2O)2]H2O [Tb(H2L4)(H2O)4]DMA+0.5, DMF, xH2O [Tb4(btec)3(H2O)12]20H2O Tb2(ap)3(H2O)2(bpy) [Tb(L3)0.5(H2O)2]2H2O [Gd2Tb2(btec)3(H2O)12]20H2O Ln(ppmc)3(phen) Tb(HL6)(H2O)2 [Tb(tf-bdc)(NO3)(DMF)2]DMF [Tb2(oda)6Cd2][Cd(H2O)6]2+, H2O [Tb2(m-bdc)3(phen)2]DMF [Tb2(oda)6Cd3(H2O)6]6H2O [TbCl3(bpy)(py)2]py Tb(ina)3(H2O)2 [Tb2(oda)6Cd3]xAc

455 365 350 348 306 340 354 306 362 375 378 378 350 355 300 366 355

514 420, 495, 485, 485, 490, 485, 485, 480, 480, 489, 485, 485, 485, 486, 485, 485,

Yellow Yellow Yellow Yellow Yellow Yellow Yellow Yellow Yellow Yellow Yellow Yellow

LC LC LC LLCT LLCT LLCT LLCT LC, MLCT LC, MLCT LLCT, MLCT LLCT, MLCT MC-S

36.6 74 95 – – 10.1 90.7 11 85 14.6 18.1 0.2

Zn(tppa)(ndc) K[Bi(tcbpe)(DMF)2]xDMF Zn2(tcbpe) In2(OH)2(tbapy) (frz connect.) In2(OH)2(tbapy) (scu connect.) [Zn2(tcbpe-F)(bpee)]xDMA [Zn6(btc)4(tppe)2(H2O)2]xH2O BiBr3(bp4mo)2 [BiBr4(bp4mo)]TBA+ [Mg[Ir(ppy)2(Hdcbpy)]2(H2O)2]3.5H2O [Mg[Ir(ppy)2(Hdcbpy)]2(DMF)2]3.5H2O [Dy4(btec)3(H2O)12]20H2O

365 455 420 405 405 455 400 352 352 468 468 306

553 553 550 525 525 530 540 540 540 544 554 480, 575

[42] [43] [44] [24] [24] [31] [45] [46] [46] [47] [47] [34]

Orange Orange

LLCT, MLCT LC

2.4 20.5

[Mg[Ir(ppy)2(Hdcbpy)]2(DEF)(H2O)]3H2O Zn4(tppa)2(sdc)3(NO3)2

468 365

575 580

[47] [42]

Red Red Red Red Red Red Red Red Red Red Red Red Red Red Red Red

MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S MC-S

7 7.8 8.4 8.6 10 11 20 22.5 25.3 38.5 40 43.9 50 53 59 75

[Eu(H2L4)(H2O)4]DMA+0.5, DMF, xH2O [Eu4(btec)3(H2O)12]20H2O Eu[Zn8(ad)4(bpdc)6O]2DMA+, 8DMF,11H2O Eu(ina)3(H2O)2 [InEu(btb)7/3(H2O)2]6DEF [Eu(L3)0.5(H2O)2]2H2O Eu(tcbb)(DMA)4.5CH3OH Eu2(ap)3(H2O)2(bpy) Eu3(OH)(ina)3(Hida)(ida)2 EuTb(ap)3(H2O)2(bpy) [Eu(btb)(H2O)]xDEF, xH2O Eu(phen)(ppmc)3 [Eu(ttb)(DEF)(H2O)]5DEF [Eu(tf-bdc)(NO3)(DMF)2]DMF Eu(hfa)3(dppy)2ZnCl2 [Eu2(m-bdc)3(phen)2]DMF

348 306 365 366 330 354 394 340 396 340 330 346 330 393 465 350

590, 590, 595, 595, 590, 590, 590, 590, 596, 590, 590, 590, 590, 593, 580, 585,

Blue ? Green Blue ? Green Blue ? Red Blue ? Red ? Green Blue ? White Blue ? White ? Yellow Green ? Yellow ? Red Yellow ? White Yellow ? White Yellow ? White Orange ? White

LC, LLCT LC, GC LC, GC LC, LLCT, LMCT – LC, GC MC-S LC, MC-S LC, MC-S LC, MLCT LC, LMCT

4 60.7 – 3 32 20.4 15.2 – – 10.9 –

[Cd2Na(ttca)2(Httca)2]5DMA+, 4DMF, 2H2O DSMy[In3(btb)4]3DMA+, 12DMF, 22H2O AFx[In3(btb)4]3DMA+, 12DMF, 22H2O [Pb2K(ttca)2Cl2]DMA+ Zn2Cl4(bpp)2 [Ir(ppy)2(bpy)]+x[(Cd2Cl)3(tatpt)4]12DMF, 18H2O [Tb4-4xEu4x(btec)3(H2O)12]20H2O [Gd0.24Eu0.5Tb0.26Cd2(imdc)2(Ac)(H2O)2]H2O La0.5Eu0.22Tb0.28Zn4(imdc)4(Him)4 [AgL5]xH2O Pb(NO3)(tzib)

260 ? 400 365 365 270 ? 410 254 ? 378 370 306 285 ? 360 294 ? 350 330 ? 355 320 ? 350

385, 495 420, 525 420, 625 381, 500, 619 410 ? 500 425, 540 485, 545, 585, 620 425, 480, 545, 590, 620, 700 430, 490, 545, 580, 615, 620, 700 427, 513, 566, 617 400, 600 ? 550

[51] [23] [23] [51] [52] [53] [34] [21] [22] [54] [55]

White

LC, LMCT

–

Pb(tzib)2

320 ? 385

380 ? 420, 525

[55]

495, 545, 545, 545, 545, 545, 545, 545, 545 544, 545, 545, 545, 544, 545, 545,

620, 620, 620, 620 614 620, 614 620 620, 620 614 615, 614 619, 615, 620,

545, 580, 580, 580, 580 580, 580, 575,

580, 615 615 620 615 620 615 615

584, 621 580 580, 620 580 580, 620 580 580

695 695 695

[31] [32] [21] [33] [34] [35] [33] [34] [36] [37] [38] [39] [40] [39] [27] [41] [39]

[33] [34] [32] [41] [48] [33] [48] [35] [49] [35] [48] [36] [48] [38] [50] [40]

695

705

700 686, 697 700 695

(continued on next page)

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

4

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Table 1 (continued) Emission Color

Emission Mech.

QY (%)

Formula

kex (nm)

kEM (nm)

Ref.

White White White White White White

LC, GC LC, GC LC, MC-S LC, MC-S MC-S LC, MC-S

15.2 17 – 32 46.2 47.3

[Ir(ppy)2(bpy)]+[Zn8(btca)6(2-NH2-bdc)3] AFxDSMy[In3(btb)4]3DMA+, 12DMF, 22H2O TbxEuy[Zn(H2thca)0.5(tib)]5H2O [Gd0.9977Tb0.0019Eu0.0004(m-bdc)3(phen)2]DMF Tb0.995Eu0.005(bptca) La0.6Tb0.3Eu0.1(bptca)

370 365 260 350 365 365

445, 420, 400, 450, 485, 440,

[56] [23] [57] [40] [58] [58]

570 525, 485, 485, 545, 485,

625 545, 545, 585, 545,

585, 585, 620, 585,

620 620, 695 645, 695 620, 645, 695

Emission colors separated by an arrow indicate that it is possible to tune the material’s emission between the listed colors. Excitation wavelengths separated by an arrow indicate that changing the excitation across the specified range will tune the emission color as indicated. Emission wavelengths separated by an arrow indicate that it is possible to tune the position of the indicated emission peak. The emission mechanisms listed are ligand centered (LC), guest centered (GC), metal centered (MC), metal centered following sensitization (MC-S), ligand–ligand charge transfer (LLCT), ligand–metal charge transfer (LMCT), and metal–ligand charge transfer (MLCT). All abbreviations given in the ‘formula’ column are listed following the acknowledgements section.

phosphors). LMOFs and LCPs provide a unique opportunity to create high quality, precisely tuned phosphors for applications in solid-state lighting technologies. This review will first present an overview of luminescence mechanisms and engineering strategies in LMOFs and LCPs, which will be followed by a discussion of recent progress in developing these materials. Table 1 below lists a number of representative LMOFs and LCPs and their properties as phosphor materials, including emission color and proposed mechanism, quantum yield where available, and excitation and emission wavelengths. 2. Luminescence mechanisms in LMOF/LCP phosphors

Fig. 2. Jablonski diagram showing the processes active in photoluminescence. Excitation (purple arrow) is followed by fast relaxation (blue arrow) to the S1 singlet state, which can emit a photon through fluorescence (green arrow), decay via a non-radiative process (dashed arrow) or undergo intersystem crossing (gray arrow) to the T1 triplet state, which can in turn emit a photon through phosphorescence (red arrow) or decay nonradiatively (dashed arrow).

physical properties can be well tuned by varying the identity of the metal ions, organic ligands, and overall connectivity of the framework. Additionally, their porous nature allows guest species to be encapsulated within them. As a result, functional MOF and CP materials have been reported for a wide variety of applications, including gas separation and storage [6,7], proton conduction [8,9], catalysis [10,8], and drug delivery [11,12], among others. This tunability also extends to their luminescent properties. Luminescence in MOFs and CPs can arise from the organic ligands, metal centers, guest species, or from processes that involve multiple structural components. Luminescent MOFs (LMOFs) and luminescent CPs (LCPs) have found promising applications in imaging [13,14], sensing [15–18], and optoelectronics [19,20], as well as in solid-state lighting (SSL) devices. LMOFs and LCPs are especially well-suited for lighting phosphors, as their excitation and emission properties can be carefully tuned by engineering appropriate metal centers, organic ligands, guest molecules, and structural connectivity. Constructing a framework provides an opportunity to orient and arrange absorber and emitter species for effective luminescence processes. Additionally, studies have shown that the luminescence efficiencies (e.g. internal quantum yield, IQY) of LMOFs/ LCPs that exhibit ligand-based emission can be significantly enhanced by rigidification or immobilization of the ligands to minimize molecular motions such as vibration and rotation that contribute to nonradiative decay. This is especially noteworthy in the case of ligands based on aggregation-induced emitters (AIE

Photoluminescence occurs when the absorption of light creates an excited electronic state, which then emits a photon as it decays back to a ground state. Photoluminescence can be broadly characterized as fluorescence or phosphorescence. In fluorescence, an electron is excited into some singlet excited state (Sn) and quickly relaxes into the lowest singlet excited singlet state (S1). Emission then occurs when the electron makes the spin-allowed transition from S1 to the ground state (S0) (Fig. 2). This process is fast, usually taking less than 1 microsecond to occur. In phosphorescence, an electron is again excited into some singlet state (Sn) and relaxes into the lowest excited singlet state S1—however, it then undergoes intersystem crossing into the triplet excited state (T1). Emission is associated with the spin-forbidden transition from T1 to the ground state (S0) (Fig. 2). As the emissive transition is spinforbidden, this process is slower, typically on the order of milliseconds or longer. However, as there is no exact cutoff in terms of the excited state lifetimes in fluorescence and phosphorescence, identifying which process is at work can be nontrivial. These emission processes also compete with non-radiative excitation decay processes; for example, transitions associated with vibrational states provide avenues for excited states to relax through releasing heat, rather than a photon (Fig. 2). Quantum yield is used to quantify the efficiency of photoluminescence processes. The internal quantum yield (IQY) of a material is the percentage of absorbed photons which are emitted, while the external quantum yield (EQY) is the percentage of incident photons which, following absorption, result in the emission of a photon from the phosphor [19]. The difference between the two is that the internal quantum yield describes only the efficiency of the emissive step in photoluminescence, while the external quantum yield provides the efficiency of the entire photoluminescence process, including the initial absorption of the photon. In LMOFs and LCPs, the composite nature of the frameworks can result in complex photoluminescence processes. Broadly speaking, LMOF and LCP luminescence can involve metal ions, ligand molecules, and guest molecules, and the process can either be localized on one species or result from charge transfer between multiple species (Fig. 3).

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

5

Fig. 3. Demonstration of the photoluminescence processes most commonly occurred in LMOFs and LCPs. The inner circle shows the excitation stage, with purple highlighting indicating the absorptive species. The outer circle shows the emission stage of photoluminescence, with yellow highlighting indicating the emissive species. Purple arrows indicate charge transfer or energy transfer processes between the absorber and emitter.

Fig. 4. Jablonski diagram showing the processes involved in sensitization, with excitation on the sensitizer/antenna species undergoing intersystem crossing from the S1 into the T1 state, followed by transfer into the acceptor species.

2.1. Localized luminescence processes Ligand-centered emission is a fairly common mechanism in LMOFs and LCPs, as the majority of MOFs and CPs are constructed on rigid ligand molecules, and the vast majority of rigid ligands are aromatic. In large, the extended conjugation of ligand molecules can often bring ligand HOMO/LUMO (highest occupied molecular orbital/lowest unoccupied molecular orbital) energy gaps into the visible range, potentially resulting in emission energies falling in

the visible range [59]. In LMOFs and LCPs with purely ligandcentered emission, there is neither orbital mixing between ligand molecules and the metal ions they are coordinated to nor energy transfer processes between ligand molecules. These frameworks can be considered to be arrays of single luminescence centers. The differences in luminescence between these frameworks and the bulk ligand are primarily related to the immobilization of the ligand molecules, as upon incorporation into a framework, certain molecular motions pertinent to the free ligand are prevented or otherwise limited. This in turn can lead to an increased quantum yield, as fewer non-radiative excitation decay pathways are available, as well as broadened emission peaks [60]. Metal-centered emission in LMOFs and LCPs generally stems from lanthanide ions within a framework. Lanthanide ion luminescence can produce emission covering the spectrum from UV to NIR emission—in the case of most trivalent lanthanide ions, this emission comes from a f–f electronic transition. Because only f orbitals participate in the transition, and because these orbitals are shielded by the closed 5s and 5p orbitals, these transitions are insulated from environmental and ligand effects; luminescence resulting from these transitions is therefore extremely sharp and characteristic for each lanthanide. However, as f–f transitions are parity-forbidden, most lanthanide ions have extremely poor absorption, resulting in weak emission under direct excitation [19]. Sensitization is an important process that can be used to circumvent this, in which excitation energy is absorbed by a ligand molecule and then transferred to a metal ion that would otherwise have weak absorption. This is also referred to as an antenna effect and enables strongly-emissive LMOFs and LCPs to be constructed which take advantage of lanthanide luminescence. In this process, initial absorbance occurs on the (antenna) ligand and is followed by intersystem crossing from the ligand singlet state into the

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

6

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

ligand triplet state. This in turn is followed by energy transfer from the ligand triplet state to the lanthanide triplet state (Fig. 4). The nature of the interaction between the antenna species and the lanthanide ion determines the type of energy transfer process that is active in the mechanism. For lanthanide ions which are bonded to antenna ligands, Dexter energy transfer mechanisms predominate, while Förster resonance energy transfers are responsible for most energy transfer for nonbonded antenna/emitter pairs [19]. Following the injection of excitation energy from the ligand triplet state into the lanthanide triplet state, characteristic La(III) emission occurs. Because the excitation event occurs on the ligand species, the parity-forbidden f–f transitions that are responsible for La(III) species’ poor absorption can be avoided, drastically increasing the emission from the lanthanide [61]. Guest-centered emission occurs when guest species present in the pores of the framework serve as luminescence centers within a LMOF or LCP. The identity of this guest species can vary widely from molecular organic phosphors, to fluorescent metal complexes, to metal ions or nanoparticles. These guests can either provide the entirety of an LMOF or LCP’s emission, or the guest’s emission energy can be chosen to compliment the emission from the framework, resulting in broad spectrum emission from the material [56,62]. These guests can be introduced into the framework in several different ways. In ‘‘ship in a bottle” synthesis, guest components are taken up into LMOF pores before self-assembly into an emissive guest species, while in ‘‘bottle around a ship” synthesis, guest species are present in the initial synthesis of the framework, and the framework forms around the guest species, incorporating them the formation of the framework. These two methods are more common for large, electrically neutral guests (molecular phosphors, metal nanoparticles), while ion exchange processes are more common for ionic guests [32,63]. In these, charge-bearing frameworks undergo ion exchange between the guest species and charge-balancing ions present in the framework’s pores. Additionally, guests whose size is smaller than the pore windows in a porous MOF can often be incorporated through diffusion-mediated processes, wherein the MOF is soaked in a solution containing the guest for some period of time, allowing the guest molecules to diffuse into the material [64]. 2.2. Charge transfer luminescence processes The previously described luminescence mechanisms focus on processes with absorption and emission on the same species. However, more complex emission mechanisms often present in LMOFs and LCPs, with absorption by one species followed by charge transfer to another species, with emission occurring on that secondary species. Additionally, emission can occur between molecular orbitals with contributions from more than one component of the framework. Ligand–ligand charge transfer (LLCT) emission mechanism involves luminescence from orbitals with contributions from multiple ligand molecules, or luminescence resulting from energy transfer from an absorber ligand to an emitter ligand. As ligands of LMOF and LCP are typically aromatic, frameworks which result in ligand stacking can exhibit orbital mixing between ligand molecules based on pi-pi interactions. If multiple ligands with different oxidation/reduction potentials participate in stacking, this can also lead to charge transfer emission processes. For ligand molecules with greater separation, Förster resonance energy transfer mechanisms can move excitation from one ligand to another [16]. However, this would depend on overlap between the emission and absorbance spectra of the two ligands, and that the ligands are oriented appropriately for dipole–dipole coupling. Ligand to metal charge transfer (LMCT), such as LLCT, can either result from orbital mixing between metal ions and ligand mole-

cules or from absorbance on ligands followed by energy transfer onto the metal, where emission occurs. In LMCT luminescence through orbital mixing, the LUMO is primarily based on orbitals from the metal ion, with the HOMO primarily composed of ligand orbitals. Transition metal ions with closed d subshells (i.e. d0 and d10 metals) are known to participate in LMCT luminescence, as the large gap between their HOMO and LUMO that comes from having a closed d subshell effectively prevents metal-centered emission processes [65]. However, as metal ions with closed-d subshells have high-lying LUMO energy levels, ligand molecules with similarly high-lying HOMO energy levels are generally required for LMCT in the visible range. As a result, a combination of closed-subshell metals that can be partially reduced, such as the d10 metals Zn(II) or Cd(II) used in combination with ligand molecules that can be easily/partially oxidized and have high-lying HOMO energy levels can often lead to LMCT luminescence [19]. In metal to ligand charge transfer (MLCT), transitions between mixed metal/ligand orbitals are responsible for luminescence, with the majority of the HOMO being contributed by metal orbitals, and the majority of the LUMO being contributed by ligand orbitals. Like LMCT, d10 metals can participate in MLCT as their large HOMO/ LUMO gap prevents metal-centered luminescence in the visible range. However, as the metal and ligand contributions to the framework HOMO/LUMO are inverted relative to LMCT, so too are the oxidation/reduction requirements. Metals that can be partially oxidized, such as Cu(I) and Ag(I) can participate in MLCT luminescence with ligands that have low-lying LUMO and can be partially reduced [66]. Mechanisms of host–guest charge transfer depend primarily on the nature of the guest species but follow the general patterns laid out in previous mechanisms. Dyes with conjugated or aromatic moieties can interact with the aromatic ligands of the framework through pi–pi interactions, resulting in Dexter energy transfer mechanisms which can transfer excitation energy between the host framework and guest species. Through careful selection of guest and ligand molecules with appropriate reduction/oxidation potentials, redox-mediated charge transfer can also play an important role in charge transfer, as can Förster resonance energy transfer over longer distances [67]. When the guest species is a metal ion, often a lanthanide, it can be chelated by free oxygen or nitrogen containing moieties in the ligand, or it can be located more generally in the pore [32,57]. As with the previously described sensitized LMCT luminescence, chelated lanthanides participate in luminescence through a Dexter energy transfer process, while free lanthanides participate in luminescence through Förster resonance energy transfer. For the purposes of this review, bound lanthanides with loading approaching 100% (full occupancy of the framework carboxylate/nitrogen moiety) will be considered heterometallic MOFs, while bound lanthanides present at lower loading levels will be considered host/guest MOFs.

3. LMOF/LCP phosphor design strategies When designing an LMOF or LCP phosphor material with specific emission characteristics, it can be challenging to work with systems in which radiative decay takes places across multiple species (LLCT, LMCT, MLCT). As multiple species are participating in the luminescence process, it is difficult to predict what the emission properties will be, unless the LMOF/LCP is based on a previously characterized material [63]. As such, one of the most reliable methods for producing new high-performance LMOF/LCP phosphors is to construct a material with ligand-centered emission from a highly emissive ligand. Upon rigidification in such a structure, the chromophoric ligand’s emissive properties (lifetime, quantum

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Fig. 5. The structures and HOMO-LUMO energy gaps for a series of AIE-based chromophoric ligands, estimated by DFT calculations. Reproduced from Ref. [44] with the permission of the Royal Society of Chemistry.

yield, etc.) are often improved, while the color of emission from the ligand is essentially maintained. The use of d10 metals such as Zn(II) and Cd(II) is ideal for enhancing emission in this fashion, as their low-lying HOMOs and high-lying LUMOs often prevent MLCT or LMCT from taking place. The only exceptions to this are ligands with sufficiently high-lying HOMO energy levels that can participate in LMCT with the metal LUMO, which also typically means ligands that are easily oxidizable [68]. Appropriate ligand design is crucial for preparing an LMOF with high performance ligand-centered emission. Recently, the highest performance LMOFs with ligand-centered emission have been prepared using ligands based on AIE chromophores [15,44,45], with quantum yields approaching unity. This is because immobilization of ligands within a MOF backbone effectively reduces or eliminates non-radiative decay due to various molecular motions of the freestanding ligands. If a new ligand molecule is being designed, density functional theory (DFT) calculations have been shown to be an effective method for estimating the emission energy of a given ligand molecule [44,45]. However, one should be aware that deprotonation, in the case of carboxylate-based ligands, and coordination with a metal may sometimes alter the electronic structure of a ligand [19]. As an example of this strategy was the design of the ligand H4tcbpe (1,1,2,2-tetrakis(4-(4-carboxyphenyl)phenyl)ethane, ligand L7 in Fig. 5) and corresponding LMOF phosphor [44]. A series of potential ligands based on AIE moieties was designed, and DFT calculations were used to estimate their HOMO-LUMO energy gaps and establish a structure–property relationship (Fig. 5). The H4tcbpe ligand was identified as having a HOMO-LUMO energy gap appropriate for yellow emission, and upon synthesis was indeed a bright yellow emitter. The ligand was incorporated into a framework using Zn2+, and the resulting LMOF maintained the yellow ligand-based emission while increasing its quantum yield to nearly 96%. While the effects on luminescence are less predictable, the use of Zr(IV) or other metals that make exceptionally strong metal– oxygen bonds are favored for the moisture and thermal stability of the resulting compounds. When used in combination with highly emissive chromophores, they can produce LMOFs and LCPs with strong MLCT or LMCT-based emission that also have exceptional stability [15,69–71].

7

Designing a lanthanide-based LMOF or LCP with given emission characteristics can be more straightforward, as the emission from trivalent lanthanide ions results from specific f–f transitions whose energies are typically independent of the environment around the lanthanide. Tm(III), Er(III), Ho(III), Sm(III), and Pr(III) have emission at various wavelengths in the visible range, but the inherent quantum efficiency of their emissive transitions is typically low, whereas emission from Eu(III), Tb(III), and Dy(III) is more efficient. Eu(III) emits in the orange-red region, with peaks corresponding to the 5D0–7Fn (n = 1, 2, 3, 4) transitions at approximately 590, 620, 650, and 695 nm, respectively [72]. Tb(III) emits in the yellowish green region, with peaks corresponding to the 5D4–7Fn (n = 3, 4, 5, 6) transitions at approximately 490, 545, 585, and 625 nm [41]. Dy(III) has two major emission peaks, with 4F9/2–6H15/2 and 4 F9/2–6H13/2 emitting at 480 nm and 575 nm, respectively [73]. It is important to note that while the energy of these transitions is fairly consistent, the Eu(III) 5D0–7F2, Tb(III) 5D4–7F5, and Dy(III) 4 F9/2–6H13/2 transitions are hypersensitive to the coordination environment and symmetry of the lanthanide in question, and their intensities can therefore vary drastically [74]. This can change the overall color of emission from the lanthanide, for example, tuning Eu(III) emission from orange to red, Tb(III) emission from green to greenish-yellow, and Dy(III) from blue to greenish-yellow. Lanthanide ions have poor absorbance under direct excitation, so it is necessary to include a ligand that is able to effectively sensitize the ion in question. For ideal sensitization, the energy level of the ligand triplet state should be approximately 3000–5000 cm1 higher than the lanthanide excited state [75]. If the energy gap between the ligand triplet state is too large, energy transfer is less efficient, while if it is too small, back transfer from the lanthanide to the ligand can occur, decreasing the efficiency of sensitization and therefore quantum yield [76]. As the emissive lanthanide excited state energy levels vary between lanthanides (for example: Sm3+, 4G5/2 = 18,021 cm1; Eu3+, 5D0 = 17,250 cm1; Tb3+, 5 D4 = 20,620 cm1; Dy3+, 4F9/2 = 21,000 cm1), no single ligand molecule will have the optimal triplet energy level for all lanthanide ions. Computational methods, such as time-dependent DFT (TD-DFT) can be used to assess ligand T1 energy levels [77]. White light emission is typically obtained by combining emission from multiple luminescent centers within a material. In the case of lanthanide-based materials, this often means the inclusion of green-emitting Tb(III) and red-emitting Eu(III) in combination with a framework that emits in the blue region. Achieving white light emission relies on balancing the concentration of the lanthanide ions present in the structure. The most common strategy is to synthesize an LMOF with blue ligand-based emission coupled with a non-emissive lanthanide such as La(III) or Gd(III) [78]. These metals are used to prevent the extinction of ligand-centered emission that is typically the case in lanthanide-based LMOFs and LCPs with effective sensitization. Eu(III) and Tb(III) are then doped into the structure during synthesis at varying ratios until the proper blend of red, green, and blue emission is obtained. White emission from rare-earth-free LMOFs and LCPs is likewise achieved by combining emission from multiple sources. However, the design of these LMOFs and LCPs is more flexible, as they aren’t reliant on the two or three lanthanide ions with acceptably high quantum yields and unchanging emission colors. As such, instead of combining red, green, and blue emission, white light is more often achieved by combining blue and yellow emission. This can occur in a number of ways. One strategy takes advantage of the encapsulation of a yellow-emitting guest within a blue-emitting framework [53,79]. In this case, the quality of the white light can be controlled by altering the concentration of the guest within the framework. Another approach relies on frameworks with multiple simultaneous emission processes [55,80]. Blue emission is provided by the ligand, while yellow emission is provided by some

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

8

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

charge transfer process between the excited ligand and another species, which could be another ligand or a metal node. As concentration-based tuning isn’t an option, controlling the quality of the white light relies on the two emission mechanisms having differing excitation spectra. As the excitation energy is tuned from one into the other, the emission from the two sources can be balanced. 4. Rare earth metal based LMOFs/LCPs Rare-earth metals (REEs) have unique luminescence behaviors. Their characteristic emission wavelengths make them attractive chromophores for phosphor materials, but their poor absorption requires carefully designed frameworks to reach their full potential. The following section will discuss various examples of REE based LMOFs and LCPs, their emission mechanisms, and design strategies for taking full advantage of their characteristic emissions. 4.1. Lanthanide LMOFs/LCPs with colored emission The first lanthanide-based LCP with noteworthy luminescence properties was reported in 1999 by the W.B. Lin group at Brandeis University [41]. They synthesized a series of onedimensional lanthanide isonicotinate (ina) CPs with the general formula Ln(ina)3(H2O)2 using the lanthanides Ce(III), Pr(III), Nd (III), Sm(III), Eu(III), and Tb(III) to create samples 1a–f. All lanthanide atoms were eight coordinate, with six carboxylate oxygen and two terminal water molecules forming square antiprismatic geometry. Photoluminescent emission data were given for 1e [Eu(III)] and 1f [Tb(III)], with each structure showing characteristic lanthanide emission under 366 nm excitation. Eu(III)-containing 1e showed red emission with peaks at 595 and 620 nm, corresponding to 5D0–7Fn (n = 1, 2) transitions, with a quantum yield of 8.6%, while Tb(III)-containing 1f showed green emission with peaks at 490, 545, 585, and 625 nm corresponding to 5D4–7Fn (n = 6, 5, 4, 3) transitions and a much higher quantum yield of 90.0%. These quantum yields are significantly higher than would be expected for direct excitation of the lanthanide ions, which suggests that sensitization of the lanthanides may be taking place.

Cahill et al. reported cross-lanthanide sensitization in their 2007 article—a first for lanthanide-based LCPs [35]. They prepared two new frameworks with the general formula Ln2(ap)3(H2O)2(bpy) (ap = adipate, bpy = 4,4-bipyridine), with Ln2 clusters linked by adipate ligands into a 2D sheet, and bpy serving as non-coordinating templating and sensitizing agents in the pores (Fig. 6). Framework 1 was composed of 50:50 Eu(III) and Tb(III), while framework 2 was purely Tb(III). The luminescent properties of these two compounds were also compared with an isostructural version with only Eu(III) that the authors had previously reported [81]. In the pure Tb (III) and pure Eu(III) compounds, characteristic emission was observed at relatively high quantum yields of 31.6 ± 7.5% and 22.5 ± 2.1%, respectively, following excitation at 285 nm. This emission was accomplished by sensitization of the lanthanides by the non-coordinating bpy ligands through a FRET-mediated process. However, for the mixed Tb(III)/Eu(III) sample, emission was almost entirely Eu-based, with the characteristic 5D4–7Fn Tb(III) peaks nearly disappearing while the quantum yield increased to 38.5 ± 2.1% (Fig. 6). The drastic decrease in emission from Tb(III) concurrent with an increase in emission from Eu(III) indicates that the Tb(III) ions sensitize emission from the Eu(III) ions. Reddy et al. used two new ligands 3,5-bis(benzyloxy)benzoic acid (HL1) and 3,5-bis(pyridine-2-ylmethoxy)benzoic acid (HL2) in combination with Eu(III), Tb(III), and Gd(III) to create six new lanthanide CPs [76]. Structures 1, 3, and 5 are composed of Eu(III), Tb(III), and Gd(III) in coordination with L1, while 2, 4, and 6 are composed of Eu(III), Tb(III), and Gd(III) in coordination with L2. This set of lanthanides—Eu(III), Tb(III), and Gd(III)—are commonly employed when investigating luminescence mechanisms in LMOFs and LCPs because of their divergent luminescence behaviors. Eu(III) and Tb(III) are both strongly luminescent under sensitization, but their emission wavelengths and excited state energy levels are different, with Eu(III) having lower energy excited states and emitting in the red, while Tb(III) has higher energy excited states and emits primarily green light. Gd(III) typically does not participate in luminescence, as its excited state energy is high, and so it can be used as a sort of control to identify the framework behaviors when not participating in sensitization. Sivakumar et al. exploited this ability and were able to quantify not only the total quantum yield for their structures but also the individual efficiencies of the energy transfer step in sensitization and the efficiency of the luminescence of the sensitized lanthanides. They show that

Fig. 6. (a) Structure of framework 1 (left) viewed down the [0 1 0] direction, with the arbitrarily colored polyhedra representing TbO9 and EuO9 clusters, black lines representing adipate, and ball and stick representing bpy (black = C, blue = N). (b) Emission spectra of the Tb(III), Eu(III), and Tb(III)/Eu(III) systems (right), with the transition corresponding to each peak labeled. Reprinted with permission from Ref. [81]. Copyright 2007, American Chemical Society.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

appropriate energy matching of the L1 ligand T1 excited state and the Tb(III) states permits more efficient sensitization (60%) than the L2 ligand with Tb(III) (27%). The same group also produced four new lanthanide CPs to study the sensitization process, with Eu(III), Tb(III), and Gd(III) coordinating with the ligand 4-(dipyridin-2-yl)aminobenzoic acid (HL) to produce frameworks 1, 2, and 3. A 50:50 Eu(III):Tb(III) (4) was also prepared [82]. They also were able to quantify the overall quantum yield, sensitization efficiency, and emission efficiency for all compounds, finding that better energy level matching led to better performance. In 4, they observed sensitization of the Eu(III) by the Tb(III), with 86% efficiency. De Andres, Monge and coworkers reported three new isostructural lanthanide LMOFs with Eu(III), Gd(III), and Tb(III) complexing with the ligand 3,5-disulfobenzoate (dsb) to form materials with the formula [Ln7(dsb)4(OH)9(H2O)15]4H2O, which possess heptanuclear [Ln7(OH)9]12+ SBUs [25]. The Eu(III) and Tb(III) ions are effectively sensitized by the dsb ligands. Additionally, when structures are prepared with a few percent of Eu(III) or Tb(III) doped into the Gd(III) framework, dramatic increases in characteristic Eu(III)/Tb(III) emission are noted (Fig. 7). While this may seem initially to be evidence for concentration quenching in the pure Eu(III) and Tb(III) compounds, the lifetimes of the pure and diluted samples are nearly identical, showing that this is not the case. Instead, the authors attribute the increase emission to a higher ratio of sensitizing ligands:emissive lanthanides. In 2014, the Hong group reported six new lanthanide coordination polymers based on Eu(III), Tb(III), and Gd(III) in combination with L1 (p-terphenyl-2,200 ,2000 ,5,500 ,5000 -hexacarboxylate acid) and L2 (p-terphenyl-3,200 ,300 ,5,500 ,5000 -hexacarboxylate acid) [33]. These ligands were used for their triplet state energy levels, 24,000 cm1 (L1) and 21,200 cm1 (L2), which are located appropriately above the emissive states of Eu(III) (17,500 cm1, 5D0) and Tb(III) (20,500 cm1, 5D4), allowing for sensitization to occur. The reported structures had the formulas [Ln(L1)0.5(H2O)2]2H2O, with Ln being Eu3+ (1), Tb3+ (2), and Gd3+ (3), and [Me2NH2] [Ln(H2L2)(H2O)4]0.5DMFxH2O, with Ln being Eu3+ (4), Tb3+ (5), and Gd3+ (6). They found the highest overall quantum yield in 2 (32%) and 5 (25%), which is attributed to the more optimal energy gap between the ligand triplet state and Ln excited state. In the case of 1 (11%) and 4 (7%), the lower quantum yield is ascribed

Fig. 7. Emission spectra of Eu-dsb LMOF (black) and emission spectra of Eu0.2Gd0.98 (red) under 270 nm excitation. Reproduced from Ref. [25] with the permission of the Royal Society of Chemistry.

9

to an increased susceptibility of the Eu(III) excited state to be deactivated by nearby OH vibrations, as the gap between the emissive and ground levels is smaller in Eu(III) than Tb(III). This results in stronger coupling with the third vibrational overtone of the OH oscillator (mOH  3300–3500 cm1) of coordinated water molecules. The strong response of Eu(III) emission to the OH oscillator has been reported in LMOFs that also have applications in proton transport [83,84]. Increased water content within the LMOFs was responsible for more than fivefold decreases in luminescence intensity, emphasizing the importance of limiting the presence of water in lanthanide and specifically europium, LMOFs. A lanthanide LMOF with tunable emission color ranging from green, through yellow, and into red was reported in 2015 by Li, Zang, Mak et al., with the tuning taking advantage of the previously discussed sensitization of Eu(III) by Tb(III), and accomplished via careful codoping of Tb(III) and Eu(III) into an LMOF [85]. They reported d three series of lanthanide coordination polymers, [Ln(L) (H2O)2]NO32H2O (Ln = La 1, Pr 2), [Ln2(L)2(NO3)(H2O)2]Cl6H2O (Ln = Nd 3), [Ln(L)(H2O)2]Cl3H2O (Ln = Sm 4, Eu 5, Gd 6, Tb 7, Dy 8, Ho 9, Er 10, Tm 11, Yb 12, and Lu 13) (H3L = 4-carboxy-1(3,5-dicarboxy-benzyl)-pyridinium chloride), with 1, 4, 5, 7, and 8 giving characteristic emission of green, pink, red, green, and yellow light, respectively. However, the excitation peaks for all of these compounds are well outside of the ligand absorbance (two bands centered q 235 nm and 265 nm), indicating that there is no effective sensitization of these lanthanides. Nonetheless, the emission of 1 can be selectively tuned from green into yellow, orange, and red by doping Eu(III) into the Tb(III) LMOF (Fig. 8). As shown in Fig. 8, as the concentration of and emission from Eu(III) slowly increases, emission from the Tb decreases, until it is neatly extinct at 56.5% Eu (III), indicating energy transfer from Tb(III) to Eu(III). A similar series of isostructural color-tuned LMOFs was reported in 2016 by the O. Guillou group, with Tb/Eu codoping into an LMOF with the formula [Ln4(btec)3(H2O)1220H2O] (btec = benzene-1-2-4-5-tetracarboxylate) providing an avenue to tune emission from green through red, and codoping with Gd(III) used to increase brightness [34]. In addition to Tb, Eu, and Tb/Eu, LMOFs with Sm(III), Gd(III), and Dy(III) were also prepared. Unlike the previous work, emission from the lanthanides was partially sensitized, as excitation spectra monitored at characteristic lanthanide emission wavelengths showed broad ligand absorption bands in addition to direct lanthanide absorption (Fig. 9). However, the similar intensity of the direct and sensitized excitation regions indicates that the sensitization is somewhat incomplete, with sensitization of Tb(III) being most effective at 81% efficiency, followed by Eu(III) at 66%. This also resulted in the Tb-btec having the highest quantum yield of 31.1%, with Eu-btec’s quantum yield being 7.8%, Dy’s quantum yield being 0.2%, and Sm’s being 0.1%. The significantly lower quantum yield for Eu-btec, despite its similar sensitization efficiency to that of Tb(III), is attributed to greater susceptibility to solvent-induced relaxation. The authors used a heterometallic strategy to tune brightness and color, with the inclusion of an inactive metal [Gd(III)] used to increase brightness by limiting concentration quenching. In both Tb/Gd-btec and Eu/Gd-btec LMOFs, inclusion of 20% Gd was able to increase quantum yields by approx. 30%. The combination of green-emitting Tb(III) and red-emitting Eu(III) was used to tune emission from green through red (Fig. 9). As demonstrated by the work discussed above, deactivation of luminescence through coupling with solvent vibrations significantly limits the quantum yield of lanthanide LMOFs—especially those that contain Eu(III). To avoid this, Cepeda and Castillo et al. used a solvent-free approach to synthesize a series of Lnpyrimidine-4,6-dicarboxylate (pydc) CPs [86]. Solvent-free melt reactions between lanthanide(III) nitrates and pydc produced three dimensional CPs with the formulas Ln(l4-pmdc)(NO3)(H2O) (Ln1)

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

10

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Fig. 8. (left) Emission spectra from 1 doped with progressively increasing concentrations of Eu(III), showing the decreasing emission from the 5D4–7Fn transitions in Tb(III) (490 nm, 545 nm, 585 nm) and increasing emission from the Eu(III) transitions 5D0–7Fn (595 nm, 620 nm). (right) CIE chromaticity diagram for the Tb(1x)Eux(L) series, with (a) x = 0%, (b) 1.46%, (c) 2.64%, (d) 14.79%, (e) 19.49%, (f) 35.46%, (g) 56.50%, (h) 100%. Reprinted with permission from Ref. [85]. Copyright 2015, American Chemical Society.

Fig. 9. (left) CIE chromaticity diagram showing the color of emission from the Tb1xEux-btec series, with images of the 11 samples with CIE coordinates shown. (center) Excitation and emission spectra for Tb-btec, showing ligand-based excitation at 300 nm and direct excitation of the Tb(III) between 350 and 400 nm. (right) Excitation and emission spectra for Eu-btec, showing ligand-based excitation at 300 nm and direct excitation of the Tb(III) between 350 and 400 nm. Reprinted with permission from Ref. [34]. Copyright 2016, American Chemical Society.

(Ln = La, Ce, Pr, Nd, Sm, and Eu) and [Ln(l4-pmdc)(NO3)(H2O)]H2O (Ln2) (Ln = Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu). In both cases, nitrate group ligands limit the coordination of water, allowing only one H2O to coordinate to the metal. In both the Ln1–Eu and Ln2–Tb, sensitization by the ligands is incomplete, as can be inferred from the excitation spectra, which show direct lanthanide excitation competing with ligand excitation. However, despite the limited sensitization, the quantum yields of Ln1–Eu and Ln2–Tb are 20% and 18%, respectively, which is higher than the previously discussed Eu(III) quantum yields and similar to those previously discussed for Tb(III). In addition to eliminating quenching from OH-vibrations, increasing quantum yields of LMOFs with framework-based emission has also been achieved by eliminating quenching from CH vibrations [87]. Ruschewitz et al. demonstrated that this is an effective strategy for improving lanthanide LMOF luminescence as well, using the perfluorinated ligand 2,3,5,6-tetrafluorotereph thalate (tfBDC) to produce 10 new coordination polymers with

the general formula [Ln(tfBDC)(NO3)(DMF)2]DMF for Ln = Eu3+ (1), Gd3+ (2), Tb3+ (3), Ho3+ (4), Tm3+ (5), [Ln(tfBDC)(CH3COO) (FA)3]3FA for Ln = Sm3+ (6), Eu3+ (7) and [Ln(tfBDC)(NO3)(DMSO)2] with Ln = Ho3+ (8), Er3+ (9) and Tm3+ (10) [38]. Compound 1, or EutfBDC, showed an impressive quantum yield of 53% following removal of adsorbed water by heating at 60 °C for one hour. Compound 3, or Tb-tfBDC, was measured to have a quantum yield of 67%. 4.2. Lanthanide LMOFs/LCPs with white emission In the previously discussed works, combined emission from multiple types of lanthanide ions was used to create emission ranging from green into red, depending on the relative contributions of Tb(III) and Eu(III) to the framework’s emission. Examining the CIE chromaticity diagrams in Figs. 7 and 8 shows a linear track from green into red as the composition of the of Tb(III) and Eu(III) in the material changes. The addition of blue light would pull that

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

11

Fig. 10. Excitation (inset) and emission spectra of the La:Eu:Tb sample with doping concentrations of 50:22:28 (left), and the corresponding CIE chromaticity diagram (right) under (a) 350 nm and (b) 294 nm excitation. Reproduced with permission from Ref. [78]. Copyright 2012, American Chemical Society.

track down into the white area of the diagram, tuning along the blue-yellow axis. This is the strategy pursued by Zheng et al. to prepare a white-emitting LMOF [78]. They reported a series of five isostructural lanthanide-zinc-organic frameworks with the formula LnZn4(imdc)4(Him)4 (Ln = La (1), Pr (2), Eu (3), Gd (4), and Tb (5); H3imdc = 4,5-imidazoledicarboxylic acid, Him = imidazole), and the photoluminescence properties of 1, 3, and 5 were investigated. As the La(III) in 1 does not have a f–f transition available to it, it does not participate in luminescence, and the 430 nm emission observed for 1 under 354 nm excitation was attributed to ligand-centered fluorescence processes, as both ligands emit at 425 nm under the same excitation energy. Compound 3 emits characteristic Eu(III) emission in the red region under 289 nm excitation, while 5 emits characteristic Tb(III) emission in the green region under the same excitation. In both cases, blue emission from the ligand was completely absent, indicating effective sensitization. In order to combine the blue framework emission with the red Eu(III) emission and green Tb(III) emission, Eu(III) and Tb(III) were doped into the La-containing framework 1 at varying concentrations. These codoped materials possessed excitation peaks at 294 and 350 nm excitation when monitored at characteristic Eu (III) or Tb(III) wavelengths, while monitoring at the framework emission wavelength of 430 nm returned only the excitation peak at 350 nm, indicating that 294 nm emission effectively sensitized emission from the lanthanides, but that 350 nm emission only partially sensitized the lanthanides, allowing framework emission to compete with energy transfer (Fig. 10). A sample with a La:Eu:Tb concentration of 50:22:28 produced yellow light when excited at 294 nm, but as the excitation wavelength was tuned toward 350 nm, framework emission was established, producing high quality white light at CIE coordinates of (0.328, 0.330)—ideal white light is defined as (0.33, 0.33)—when excited at 350 nm (Fig. 10). A similar paper was simultaneously published by Chen and Qian et al. They reported a new two dimensional lanthanide coordination polymer ZJU-1 [La2(pda)3(H2O)5] (pda = pyridine-2,6dicarboxylic acid) which exhibits ligand centered emission at 408 nm under 312 nm excitation [88]. Upon progressive doping with Tb(III) under the same excitation energy, emission can be tuned from the blue into the green region, resulting from the blending of ligand-centered emission with Tb(III) characteristic emission, and its main peak at 543 nm (Fig. 11). Likewise, progressive doping with Eu(III) can tune emission from blue through purple and toward red, as characteristic Eu(III) emission and its main

peak at 614 nm blend with blue ligand-centered emission (Fig. 11). At La:Tb:Eu concentrations of 96.5:1.5:2.0, white light with CIE coordinates of (0.3109, 0.3332) is produced with a quantum yield of 6.80%, while La:Tb:Eu concentrations of 97:1:2 produce white light with CIE coordinates of (0.3269, 0.3123) and a quantum yield of 6.1% (Fig. 11). Sun et al. demonstrated that white light emission from heterometallic lanthanide LMOFs is also possible without relying on ligand emission for the blue component. They reported a series of 14 isostructural, anhydrous metal–organic frameworks based on the semiflexible ligand 1,3-di-(4-carboxyphenoxy)-2-(4-carboxy phenoxymethyl)-2-methylpropane (H3L) in combination with Y, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, or Yb (all lanthanides excluding Sc and the radioactive Pm), and used combinations of Dy/Eu or Dy/Sm codoped into the Gd framework to produce white light [89]. In all structures, the lanthanide ion is nine-coordinated by oxygen from six carboxylate groups with three oxygen bridging each lanthanide with its neighboring lanthanides to form a 1d chain along the c axis, with each ligand bridging three chains to form a 3D framework. To create a white phosphor, the GdL was used as the host framework, with Dy(III) doped in to provide both blue and yellow luminescence from its characteristic 4F9/2–6H13/2 transition (480 nm) and 4F9/2–6H15/2 transition (572 nm), which is permitted because the Dy(III) occupies a site with low symmetry and no inversion center [90]. Either Sm(III) or Eu(III) was codoped with the Dy(III) to provide red emission, resulting in white light emission from the combined metals under 294 nm excitation. The CIE chromaticity coordinates for Dy0.02Eu0.05Gd0.93L are (0.355, 0.313), while those for Dy0.01Sm0.10Gd0.89L are (0.328, 0.320). None of the prepared compounds showed any ligandbased emission, indicating that they were all effectively sensitized. Additionally, it should be noted that samples of GdL doped with the individual metals Eu(III), Sm(III), and Dy(III), and LaL doped with Tb(III) were prepared, which all showed characteristic emission from the dopant metal. All four samples showed at least several-fold QY enhancement over their native EuL, TbL, SmL, and DyL frameworks, with DyL showing the most significant enhancement from 1.2% in DyL to 18.2% in Dy0.02Gd0.98L. This is likely a result of the 1D chain structure leading to concentration quenching in the native frameworks. The quantum yield of Tb0.08La0.92L is also noteworthy, as quantum yield of 79.9% (increased from 38.6% in the native framework) was among the highest reported for any lanthanide LMOF at the time.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

12

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Fig. 11. (a) Emission spectra of ZJU-1 with progressively increasing concentrations of Tb(III) under 312 nm excitation. (b) CIE chromaticity diagram demonstrating the tuning of ZJU-1 emission from blue into green under progressive doping with Tb(III) – images of the material are inset. (c) Emission spectra of ZJU-1 containing progressively increasing concentrations of Eu(III) under 312 nm excitation. (d) CIE chromaticity diagram demonstrating the tuning of ZJU-1 emission from blue into purple and red under progressive doping with Eu(III), with images of the material inset. (e) Emission spectrum of ZJU-1:1.0% Tb(III), 2.0% Eu(III) under 312 nm excitation. f) CIE chromaticity diagram showing the positions of the white light emitted by ZJU-1:1.0% Tb(III), 2.0% Eu(III) (a) and ZJU-1:1.5% Tb(III), 2.0% Eu(III) (b). Reproduced from Ref. [88] with the permission of the Royal Society of Chemistry.

While the previous works relied on ligand doping to provide white emission, Su et al. demonstrated white light emission from a single-phase lanthanide network [91]. The new bifunctional ligand L (tris((pyridin-3-ylmethyl)benzoimidazol-2-ylmethyl)ami ne) possesses coordination sites with discernable coordination environments; each ligand contains a tripodal benzimidazolyl moiety and three monodentate pyridine moietie. Direct reaction of L with Eu(ClO4) or Gd(ClO4) produces monomeric clusters [EuL2] (ClO4)32.5MeCN (1-Eu) and [GdL2](ClO4)32MeCN2CHCl3 (1-Gd).

These clusters form with the lanthanides coordinated to the central amine and benzoimidazol moieties of two ligands, with the pyridine moieties oriented outwards. Upon reaction with AgClO4, these clusters are linked into a three dimensional framework, as each Ag (I) ion is bridgind between two pyridine groups from different clusters. This reaction gives [EuAg3L2(H2O)(MeCN)](ClO4)64MeCN (2-Eu-Ag) and [GdAg3L2(H2O)(MeCN)](ClO4)64MeCN (2-Gd-Ag). The lack of ligand emission in the emission spectra of 1-Eu demonstrates that the ligand is an effective sensitizer of Eu(III) emission

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Fig. 12. Emission of (a) H3L, (b) 1-Gd, (c) 2-Gd-Ag, (d) 1-Eu, (e) 2-Eu-Ag under 310 nm excitation, with the CIE chromaticity diagram inset above, and an emission photograph of 2-Eu-Ag under 310 nm excitation with a 450 W xenon lamp. Reprinted with permission from Ref. [91] Copyright 2012, Journal of the American Chemical Society.

(Fig. 12). However, upon the formation of the framework 2-Eu-Ag, some ligand fluorescence is recovered, while the intensity of emission from Eu(III) increases (Fig. 12). As the emission from Eu(III) was not negatively affected, it appears that the Ag(I) does not interfere with sensitization of the lanthanide by the ligand but instead increases the efficiency of all ligand luminescence processes. Given that the ligand-based fluorescence also increased in 2-Gd-Ag (Fig. 12), where sensitization of the lanthanide was not occurring, the authors attribute this increase to the rigidification of the pendant pyridyl groups. A mixed lanthanide LMOF with white light emission with a relatively high quantum yield was reported by Liu and Zheng et al. in 2013, who developed a series of isostructural lanthanide LMOFs using the ligand H3BTPCA (1,10 ,100 -(benzene-1,3,5-triyl)tripiperi dine-4-carboxylic acid) in combination with Eu(III), La(III), Tb(III), and Sm(III) [58]. La-BTPCA has broad blue emission from the ligand p–p⁄ transitions centered at 445 nm under 365 nm emission, while Eu-BPTCA and Tb-BPTCA exhibited characteristic emission in the red and green regions with no simultaneous ligand emission, indicating effective sensitization by the ligand. As the LMOFs were isostructural with red, green, and blue emission, the authors sought to use the codoping strategy to prepare a white-emitting LMOF, finding that La0.6Tb0.3Eu0.1-BTPCA produced white light with CIE coordinates of (0.3161, 0.3212) and quantum yield of 47.33% under 365 nm excitation. A strategy to tune the energy transfer from sensitizing ligands to Ln metal centers in emissive lanthanide LMOFs was recently developed by Li and Yan et al. and used to produce white light without relying on mixed-metal doping [55]. By altering the size of the ligand, they showed that it is possible to systematically alter the efficiency of sensitization. They report four isoreticular onedimensional Ln(III) CPs based on the ligands 2-(1,8naphthalimido)ethanoic acid (HL1) or 3-(1,8-naphthalimido) propanoic acid (HL2) in coordination with Eu(III) or Gd(III). In both ligands, the 1,8-napthalimido moiety serves as the luminescence center, with the only differences between the ligands being the length of the carbon chain linking the 1,8-napthalimido moiety with the lanthanide-coordinating carboxylate (two carbons in L1, three carbons in L2). Gd–L1 and Gd–L2 both display ligandcentered emission with broad peaks at 475 which cover much of the blue-green region. Eu–L1 exhibits pink emission, with characteristic Eu-centered emission mixing with a small amount of emis-

13

sion from the ligand. However, Eu–L2 emits white light with 11.9% quantum yield, as the increased distance between the absorbing ligand moiety and emissive Eu(III) decreases the efficiency of the energy transfer, resulting in relatively stronger blue-green ligand emission and weaker Eu(III) red emission. In addition to inclusion in metal nodes, emissive lanthanide ions can be introduced into LMOFs via post-synthetic modification. Yan et al. used the UiO-type MOF UiO-67-bpydc (1, [Zr6O4(OH)4 (bpydc)6] bpydy = 2,20 -bipyridine-5,50 -dicarboxylate) as a base framework, as the 2,20 -bipyridyl moiety could effectively chelate lanthanide ions [92]. They prepared three samples containing Eu3+, Tb3+, and Eu3+/Tb3+ by exposing 1 to solutions of the lanthanide chloride salts in methanol at 60 °C, producing (Eu@1, Tb@1, and Eu/Tb@1). As UiO-67-bpydc emitted in the blue-green region, Eu@1 was investigated for white light emission. Under 340 nm excitation, emission is entirely based on the characteristic Eu(III) transitions (Fig. 13). However, as the excitation wavelength is tuned toward the blue, the efficiency of sensitization drops, allowing blue-green framework emission to mix with red Eu(III) emission and producing white light with CIE coordinates of (0.3481, 0.3292) and a quantum yield of 14.34%. A WLED device was also prepared using Eu@1, which was coated onto a UV LED chip using a methanol dip-coating method (Fig. 13). The resulting device had a correlated color temperature (CCT) of 4919 K, color rendering index of 75, and luminous efficiency of 32 lm W1. 4.3. Actinide LMOFs/LCPs Though the bulk of this review has focused on lanthanide rareearth LMOFs and LCPs, there are several examples of actinide LMOFs and LCPs with interesting luminescent properties as well. Unlike lanthanides, actinide excitation does not require f–f transitions, and so they rely less on sensitization. Shun et al. reported a heterometallic uranyl-silver CPs, with mixed Ag-UO2 SBU with green metal-centered emission from the S11-S00, S11-S01, S11-S02, S11-S03, and S11-S04 transitions on U(VI) [93]. The LMOFs were constructed from Uranyl-bpdc (2,20 -bipyridine-3,30 -dicarboxylate) helixes, both right- and left-handed, bridged by Ag+ ions. Because of the hardness of the uranyl cations, the uranyl only coordinated with the bpdc carboxylate groups, allowing the pyridyl moieties of the ligands to coordinate with the bridging Ag+ ions. Uranyl-based LCPs with U-centered emission processes were also reported by Burns et al. [94] A 1D uranyl-2,20 -bipyridine (bpy) coordination polymer was prepared under hydrothermal conditions. The pentagonal bipyrimidyl uranyl cations form an edge-sharing 1D chain held together through cation-cation interactions, with the coordinating bpy ligands surrounding the central uranyl chain, limiting its dimensionality (Fig. 14). The emission is again centered on the uranium; however, the presence of the bpy ligand significantly alters the strength of the uranium transitions, depressing higher energy peaks and enhancing lower energy peaks (Fig. 14). While uranium typically possesses green emission, the 1D LCP produced orange emission, with main peaks at 540, 580, and 620 nm under UV excitation. Sun et al. similarly reported a luminescent uranyl-based LMOF CPP-U4 [UO2(L)(bbi)0.5, L = (2-carboxyethyl)(phenyl)-phosphinic acid, bbi = 1,10 -(1,4-butanediyl)bis(imid-azole)] [95]. CPP-U4 is a three-dimensional structure with 2D layers composed of U linked by the ligand L, pillared by the ligand bbi. Like the previously described uranyl LCP, the presence of ligand molecules alters the uranium-centered emission. However, unlike the previously described LCP, CPP-U4 does not influence the peak strength but instead affects the transition energy. The green emission from CPP-U4 is redshifted by 12 nm relative to the original UO2(OAc)22H2O (Fig. 15), without having any significant impact on relative peak intensity.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

14

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Fig. 13. (a) Emission spectra of Eu@1 under various excitation conditions, with intensity normalized to the 5D0–7F2 transition of Eu(III) at 620 nm. (b) CIE chromaticity diagram showing the color of light emitted from Eu@1 excitation by 400 nm (rightmost), 390 nm, 380 nm, 370 nm, 360 nm, 350 nm, and 340 nm. (c) Photographs of the UVLED chip before in the ‘off’ state before (i) and after (ii) being coated with Eu@1, and the same chip in the ‘on’ state (iii), emitting white light. Reproduced from Ref. [92] with the permission of the Royal Society of Chemistry.

Fig. 14. Comparison of emission spectra from UO2(NO3)2.6H2O (red) and 1D uranyl-2,20 -bipyridine (bpy) coordination polymer under 365 nm excitation. Reproduced with permission from Ref. [94]. Copyright 2012, American Chemical Society.

Fig. 15. Comparison of emission spectra from CPP-U4 and UO2(OAc)22H2O under 365 nm excitation. Reproduced with permission from Ref. [95]. Copyright 2014, Wiley-VCH Verlag GmbH & Co. KGaA.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Two new 3D thorium-based LMOFs were reported by Liu and Wang et al. Compound 1 ([Th(TPO)(OH)(H2O)]8H2O, TPO = tris-(4 -carboxylphenyl)phosphine oxide) was synthesized in hydrothermal conditions, while 2 ([C9H17N2][Th(TPO)Cl2]18H2O) was synthesized in ionothermal conditions [96]. Although Th(IV) is typically nonemissive, the reported LMOFs displayed broad bluegreen emission with peaks at 450 nm and 500 nm, respectively. The authors believe that the emission was based a ligand-tometal charge transfer process. Other actinide-based LMOFs with interesting luminescence properties have been reported. Reger et al. studied uranyl and thorium-based 1D CPs using a new ligand with a fluorescent 1,8naphthalimide moiety [97]. While the thorium-based LMOF demonstrates pure ligand-centered emission, the uranyl MOF is non-emissive, which the authors attribute to charge transfer relaxation. Two threefold interpenetrated uranyl-based LMOFs were also reported by Sun and coworkers which possess ligandcentered emission only [98]. Both LMOFs (UOF-1 and UOF-2) were synthesized with extremely flexible ligands (UOF-1, 4,40 -[[2-[(4-c arboxyphenoxy)ethyl]-2-methylpropane-1,3-diyl]dioxy]dibenzoic acid; UOF-2, hexakis[4-(carboxyphenyl)oxamethyl]-3-oxapen tane). The flexibility of these ligands enabled interpenetration of the structures, which represent the first reported threefold interpenetrated uranium MOFs. 5. Rare earth metal free LMOFs/LCPs While REE free LMOFs and LCPs have less predictable luminescence behaviors than those with lanthanide-based emission, they can emit at essentially any wavelength and their emission energy is readily tunable. They often have a lower material cost and simpler coordination geometries. Materials with ligand-centered emission can be rationally designed, and quantum yields for these materials are among the highest reported for any class of phosphor. This section will discuss various examples of the diverse luminescence mechanisms extant in REE free LMOFs and LCPs and describe several design strategies. 5.1. Rare-earth metal free LMOFs/LCPs with colored emission One of the earliest fluorescent lanthanide-free framework materials in 2001, in which the blue-emitting molecule norfloxacin (H-Norf) was constructed into a 2D coordination polymers using Zn2+ [99]. The LCP emits ligand-centered blue emission at

15

420 nm based on the p–p⁄ transition with a quantum yield of 26%, which is significantly improved from the free ligand. The authors attribute this to the rigidification of the ligand molecule, reducing non-radiative decay of the ligand excited state. The expectation that ligand rigidification could improve quantum yield was also utilized to create a pair of LMOFs based on Zn and the ligand trans-stillbene [(E)-4,40 -(ethene-1,2-diyl)dibenzoic acid, L] [100]. Upon excitation, the free ligand rapidly interconverts from the trans to the cis isomer, quenching the emission. However, after incorporation into the 2D framework Zn3L3(DMF)2(1) and the 3D framework Zn4OL3, ligand centered emission increases. A similar rationale was developed by the M. Dinca group in 2011, who sought to identify the utility of aggregation induced emission phosphors as ligands in LMOFs [101]. In solution, AIE phosphors have limited emission due to molecular vibrations or tortions providing facile nonradiative excitation decay. However, in the bulk phase, these are prevented, resulting in emission turn on. The Dinca group used a ligand based on tetraphenylethylene, in which the excited state relaxes through rotation of the phenyl groups while this rotation is prevented in the bulk phase, turning emission on. Mounting AIE phosphors into LMOFs can also prevent this rotation. They prepared the tetracarboxylate ligand TCPE (tetra kis(4-carboxyphenyl)ethylene) and reacted it with Zn2+ and Cd2+ to form 1 (Zn2(TCPE)(H2O)24DEF, DEF = diethylformamide) and 2 (Cd2(TCPE)(DEF)(EtOH)4DEF, EtOH = ethanol), respectively. The excited state lifetimes of both 1 and 2 closely matched that of bulk H4TCPE, indicating that the mechanism of luminescence was similar to the bulk ligand luminescence despite the open frameworks of 1 and 2 preventing the direct ligand stacking that restricts rotations in the bulk phase (Fig. 16). The emission was similarly related, with both H4TCPE and 1 displaying emission peaks at 480 nm, while the emission from 2 was slightly blueshifted to 455 nm (Fig. 16). To develop highly-efficient REE-free yellow phosphors, an AIE phosphor ligand related to tpe was selected by the Li group in 2014 to construct a yellow-emitting LMOF with exceptionally high quantum yield [45]. The ligand, 1,1,2,2-tetrakis(4-(pyridin-4-yl)p henyl)ethane (tppe), typically emits in the green region with a peak at 500 nm and a quantum yield of 57.2% in the bulk phase. To achieve yellow emission with high quantum yield, tppe was immobilized into a framework with a second ligand which would serve as a bandgap modulator. DFT calculations suggested that the LUMO energy level of 1,3,5-benzenetricarboxylic acid (H3btc) was lower than the LUMO energy level, enabling a LLCT emission

Fig. 16. (left) Single crystal structures of H4TCPE, 1, and 2, demonstrating the shortest phenyl–phenyl (top) and H–H (bottom) distances (C = gray, H = white, O = red, Zn = orange, Cd = green). (right) Normalized emission (solid lines) and absorption (dashed lines) spectra for H4TCPE (blue), 1 (red), and 2 (green), with an epifluorescence microscopy image of a crystal of 2 inset. Reproduced from Ref. [101] Copyright 2011, American Chemical Society.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

16

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Fig. 17. (a) Structure of LMOF-251. C = gray, N = dark blue, O = red, Zn = light blue, H omitted for clarity. (b) Diagram demonstrating redshifting of emission from LMOF-251 relative to tppe through incorporation of the btc ligand. (c) Comparison of the excitation and emission spectra from commercial yellow phosphor YAG:Ce (dark red, pink) and LMOF-241 (blue, green). (d) Excitation and emission spectra of the bulk tppe ligand (dotted) and LMOF-251 (solid) under 360 (black), 400 (red), 420 (blue), and 440 nm (green). Reproduced with permission from Ref. [45]. Copyright 2014, American Chemical Society.

to occur and redshifting the LMOF emission relative to the bulk tppe ligand (Fig. 17). The resulting structure LMOF-251 has the formula of [Zn6(btc)4(tppe)2(sol)2]. It exhibits bright yellow emission at 540 nm with extremely high quantum yield of 90.7% under 400 nm excitation. Suspensions of the LMOF powder are also stable in ethyl acetate, and a dip-coating method was used to create a prototype small power WLED device. Following ultrasonication of LMOF-251 in ethyl acetate for one hour, a small blue LED bulb was dipped into the suspension several times until a thin layer of LMOF-251 formed on the surface. Under 3 V bias, bright white light was produced with luminous efficiency of 47.4 lm W1, exceeding the minimum requirement for small LED directional lamps (40 lm W1) [21]. The same dip-coating method was used to coat the LMOF onto a small piece of string, indicating that the material is able to be coated onto flexible substrates as well, suggesting significant commercialization potential. Another highly emissive LMOF with a chromophoric ligand based on a tpe core was reported in 2014 by the Zhou group, with UV-excitable blue emission at 470 nm and an impressive quantum yield of 99.9 ± 0.6% under Ar and 76.1 ± 3.5% in air [15]. The LMOF PCN-94 is constructed from 12-coordinated Zr6 clusters, with each linked to 18 other clusters by 12 ETTC ligands [H4ETTC = 40 , 4000 ,400000 ,40000000 -(ethane-1,1,2,2-tetrayl)tetrakis(([1,10 -biphenyl]-4-carboxylic acid))] (Fig. 18). While the bulk ligand H4ETTC emits yellow light (545 nm) under blue excitation, PCN-94 has its emission significantly blueshifted to 470 nm (Fig. 18). Despite the large shift in energy, emission from PCN-94 is the result of ligand-centered p–p⁄ transitions. The observed blueshift is attributed to two factors; first, in the open PCN-94 framework, there is no direct interaction

between any ligand molecules, preventing excimer-based emission that occurs in the bulk ligand. Second, upon incorporation into PCN-94, the ligand molecules were twisted such that the dihedral angle between neighboring phenyl rings increased from 35.93° in the free conformation to 54.45° in the LMOF. This broke the conjugation between neighboring rings and increased the energy gap between the HOMO/LUMO of the ligand. Additionally, the high quantum yield of the material was attributed to the rigidification of the ligand into a framework that prevented the phenyl ring rotoring behavior that is the common excited state deactivation pathway in tpe-based phosphors. The ligand H4ETTC was simultaneously developed by the Li group, which was named H4tcbpe [1,1,2,2-tetrakis(4-(4-carboxy phenyl)phenyl)ethene]. The Zn2+-based LMOF built on this ligand (LMOF-231) emits bright yellow color (550 nm) with a 76.41% quantum yield under 455 nm excitation—the highest value reported to date for blue-excitable, yellow-emitting, REE free LMOF phosphors [44]. The LMOF consists of tetrahedrally coordinated Zn2+ ions linked into an infinite 1D chain by bridging carboxylate groups from four ligand molecules, with the ligand molecules forming a close-packed layer structure (Fig. 19). Much like PCN-94, the rigidification of the tcbpe ligand upon incorporation into the framework drastically increased its quantum yield; however, unlike PCN-94, LMOF-231 exhibits a ligand-based emission without the involvement of metal orbitals in the frontier energy states. To demonstrate its suitability for commercialization, a small blue LED was coated with LMOF-231 using a dipcoating method, producing a WLED with 58.9 ± 1.5 lm W1 under 3 V bias.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

17

Fig. 18. (a) Structure of the ligand ETTC4. (b) Structure of the Zr6 cluster. (c) Structure of the PCN-94 framework, with the pore cavity shown in yellow. C: black, O: red, Zn: blue polyhedron. (d) Absorption (dotted) and emission (solid) spectra for the bulk H4ETTC ligand (orange) and PNC-94 (blue). Reproduced with permission from Ref. [15]. Copyright 2014, American Chemical Society.

A Mg2+-based luminescent coordination polymer with emission tunable between green and yellow was reported by Huang et al. in 2015 with the formula [Mg2(H2O)(1,4-NDC)2(1,10-phen)] (2; 1,4-NDC = 1,4-naphthalenedicarboxylate; 1,10-phen = 1,10phenanthroline) [102]. Mg2+ was selected because of its 3d0 electron configuration, which limits interference with the desired ligand-based emission, and for its low cost, low toxicity, and high abundance. Under 335 nm excitation, emission from the LCP showed a high energy blue-violet peak at 380 nm and a low energy yellow peak at 540 nm resulting in net green emission. These two emission bands closely matched the emission of the free 1,10-phen and 1,4-NDC ligands, respectively. The fluorescence decay profile of the LCP monitored at 380 nm also closely resembles to that of 1,10-phen, and the decay profile of the LCP monitored at 540 nm closely resembles to that of 1,4-NDC. The similarities in both emission wavelength and lifetime indicate that both emission peaks are the result of ligand-centered emission from the two types of ligands, with 1,10-phen contributing the high energy emission and 1,4-NDC contributing the low energy emission. As the excitation wavelength was increased, the intensity of the high energy band fell away and the intensity of the low energy band rose, allowing the emission to be tuned from green into yellow, with

most of the high-energy emission absent under 370 nm excitation. When monitored at 540 nm, the excitation spectrum for the LCP showed a peak at 450 nm, and when excited at 450 nm, the LCP’s quantum yield was 20%. Another blue-excitable, yellow emitting LMOF based on the chromophoric ligand H4tcbpe was reported in 2016 by the Li group. LMOF-401 is composed of tcbpe4 in conjunction with Bi3+and K+, with a formula of K[Bi(tcbpe)(DMF)2]xDMF (DMF = dimethylformamide) [43]. Fluorescence from LMOF-401 is ligand based, with extreme sensitivity to the solvation state of the framework (Fig. 20), with tuning the LMOF’s emission from blue to yellow made possible by altering the amount of DMF present in the material’s pores. In the as-made LMOF, which has approximately 5 DMF solvent molecules per formula unit, exposure to 395 nm excitation results in blue emission at 459 nm with a quantum yield of 57%. As the structure is desolvated, the emission redshifts and increases in intensity. With 3 DMF molecules per formula unit, emission from the LMOF peaks at 500 nm, and when the structure is entirely desolvated, the LMOF emits at 553 nm with a quantum yield of 74% under 455 nm excitation. The changes in emission are reversible for solvation states with one or more DMF molecules per formula unit; however, upon removal of the final DMF molecule,

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

18

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Fig. 19. (a) Structure of the 1D zinc-carboxylate chain infinite PBU. (b) Representation of the infinite PBU as edge-sharing tetrahedra. (c) The simplified tcbpe4 ligand. (d) The complete structure of LMOF-231, showing the zinc-carboxylate chains running along the c axis, bridge by tcbpe ligands in ordered stacks. (e) Excitation (dotted) and emission (solid) spectra of LMOF-231 (red) and the commercial yellow phosphor YAG:Ce (black). Reproduced from Ref. [44] with the permission of the Royal Society of Chemistry.

PXRD of the samples shows a structure change that is not reversible. This is likely because the final DMF molecule coordinates directly with the Bi3+, and its removal causes the structure to irreversibly change. Two new isoreticular yellow-emitting LMOFs based on a fluorinated derivative of the H4tcbpe chromophore in combination with bipyridine-based secondary ligands and Zn2+ were recently reported [31]. LMOF-301 ([Zn2(tcbpe-F)(bpy)]xDMA, tcbpe-F = 1,1,2,2-tetrakis(4-(3-fluoro-4-carboxyphenyl)phenyl)ethane, bpy = 4,40 -bipyridyl, DMA = dimethylacetamide) and LMOF-302 ([Zn2(tcbpe-F)(bpee)]xDMA, bpee = 4,40 -bipyridylethylene) both crystallize with 2D Zn2(tcbpe-F) sheets pillared by the secondary ligand to give 3 dimensional framework with paddlewheel SBUs. LMOF-301 and 303 fluoresce at 514 and 530 nm, respectively, under 455 nm excitation, which is slightly blueshifted relative to the bulk ligand, which fluoresces at 540 nm. This indicates that some ligand–ligand charge transfer between the tcbpe-F and pillaring bipyridine ligands may play a role in the emission. This is supported by their quantum yields. LMOF-301 has a quantum yield of 50.9% under 455 nm excitation, while LMOF-303 has a quantum yield of 10.1%; in both cases, the emission efficiency is lower than reported for tcbpe-based LMOFs without a secondary ligand.

A luminescent coordination polymer based on a BiBr4 cluster was recently reported that emits 540 nm yellow light under UV excitation with high quantum yield (85%) [46]. The LCP has a formula of (TBA)-[BiBr4(bp4mo)] (TBA = tetrabutylammonium, bp4mo = N-oxide-4,40 -bipyridine), with BiBr4 clusters linked together by the neutral zwitterionic ligand to form a 1D chain. In the solid state, the LCP chains crystallize into layers with TBA cations between them for charge balancing purposes. In solution, emission from the LCP is blueshifted to 400 nm and the quantum yield of the material drops significantly. However, when frozen in solution, the emission from the LCP resembles that of the solid state, indicating that an aggregation-induced emission process is responsible for the exceptional quantum yield. The LCP’s yellow emission at 540 nm has a lifetime of 18 ms, indicating a phosphorescence mechanism that is likely induced by the heavy Bi3+ ion. TD-DFT calculations show that the lowest-energy excitation processes involved charge transfer from the inorganic BiBr4 cluster to the ligand, indicating that the AIE-phosphorescence previously identified occurs through a metal-to-ligand charge transfer process. Finally, the CP is mechanochromic; upon grinding, it become amorphous and the emission redshifts into the orange region with a peak at 585 nm. However, upon heating at 80 °C or exposing the material to atmosphere saturated with water vapor, the original structure and luminescence were recovered. The triangular chromophoric ligand tppa [tri(4-pyridylphenyl) amine] was recently used to synthesize two blue-excitable yellowemitting LMOFs with emission centered on the tppa ligand [42]. Under 455 nm excitation, both 1 ([Zn(tppa)(ndc)](DMF)4, ndc = 2,6-naphthalenedicarboxylate, DMF = dimethylformamide) and 2 ([Zn4(tppa)2(sdc)3(NO3)2](DMF)4(ACN)2, sdc = (E)-4,40 -(ethe ne-1,2-diyl)dibenzoic acid, ACN = acetonitrile) emit yellow light with similar CIE coordinates to the commercial yellow phosphor YAG:Ce. Compound 1 emits at 548 nm and 11.4% quantum yield, which slightly redshifts to 553 nm upon removal of the solvent DMF; the redshift is accompanied by an improvement in the quantum yield to 27.6%. Compound 2 emits at 535 nm with 20.6% quantum yield, which is drastically redshifted to 580 nm following solvent removal. Unlike 1, the quantum yield of 2 drops to 12.2% upon activation. A recent publication has reported the synthesis of an LMOF with three distinct active emission processes which can be selected between by altering excitation wavelength [51]. Compound 1 [(Me2NH2)3[Pb2K(TTCA)2Cl2, TCA = triphenylene-2,6,10-tricar boxylic acid] is composed of trinuclear Pb2KCl2 clusters coordinated to six ligand carboxylates in a 2D bilayer framework and has three distinct emission peaks at 381 nm, 502 nm, and 619 nm (Fig. 21). The emission at 381 nm is attributed to ligandcentered emission, while the emission at 502 nm is attributed to ligand-to-ligand charge transfer, and both peaks are present in the bulk H3TTCA emission spectrum. The 619 nm emission is attributed to ligand-to-metal charge transfer between the TTCA3 and Pb2+ ion. Interestingly, the ligand-centered emission at 381 nm is favored at relatively high excitation energies (kex = 270–320 nm), while the LMCT emission process at 619 nm is favored at mid-range excitation energies (kex = 350–370 nm), and the LLCT emission process at 508 nm is favored by relatively low excitation energies (kex = 390–410 nm). As a result, when the excitation energy is tuned from high energy to low energy, emission from 1 passes from blue into purple/red and finally to green. While the overall quantum yield is relatively low (3% under 410 nm excitation), the unusual luminescence properties of 1 suggest that more complex LMOF phosphor behaviors are attainable. 5.2. Rare-earth metal free LMOFs/LCPs with white emission One of the first rare-earth free white-emitting coordination polymers was reported by Liu et al. in 2008 [52]. Their phosphor

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

19

Fig. 20. The colored legend at bottom indicates the weight percent of DMF in the material, and the estimated number of DMF molecules per formula unit. (a) Images of LMOF241 at various solvation states under ambient light and 395 nm excitation. (b) Emission spectra from LMOF-241 at various solvation states. (c) CIE chromaticity diagram demonstrating how emission color changes as a function of solvation. (d) PXRD patterns of LMOF-241 under various solvation states. Reproduced with permission from Ref. [43]. Copyright 2016, American Chemical Society.

Fig. 21. Emission from compound 1 under varying excitation wavelengths. Reproduced from Ref. [51] with the permission of the Royal Society of Chemistry.

linked ZnCl2 together in a 1D chain with the ligand 1,3-di(4,40 -pyr idyl)propane (dpp), with Zn2+ coordinated in a tetrahedral fashion to two pyridyl groups and two Cl ions. Under 400 nm excitation, the LCP Zn2Cl4bpp2 emits white light in a broad spectrum peaking at 480 nm. The quantum yield of white light emission is 32%, and

the CIE chromaticity coordinates are (0.328, 0.328), indicating a slightly cold white light. The authors also note that the phosphor material is stable at up to 200 °C. A 2D layered CP of Ag+ and 4-cyanobenzoate with emission tunable from yellow into white by variation of the excitation wavelength was reported by Guo et al. in 2009 [54]. Under 350 nm excitation, compound 1 (AgLnH2O, L = 4-cyanobenzoate) emits through two distinct luminescence processes to generate white light with a quantum yield of 10.86%. The first is a ligand-based fluorescence (s = 0.87 ns) at 427 nm with a, and the second is a metalto-ligand charge transfer (MLCT) phosphorescence (s = 2.60 ms) that generates a main peak at 566 nm, with shoulders at 513 and 617 nm. As the excitation wavelength is decreased from 350 nm, the contribution from ligand-based fluorescence diminishes while the MLCT phosphorescence increases, resulting in purely yellow MLCT emission under 330 nm excitation (Fig. 22). Because of this sensitivity to relatively small changes in excitation wavelength, the balance between ligand-centered/MLCT emission—and therefore the color of white light generated—can be finely adjusted. When excited at 350 or 349 nm, the CIE coordinates of the white light are (0.31, 0.33) or (0.33, 0.34), respectively. While the previous example produced white light through a combination of ligand-centered and metal-to-ligand charge transfer-based emission, Shi et al. reported a 1D Zn2+-based luminescent coordination polymer which produced white light by combining ligand-centered and ligand-to-metal charge transfer emission processes [68]. The LCP in question is a 1D helical chain

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

20

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Fig. 22. Emission spectra of 1 under varying excitation energies showing ligand-centered fluorescence at 427 nm and MLCT phosphorescence at 513–617 nm. Inset is an image of 1 under 350 and 330 nm excitation. Reproduced with permission from Ref. [54]. Copyright 2009, American Chemical Society.

Fig. 23. Emission spectra of 1a under varying excitation energy, with image of a sample excited at 360 nm inset. Reproduced from Ref. [68] with the permission of the Royal Society of Chemistry.

consisting of tetrahedrally coordinated Zn2+ in combination with the radical anion ligand 2,30 -biimidazo[1,2-a]pyridin-20 -one (bipo) and a terminal formate group (1), with the formate exchangeable with SCN to give 1a or N 3 to give 1b. Emission from 1 under 360 nm excitation is entirely ligand-centered, giving green light in a broad peak at 507 nm with a small shoulder at 468 nm. Upon exchange of the terminal formate with N 3 to create 1b, an additional emission peak appears at 593 nm, shifting emission into the yellow-green region. This additional peak is intensified in 1a, with additional shoulders appearing at 550 nm and 640 nm leading to the production of warm white light with CIE coordinates of (0.34, 0.36). Interestingly, unlike most other LMOFs with dual emission processes resulting in net white light emission, the ratio of ligand-based to LMCT emission is not wavelength dependent;

the color of white light emitted by 1a is wavelengthindependent, with changes to excitation energy only changing the total emission intensity (Fig. 23). A white light emitting 3D coordination polymer was recently reported by Xu et al. that is based on a thioether-functionalized benzenedicarboxylate ligand L2 (2,5-bis(((S)-2-hydroxypropyl)thi o)terephthalic acid) and Pb2+ [103]. In the resulting structure, the Pb2+ ions are linked into a 1D chain, with four ligands attached to each chain and each ligand bridging between two chains. Interestingly, the thioether and hydroxyl groups on the ligand side chains all participate in coordination with the lead ion. Emission from Pb–L2 is broad, with peaks at 459 nm and 515 nm and a long tail that extends to nearly 700 nm. While the intensity of the 515 nm peak and tail region are mostly independent of excitation wavelength, the intensity of the ligand-centered 459 nm peak can be tuned by adjusting the excitation wavelength between 300 and 360 nm. While the reported quantum yield is low, on the order of 2–3%, the author points out that there is significant overlap between the excitation and emission spectra of Pb–L2, which likely results in an underestimation of the actual quantum yield. Two Pb2+ luminescent coordination polymers using the ligand 1-tetrazole-4-imidazole-benzene (Htizb) were recently found to emit both high energy (blue) and low energy (yellow) bands based on intraligand p–p⁄ transitions and ligand-to-metal charge transfer, respectively, which could be combined to produce white light under specific excitation conditions [55]. Both LCPs are solventfree 2D layer structures, with formulas of Pb(NO3)(tzib) (1) and Pb(tzib)2 (2). In 1, excitation at 320 nm results in orange emission at 595 nm, which stems from a LMCT process between ligand p orbitals and the Pb2+ p orbitals. Upon increasing the excitation wavelength, the emission intensity resulting from this process increases and slightly blueshifts, while an intraligand p–p⁄ process causes emission in the blue region at 400 nm, resulting in white light (Fig. 24). Unlike 1, both ligand-centered and LMCT processes operate in 2 at all excitation wavelengths. Emission intensity from

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

21

Fig. 24. (a) Emission spectra from 1 under varying excitation wavelengths. (b) CIE chromaticity diagram for 1 under 320, 330, 335, 338, 340, and 350 nm excitation, with the photoluminescence images of 1 under the same excitation conditions inset. (c) Emission spectra from 2 under excitation wavelengths varying from 320 to 350 nm. (d) Emission spectra from 2 under excitation wavelengths varying from 360 to 385 nm. (e) CIE chromaticity diagram for 2 under 320, 350, and 385 nm excitation, with the photoluminescence images of 2 under the same excitation conditions inset. Reproduced from Ref. [55] with the permission of the Royal Society of Chemistry.

Fig. 25. Emission from [Cd(tzphtpy)2]6.5H2O under varying excitation wavelengths, with the corresponding CIE chromaticity diagram and photoluminescence images of the LMOF under several excitation energies inset. Points 1 through 8 correspond to excitation at 286, 326, 356, 386, 396, 413, 440, and 460 nm, respectively. Reproduced with permission from Ref. [80]. Copyright 2016, American Chemical Society.

2 increases with increasing excitation wavelength from 320 to 350 nm, with the intensity of the high energy ligand-centered p– p⁄ band increasing faster (Fig. 24). However, as the excitation wavelength passes 360 nm, emission from the high energy band redshifts significantly while emission from the low energy LMCT band decreases. The authors attribute this difference in behavior between 1 and 2 to the presence of the nitrate anion in 1, as it could influence both emission mechanisms. A third variety of LMOF with dual photoluminescence processes resulting in the direct emission of white light was reported by the

same group, in which ligand-centered emission combines with emission from a ligand-to-ligand charge transfer process to give white light [104]. In the LMOF Zn(L)(HBTC)(H2O)2 (5, L = N4,N40 di(pyridin-4-yl)-[1,10 -biphenyl]-4,40 -dicarboxamide, H3BTC = 1,3,5-benzenetricarboxylic acid), blue emission at 436 nm arises from intraligand p–p⁄ transitions on L, while yellow emission at 539 nm arises from LLCT. This LLCT process is believed to be mediated by p–p interactions between pyridyl rings on neighboring L molecules. Emission from this LMOF is extremely dependent on excitation wavelength and can be tuned between majority blue emission at 370 nm excitation, through white, and into majority yellow emission at 381 nm excitation. Another LMOF that was recently reported to exhibit direct white light emission through a combination of ligand-centered and LLCT processes is based on the new azole/terpyridine ligand Htzphty [4-(tetrazol)-5-yl)pheyl-2,20 :60 ,200 -terpyridine] [80]. In coordination with Cd2+, this ligand forms 1D chains with the formula [Cd(tzphtpy)2]6.5H2O, which forms strong p–p interactions between ligand molecules both within a single chain and between multiple neighboring chains. Emission from this material occurs through a ligand-centered p–p⁄ transition, which emits at 454 nm, and an interligand p–p-mediated LLCT transition, which emits at 554 nm (Fig. 25). Under 326 nm excitation, the resulting white light has CIE coordinates of (0.33, 0.36). As the excitation wavelength increases, ligand-centered emission decreases and LLCT emission increases. Increasingly warm white light emission is maintained between 286 and 386 nm, beyond which yellow light is generated. Theoretical methods including density of states (DOS) and time-dependent density functional theory (TD-DFT) calculations were used to investigate the luminescence mechanism, demonstrating that the Cd2+ makes no contribution in the vicinity of the valence/conduction bands, and that the 554 nm emission is based on an interligand p–pmediated charge transfer process.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

22

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Fig. 26. (a) Schematic showing structure of Bio-MOF-1 and its loading with Sm(III), Tb(III), and Eu(III) to produced LMOFs with lanthanide-based luminescence. (b) Excitation and emission spectra of Sm@Bio–MOF-1 (i), Tb@Bio–MOF-1 (ii), and Eu@Bio–MOF-1 (iii). Reproduced with permission from Ref. [32]. Copyright 2011, American Chemical Society.

6. Host–guest LMOF/LCP phosphors The permanent porosity is one of the most distinctive qualities of MOFs and is central to most other MOF applications (gas adsorption, sensing, catalysis, etc.). Pore size, geometry, and chemical environment can all be fine-tuned. While some LMOFs don’t take advantage of this attribute, the pores in such frameworks often present unique opportunities to include certain chemical species which can influence or compliment framework emission, or emit characteristic light of their own.

6.1. Host–guest LMOFs/LCPs with colored emission The loading of lanthanide ions into a framework material is a commonly utilized method to introduce or alter luminescence processes in that material. In the case of anionic frameworks with charge-balancing cations present in the pores, cation exchange provides a method to easily incorporate lanthanide cations within the framework, with ion-ion interactions securing the lanthanide ions within the pores and encouraging their inclusion. However, the lack of a strong bond between the lanthanide ion and the

framework can limit the degree to which the lanthanide ions are sensitized. An example of this method of loading, with its corresponding advantages and disadvantages, was recently reported using BioMOF-1 [Zn8(ad)4(bpdc)6O2Me2NH2,8DMF,11H2O] (ad = adeninate, bpdc = biphenyldicarboxylate, DMF = dimethylformamide); a highly porous (BET surface area  1700 m2g1) anionic MOF constructed of zinc-adenate clusters linked into a 3D framework by bpdc ligands, with charge-balancing dimethylammonium cations resident in the structure’s pores [32]. These cations were exchanged for the lanthanide(III) cations in aqueous solvent, giving Tb@Bio–MOF-1, Sm@Bio–MOF-1, and Eu@Bio–MOF-1. The excitation and emission spectra indicate that Bio-MOF-1 is able to effectively sensitize Eu(III), as no framework-based peak is present in the emission spectrum. Only slight sensitization of Tb(III) occurs, as a significant framework emission peak is present at approximately 420 nm, and minimal sensitization of Sm(III) (Fig. 26). The limited sensitization in Tb@Bio–MOF-1 and Sm@Bio–MOF-1 attributed to increased thermally activated back transfer from the excited lanthanide to the ligand triplet state, which is supported by lifetime data. Given the lack of sensitization and aqueous environment, the quantum yields of Tb@Bio–MOF-1 and

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Sm@Bio–MOF-1 are low (1.7% and 0.28%, respectively), but the quantum yield of Eu@Bio–MOF-1 is consistent with previously discussed examples of Eu(III) emission in aqueous environments (8.4%). To improve the sensitization of lanthanide guests, the lanthanide ions can be loaded into frameworks which possess open binding sites. This can be accomplished by choosing MOFs with dangling, nonbonded carboxylates or other heteroatom-rich ligands. This was demonstrated by Dong et al., who reported a Cd-LMOF with Cd2+ in combination with the heteroatom-rich ligand 3,5-bis(3-carboxyphenyl)-1,2,4-triazole and 4,40 -bis(4pyridyl)ethane [64]. The lanthanides Tb(III) and Eu(III) were incorporated into the structure postsynthetically through a diffusion-mediated process in aqueous media, and under UV excitation, the resulting structures showed characteristic Tb(III) and Eu(III) emission, with quantum yields of 10% and 1%, respectively. The emission spectra indicate that Tb(III) was efficiently sensitized by the material, as no framework emission was observed, while the Eu(III) was only partially sensitized. Additionally, Tb(III) and Eu(III) could be simultaneously loaded into the structure, again under aqueous conditions, to produce a yellowemitting material with a quantum yield of up to 4%. In addition to introducing lanthanide guests to influence LMOF emission, emission from framework lanthanides may be influenced by the inclusion of guest molecules which can alter the rigidity of the framework. By improving the rigidity of the framework, the quantum yield can be improved. An example of the role that solvent rigidification can play in luminescence can be found in work by Liu et al., who used an anionic lanthanide metalloligand [Ln(ODA)3]3 (Ln = Gd(III) or Tb(III), ODA = oxydiacetate) in combination with Cd in H2O to give two structures [39]. Compound 1 is a neutral framework which was obtained through slow evaporation of the solvent water, with 1a being the Gd(III) variety and 1b being the Tb(III) variety. Compound 2 is an anionic framework with [Cd(H2O)6]2+ cations in the pore that was the result of solvothermal synthesis. Compounds 2a and 2b are likewise the Gd(III) and Tb(III) versions, respectively. Compound 1b has a high quantum yield of 78.98%, with characteristic Tb(III) emission, and a long lifetime of 2.84 ms. Following activation by heating under vacuum at 220 °C, the quantum yield of 1b is decreased to 57.6%, but then increases to 95.2% and 93.4% upon exchange with acetone or ethanol. The authors believe that the anomalous increase in quantum yield observed in the presence of molecules with OH vibrations is due to decreased rigidity in the absence of solvent molecules. Sheng et al. also investigated the effects that solvent guests have on the emission of lanthanides with coordinated water molecules [105]. They synthesized a series of isostructural Tb(III) and Eu(III) LMOFs with 2 coordinated waters per lanthanide in NMP (N-methyl-2-pyrrolidone), DMF (dimethylformamide), DMA (dimethylacetamide), and water, so the pores of the LMOF contained the solvent it was synthesized in. For all 8 LMOFs (Tb(III) or Eu(III) in all four solvent conditions), they measured QY and lifetime. They found samples with water-filled pores produced the lowest quantum yield, followed by DMF and then DMA, with NMP producing the highest quantum yield. For the Tb(III) containing structures, the highest quantum yield was 56.5%, while the lowest was 13.9%, while the Eu(III) structures ranged between 3.1% and 15.7%. Single crystal data showed that there is an uncoordinated carboxylate group from the ligand that makes a hydrogen bond with the free solvent, and that free solvent is also able to make a hydrogen bond with the coordinated water. As the quantum yield of the resulting solvated structures is in trend with the flexibility of the solvent molecules, and because the quantum yield of the solvated structures increases significantly upon reduced temperature (suggesting that molecular vibrations/rotations/tortions are responsible for some nonradiative excitation decay) the

23

authors believe that the solvent molecules influence the quantum yield by increasing the framework rigidity. Guest species can also be used to influence emission mechanisms more directly. In the case of lanthanide-based LMOFs, effective sensitization of the lanthanide emitter relies on populating the ligand triplet state. Heavy atom guests, including large halides such as Br, can be introduced into the pores of these LMOFs; the heavy guest can then increase triplet populations on the ligand through the heavy atom effect. This was recently explored in a new series of isostructural LCPs based on Eu(III), Tb(III), and Gd(III), with the formula [Ln(Bcpi)2(H2O)]BrxCl1x(H2O) (Ln = Eu, Tb, or Gd; Bcpi = 1,3-bis(4-carboxyphenyl)imidazolium; x = 0, 0.5, 1) [106]. The resulting LCPs are cationic 2d nets, which stack to create 1D channels occupied by the halide anion and solvent water, and show characteristic Eu/Tb emission. It was found that increasing bromide content increases QY; for 1 (Ln = Eu, x = 0), the quantum yield is 3.50%, which increases substantially to 15.80% in 2 (Ln = Eu, x = 0.5) and 23.46% in 3 (Ln = Eu, x = 1). The Tb series shows the same enhancement, with QY increasing from 3.08% in the 0% Br structure (4) to 13.83% and 19.63% in the 50% (5) and 100% (6) Br structures, respectively. The photoluminescent lifetimes of these materials were similarly affected. Compound 1, 2, and 3 had lifetimes of 501, 572, and 583 ms, while 4, 5, and 6 had lifetimes of 716, 825, and 894 ms. The authors attribute the influence of Br anions on the excited state lifetimes and quantum yields to the intra-heavy atom effect, which would serve to increase the population rates of the ligand excited state and allow for more effective sensitization. This is supported by the calculated sensitizations efficiencies of 1, 2, and 3 of 7.6%, 28.4%, and 47.6%, respectively. In addition to lanthanides, other small, emissive species can be introduced into LMOF pores. The J. C. Tan group recently encapsulated emissive ZnQ (Zn-bis-8-hydroxyquinoline) clusters inside the zeolitic imidazole framework ZIF-8 through an in situ process, in which the ZnQ clusters were formed simultaneously with the framework, resulting in single ZnQ clusters encapsulated inside ZIF-8 pores, and pore windows small enough to prevent migration of the ZnQ out of the structure [107]. Optical spectra taken of ZIF-8, the ZnQ guest, and the ZnQ  ZIF-8 material indicate that upon encapsulation, the optical band gap of ZIF-8 decreases by 0.63 eV, while the optical band gap of ZnQ increases by 0.27 eV (Fig. 27). Despite the apparently increased bandgap of the emissive ZnQ clusters, emission from ZnQ  ZIF-8 is redshifted by 36 nm relative to the free cluster (Fig. 27). The authors attribute this to a charge transfer mechanism from the high-lying LUMO of the ZIF-8 framework into the lower-lying LUMO of the ZnQ cluster. The previous examples discussed cases in which either a luminescent guest was introduced into a framework or a nonluminescent guest was used to influence emission from a luminescent framework. However, it is also possible to design a system by which the emission is not guest or framework centered but is instead entirely based on charge transfer between the two. By taking advantage of materials with appropriate redox potentials, it is also possible to tune that charge transfer process and select the emission wavelength of the resulting material. A set of guest-loaded LCPs with luminescence from a guest–host charge transfer mechanism was recently engineered to demonstrate this by including electron-rich guests in a framework constructed from an electrondeficient ligand [63]. The ligand 2,4,6-tri(pyridin-4-yl)-1,3,5-tria zine (tpt) was selected, as the triazine moiety is known for being electron deficient and for having exceptional pi-pi stacking abilities, which improve charge transfer. It was immobilized into a framework using the secondary ligand 1,4-pda (1,4-H2pda = 1,4,phenylenediacetic acid) and the d10 metal Cd2+, which was selected for its nonparticipation in luminescence, as discussed in section 3. The resulting LCP was the 2D layered structure [Cd(tpt)(1,4-pda) (H2O)2](tpt)2(H2O) (1), with tpt as both a ligand and present as a

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

24

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Fig. 27. (a) UV–Vis absorption spectra of ZIF-8, the ZnQ guest cluster, and the guest-loaded ZIF-8; the inset spectra show the region from 200 to 550 nm with intensity normalized. (b) Kubelka-Munk (KM) function of the three samples, demonstrating the shift in bandgap for the cluster and framework in the guest-loaded sample. (c) Emission spectra for ZIF-8, the ZnQ guest cluster, and the guest-loaded ZIF-8 under 350 nm excitation. (d) The emission spectrum of the ZnQ cluster and the absorption spectrum of ZIF8, demonstrating the lack of overlap between the two. Reproduced from Ref. [107] with the permission of the Royal Society of Chemistry.

guest molecule between the layers (Fig. 28). A series of polycylic aromatic hydrocarbon (PAH) molecules were chosen to function as the electron-rich guest molecules, as their planar structure would fit well between the 2D LCP layers, and they are known to participate in donor–acceptor luminescence systems. When the reaction was performed in the presence of PAH guest molecules, the resulting structure was a 2D bilayer [Cd2(tpt)2(1,4-pda)2] (guest) (2, guest = triphenylene; 3, guest = pyrene; 4, guest = coronene; 5, guest = perylene) with guest molecules located between each bilayer (Fig. 28). The authors found that the absorption bands of the guest-loaded samples differed significantly from the both 1 and the bulk guest molecule in question, suggesting the formation of donor–acceptor systems in the LCP. This is further supported by the emission spectra of 2–5, which show emission from an exciplex formed between the electron-donating guest and electron-withdrawing tpt ligand. As the emission mechanism is based on charge-transfer between the guests and tpt ligand, the emission from the guest-loaded materials is linearly correlated to the ionization potential of the relevant guest (Fig. 28), providing a route to rationally designed luminescent species with specific emission wavelengths. 6.2. Host–guest LMOFs/LCPs with white emission In pursuing white light emission from host-guest materials, the most common technique is to find host frameworks and guest spe-

cies with complementary emission properties, i.e. a set of frameworks and guests with either blue/yellow or red/green/blue emission properties. As it is rare for lanthanide ions to possess strong yellow emission—only Dy(III) in certain coordination environments emits strongly in the yellow—most blue/yellow systems are lanthanide-free. Instead, lanthanide-based white-emitting host–guest materials more commonly use red/green/blue systems, as Tb and Eu can provide green and red emission, with blue emission coming from the framework. The first lanthanide-free host–guest LMOF to produce white light through was reported in 2013 by the Su and Li groups [53]. First, a mesoporous blue-emitting LMOF was prepared from Cd2+ and the ligand 2,4,6-tris(2,5-dicarboxylphenylamino)-1,3,5-tria zine (H6TATPT). The resulting anionic framework has a formula of DMA+15[(Cd2Cl)3(TATPT)4]12DMF18H2O (1, DMA+ = dimethylammonium, DMF = dimethylformamide). While the ligand weakly emits at 490 nm, the framework is highly emissive at 425 nm (QY = 15.1%) and emits through a MLCT mechanism. Following the synthesis of the anionic LMOF, the yellow-emitting iridium complex [Ir(ppy)2(bpy)]+ (Hppy = 2-phenylpyridine, bpy = 2,20 -bipyridine) is encapsulated in the pore through a cation exchange process with DMA+. This cluster was selected for its yellow emission at 570 nm, cationic character, and because its size of approximately 10  11 Å is small enough to fit into the pore windows in 1. The combination of blue framework-based emission and yellow guest-based emission results in white light with CIE coordinates of (0.31, 0.33). While

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

25

Fig. 28. (a) Structure of 1 (top) and 2–5 (bottom), showing the guest molecules in green located between the LCP layers, with the layer (1)/bilayer (2–5) structures shown at right. (b) Emission spectra of 1–5 under 370 nm excitation. (c) Relationship between ionization potential of the guest PAH and emission wavelength for 2–5. Reproduced with permission from Ref. [63]. Copyright 2017, American Chemical Society.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

26

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Fig. 29. Emission spectra (left) and CIE chromaticity diagram (right) of 1 with varying [Ir(ppy)2(bpy)]+ loading concentrations. Reproduced with permission from Ref. [53]. Copyright 2013, Macmillan Publishers Limited.

Fig. 30. (a) Emission spectra and (b) CIE chromaticity diagram of Eu/Tb@MIL-124 under varying excitation wavelengths, with photographs of the sample under 297 and 320 nm excitation inset. Reproduced from Ref. [108] with the permission of the Royal Society of Chemistry.

emission from the free cluster peaks at 570 nm, emission from the framework-encapsulated cluster blueshifted to 530 nm, which is attributed to the confinement of the cluster inducing a phenomenon

related to the rigidochromic effect. The cation exchange was optimized to provide varying concentrations of the Ir(III) cluster in 1, with 3.5% w/w providing the highest quality white light (0.31,

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

0.33) (Fig. 29), at a quantum yield of 20.4%, which was the highest then reported for white-emitting LMOFs. Additionally, this quantum yield was maintained up to temperatures of 115 °C. A blue emitting gallium MOF Ga2(OH)4(Hbtc) (MIL-124, H3btc = 1,2,4-benzenetricarboxylic acid) was selected to encapsulate Eu(III) and Tb(III) on the basis of its dangling carboxylate group, which provided a convenient location to incorporate the red and green emitting lanthanides and produce a white-light emitting LMOF [108]. The framework emits at approximately 400 nm through a MLCT mechanism, while the encapsulated lanthanides emit at their characteristic wavelengths, with the Eu(III) 5 D0–7F2, and Tb(III) 5D4–7F5 transitions providing the strongest peaks at 614 and 545 nm, respectively. The lanthanides are doped at a level of approx. 1:60 Ln:Ga, with a 1:1 ration of Eu:Tb. The relative efficiency of luminescence from the framework, Tb(III), and Eu(III) can be selected by tuning the excitation energy, with excitation between 297 and 315 nm producing white light with CIE coordinates of (0.3693, 0.3362) and a quantum yield of 22% (Fig. 30). In addition to Tb(III) and Eu(III), green and red emission can be provided by dye molecules. Qian et al. recently reported a warm white light emitting LMOF phosphor which was created by simultaneously encapsulating the red-emitting (625 nm) cationic dye 4(p-dimethylaminostyryl)-1-methylpyridinium (DSM) and the green-emitting (535 nm) cationic dye acriflavine (AF) into the blue-emitting (415 nm) anionic MOF ZJU-28 through a cation exchange process, which has the previously described advantages [23]. ZJU-28 has a formula of DMA+3[In3(BTB)4]12DMF22H2O (DMA+ = dimethylammonium, BTB = benzenetricarboxylate, DMF = dimethylformamide), and its 415 nm emission overlaps strongly with the absorption spectra of AF and DSM. As a result, loading of either dye molecule at concentrations higher than approximately 0.20% w/w results in efficient energy transfer from the framework to the dye, with solely dye-based emission observed. The addition of each dye individually tunes emission toward that dye’s color, with the encapsulation of AF tuning from blue to green, and encapsulation of DSM tuning from blue to red. By doping at a mixed concentration of 0.02% DSM, 0.06% AF, warm white light with CIE coordinates of (0.34, 0.33) and a quantum yield of 17.4% is achieved. Another host–guest LMOF system that takes advantage of combined emission from an emissive metal complex guest in combination with an emissive MOF framework was recently reported using NENU-521 ([(Zn4O)5(TPA)8(TDA)3(H2O)6]12DMF, H3TPA = 4,40 ,400 -

27

Fig. 32. Emission spectra (top), CIE chromaticity diagram (bottom), and photographs (inset) of the 1.87% CuI@1 samples under varying excitation wavelengths. Reproduced from Ref. [79] with the permission of the Royal Society of Chemistry.

Fig. 31. Emission spectra (left) and CIE chromaticity diagram (right) of NENU-521, AlQ3, and Alq3@NENU-521 at varying loading concentrations under 370 nm excitation. Reproduced from Ref. [109] with the permission of the Royal Society of Chemistry.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

28

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

nitrilotribenzoic acid, H2TDA = thiophene-2,5-dicarboxylic acid) in combination with AlQ3 (tris(8-hydroxyquinoline)aluminum) [109]. NENU-521 is mesoporous, with ligand-centered blue emission at 435, while AlQ3 has strong green-yellow emission at 530 nm and is less expensive than the iridium complex discussed previously. As both LMOF and cluster are neutral species, AlQ3 was loaded into the framework in a diffusion-mediated process. Unlike the Ir-complex encapsulation, which produced white light as a mix of blue framework emission and yellow complex emission, Alq3@NENU-521 showed a single broad peak (Fig. 31), indicating that the emission mechanism likely involved the formation of an exciplex between the cluster and framework. At 4.14% loading, white light with CIE coordinates of (0.291, 0.327) was generated, with a quantum yield of 11.4%. Additionally, the framework showed good stability, with almost no loss in performance following one month of continual operation under air. It is also possible to modify guest molecules to become emissive post-synthetically. Huang et al. reported the postsynthetic functionalization of a guest molecule within an emissive LMOF to create a white-emitting material [79]. A Mg-based coordination with formula [Mg2(1,4-NDC)2(H2O)2](bpy)(H2O)4 (1; 1,4-NDC = 1,4naphthalenedicarboxylate, bpy = 4,40 -dipyridyl) emits blue light from a ligand-centered mechanism based on 1,4-NDC ligand. The bpy guest molecules present from synthesis in the material’s pores serve as templating agents, are not bound to the framework, and do not participate in luminescence. However, the bpy molecules can be postsynthetically modified with CuI, introduced in a diffusion-mediated process, resulting in yellow emission at 550 nm from the modified guest. The blue/yellow emission results in white light at a loading of 1.87% w/w CuI. Because the excitation spectra of the guest/framework are distinct, the intensity of emission from each can be tailored by changing excitation wavelengths (Fig. 32). At 340 nm, emission from the framework is favored, while at 380 nm, emission from the guest is favored. When excited at 360 nm, the CIE coordinates of the resulting white light at (0.32, 0.33) with a QY of 5.61%. As the loading percentage of CuI was low, it was not possible to resolve the structure of the CuI cluster following addition to the LMOF. However, XPS indicated that the copper was bound to the pyridyl nitrogen.

7. Other types of LMOF/LCP materials for lighting devices Though this review has focused on the luminescence mechanisms of LMOFs and LCPs as phosphors in solid-state lighting

devices, MOFs and CPs are functional materials with wider use in lighting applications. They can be incorporated into hybrid materials which alter or improve their performance, stability, or processability, and there has also been research into preparing electroluminescent MOFs and CPs for use in light emitting diodes. While there is typically poor conductivity across metal–ligand bonds in LMOFs, creative ligand design has begun to address these difficulties.

7.1. Hybrid LMOF/LCP materials Rodrigues, Júnior et al. recently reported an electrospinning method to prepare PVA nanofibers loaded with lanthanide-based MOF nanocrystals [110]. The LnMOF-loaded nanofibers were produced with polyvinyl alcohol (PVA) in combination with one of two isostructural 2D LnMOFs ([Eu(DPA)(HDPA)] = EuMOF@PVA, ([Tb(DPA)(HDPA)] = TbMOF@PVA, H2DPA = pyridine 2,6dicarboxylic acid), or the codoped Tb/EuMOF@PVA, which was prepared with 95:5, 80:20, and 50:50 Tb:Eu. The LnMOFs were first synthesized using a microwave-assisted process which produced nanocrystals measuring between 100 and 400 nm across. A suspension of the LnMOF nanocrystals was then added to an aqueous solution of PVA, which was electrospun onto a glass substrate. The resulting composite material consisted of lanthanide crystals measuring between 100 and 400 nm embedded in PVA nanofibers. Shifts in the excitation spectra of both EuMOF@PVA and TbMOF@PVA relative to the free LnMOFs indicated that in both cases, the PVA matrix improved sensitization efficiency. However, a small decrease in quantum yield was observed (from 40.3% in EuMOF to 33.0% in EuMOF@PVA), which the authors attributed to decreased rigidity, as the chelation of the ligands was perturbed by incorporation into the PVA matrix. A method to prepared lanthanide LCPs covalently bound to mesoporous silica substrates was recently reported [111]. Three new 2D lanthanide CPs, using Tb(III), Eu(III), or Nd(III), were synthesized with an the aldehyde-functionalized ligand 5-((4-formyl phenoxy)methyl)isophthalic acid (H2L). When samples of the three LCPs (Tb-L, Eu-L, and Nd-L) were synthesized in the presence of amine-decorated mesoporous silicates SBA-15 and MCM-14, a Schiff base reaction between the ligand aldehyde and substrate amine groups linked the growing LCP crystals to the silica substrate, resulting in LCPs that were covalently bound to either SBA-15 or MCM-14 (Fig. 33). In the case of Eu-L, there was no change in emission for the surface-mounted LCP, while for Tb-L,

Fig. 33. Synthetic scheme to mount 2D lanthanide CPs on mesoporous silica substrates through the formation of a Schiff base. Reproduced from Ref. [111] with the permission of the Royal Society of Chemistry.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

29

Fig. 34. Demonstration of how white-emitting quantum dot/ZIF-8 thin films are produced. Reproduced with permission from Ref. [62]. Copyright 2017, Wiley-VCH Verlag GmbH &Co. KGaA.

a drop in emission intensity suggested that the formation of the bond to the substrate decreased the efficiency with which the ligand sensitized the lanthanide ion. The composite materials were found to be stable in boiling water and in pH 3–11. In addition to their roles as phosphors, metal–organic frameworks can serve as effective support materials for other phosphors. He and Peng et al. recently synthesized a ZIF-8 film containing red, green, and blue emitting CdSexS1x quantum dots measuring approximately 5 nm in diameter in a single pot reaction, creating a warm white light emitting phosphor [62]. As the quantum dots measure 5 nm across, they cannot be contained within the ZIF-8 pores, but instead form a mixed composite material, with the MOF effectively protecting the quantum dots and preventing aggregation. The composite material film was formed a room temperature, with zinc hydroxide nanostrands reacting with methylimidazole to form the film. As the zinc hydroxide nanostrands carry a strong positive charge, they attract the anionic quantum dots and prevent aggregation while the film is forming, resulting in a highly uniform distribution (Fig. 34). As the luminescence is entirely based on the quantum dots, white light with nearly ideal

CIE coordinates of (0.330, 0.329) and an extremely high quantum yield of 90% can be created by carefully controlling the ratios of red:green:blue-emitting quantum dots present in the ZIF-8 matrix. 7.2. Electroluminescent LMOFs and LCPs One of the first electroluminescent coordination polymers (eLCPs) was reported in 2009 by Anzenbacher and coworkers [112]. The 1D eLCP was created by linking Al(III) bis(8-quinolino late)acetylacetone clusters (Alq) with fluorine oligomers varying between one and nine units long (1a–e; a, b = 1, 3 units, c–e = 5– 9 units). Simple OLED devices were prepared using films of 1a, 1b, and 1c, which were created through a spin-coating process. All three devices gave similar performance, emitting yellow light at 550 nm, with a turn of voltage of approximately 6 V, maximum luminescence of 6000 cd/m2, and a quantum efficiency of 1.2%. An electroluminescent metal–organic framework (eLMOF) that directly produced white light was recently reported [113]. The eLMOF is based on Sr2+ in combination with the semiconductor ligand 1,4,5,8-naphthalenetetracarboxylate (ntca). The resulting

Fig. 35. (left) Electroluminescence spectra of 1 compared with the fitted spectrum, demonstrating the various emission mechanisms at play. (right) Schematic demonstrating the charge transfer processes leading to LC, MC, and MLCT emission. Reproduced with permission from Ref. [113]. Copyright 2016, American Chemical Society.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

30

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

Fig. 36. (a) Framework of NNU-27 viewed along the c axis. (b) Demonstration of the zigzag ligand stacking in NNU-27. (c) 1D metal-carboxylate chain composed of 2:1 Na:Zn that is present in NNU-27. Reproduced from Ref. [114] with the permission of the Royal Society of Chemistry.

framework 1 has a formula of [Sr(ntca)(H2O)2]H2O, and forms 2d sheets which stack under the influence of interlayer p–p interactions, which also provides the conductive avenue through the material. To create a prototype device, 1 single crystals were ground to a powder and then spin-coated onto a 140 nm film of ZnO nanoparticles over 100 nm Ag@Si/SiO2 substrate, and graphene was the top electrode. It produces direct, broad spectrum white light with a peak in the yellow region at 550 nm, CIE coordinates of (0.321, 0.370), and quantum efficiency of 1.2% under a 25 mA injection current. Lifetime analysis indicated that ligandcentered, metal-centered, and metal-to-ligand charge transfer transitions play a role in the observed electroluminescence (Fig. 35). Xing and Wang et al. also utilized interligand p–p interactions to produce an electroluminescent MOF [114]. NNU-27 ([ZnNa2 (L)2(DEF)2]DEF L = 4,40 -(anthracene-9,10 diylbis(ethyne-2,1diyl))-dibenzoate, DEF = diethylformamide) was prepared with the anthracene-based ligand L in the hopes that the extended p–p conjugation would result in a conductive MOF. Upon synthesis, the structure crystallizes into 1D metal-carboxylate chains linked by ligand molecules, with Na+ and Zn2+ in a ratio of 2:1, and neighboring ligand molecules interacting via a zigzag p–p interaction, with face to face distances of 3.420 Å (Fig. 36). The authors prepared 12 simple prototype LED devices consisting of NNU-27 single crystals between two ITO (indium tin oxide) slides, with the NNU-27 layer measuring more than 100 mm in thickness. Electroluminescence turn-on was observed at approximately 27 V, producing orange-red emission with a peak at 575 nm. The authors attribute the exceptionally high turn-on voltage to the thickness of the device, which was not optimized.

8. Conclusions LMOFs and LCPs are extremely promising phosphor materials for energy-saving lighting technologies. Their luminescence properties are vastly tunable through variation of metals, ligands and ligand functionalization, guests, and even excitation energies. LMOF and LCP phosphors have been prepared that have very diverse luminescence behaviors, with efficient colored emission across the entire visible spectrum in addition to direct white light

emission. Blue, green, and yellow emitting LMOFs and LCPs have been synthesized with high quantum yields, with REE free compounds built on highly emissive chromophores in particular often exceeding 90%. Both cold and warm white lights have been directly produced using LMOF and LCP phosphors, with quantum yields of up to 47%. Additionally, many of these phosphor materials have been coated onto LED chips to create functional prototype devices with performance rivalling those of commercially available WLED bulbs. Challenges remain in commercially realizing the promise of these materials. Their stability, especially in the presence of moisture and/or under heat, has historically been one of the greatest liabilities. However, recent work has uncovered strategies to produce luminescent framework materials with exceptionally high stability [16]. By developing LMOFs based on oxophilic, high-valence metals, such as Zr(IV), the chemical bond between the ligand carboxylate and metal center can be greatly strengthened [69–71]. The development of such zirconium-based LMOFs has been emerging rapidly and has already provided the highest-efficiency blueemitting LMOF to date [15]. The synthesis of LMOFs and LCPs has also been cited as posing a challenge, as the organic ligands can be expensive, and scaling up the hydrothermal or solvothermal syntheses of most LMOFs/LCPs can be difficult. However, the cost of organic ligand molecules is currently high because they are usually produced at relatively small scale, and there has been a lack of significant, concerted efforts to scale up the synthesis of high-performance LMOF and LCP phosphor materials. Additionally, new synthetic methods, including mechanochemical synthesis, can provide economically feasible routes to large scale LMOF preparation [115,116]. The quantum yield of low-energy (orange and red-emitting) phosphors also needs to be improved, as they currently lag behind the performance of blue, green, and yellow emitting phosphors and can limit the overall efficiencies of the lighting devices. One approach would be to apply the chromophore-based strategy used to develop some of the highly efficient yellow-emitting LMOFs to the synthesis of high-performance red-emitting LMOFs. Strongly emissive ligands may be designed from red-emitting organic molecules and incorporated into LMOF/LCP backbone to effectively reduce or eliminate non-radiative decays due to intermolecular motions, thus drastically improving their luminescence quantum

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

efficiency. This chromophore-based strategy also has the benefit of decreasing the cost of red LMOF phosphors, as the current highest performing red-emitting LMOFs are based on Eu(III). By replacing the expensive rare-earth metal with an inexpensive metal like zinc, the cost of preparing the LMOF will be reduced. Finally, despite some of unresolved challenges, the exceptionally rapid advancement of research in the field of MOF luminescence bodes well for the future applications of these exciting materials. Acknowledgement We gratefully acknowledge financial support from the National Science Foundation (Grant DMR-1507210). References [1] 2009 Residential Energy Consumption Survey, U.S. Energy Information Administration, https://www.eia.gov/consumption/residential/data/2009/, (accessed 03.24.17). [2] 2016 U.S. primary energy consumption by source and sector, U.S. Energy Information Administration, https://www.eia.gov/totalenergy/data/monthly/ pdf/flow/css_2016_energy.pdf, (accessed 03.24.17). [3] T. Smith, J. Guild, Trans. Opt. Soc. 33 (1931) 73. [4] K. Choi, J. Lee, H.-J. Suk, Appl. Ergon. 52 (2016) 222–231. [5] N. Bardsley, S. Bland, M. Hansen, L. Pattison, M. Pattison, K. Stober, M. Yamada, Solid State Lighting R&D Plan, US Department of Energy, 2015. [6] Z. Bao, G. Chang, H. Xing, R. Krishna, Q. Ren, B. Chen, Energy Environ. Sci. (2016), https://doi.org/10.1039/C6EE01886F. [7] Y. He, W. Zhou, R. Krishna, B. Chen, Chem. Commun. 48 (2012) 11813–11831. [8] A. Morozan, F. Jaouen, Energy Environ. Sci. 5 (2012) 9269–9290. [9] B.-Q. Song, X.-L. Wang, G.-S. Yang, H.-N. Wang, J. Liang, K.-Z. Shao, Z.-M. Su, CrystEngComm 16 (2014) 6882. [10] J. Liu, L. Chen, H. Cui, J. Zhang, L. Zhang, C.-Y. Su, Chem. Soc. Rev. 43 (2014) 6011–6061. [11] F. Ke, Y.-P. Yuan, L.-G. Qiu, Y.-H. Shen, A.-J. Xie, J.-F. Zhu, X.-Y. Tian, L.-D. Zhang, J. Mater. Chem. 21 (2011) 3843–3848. [12] A. Ray Chowdhuri, D. Bhattacharya, S.K. Sahu, Dalton Trans. 45 (2016) 2963– 2973. [13] Y. Yang, Q. Zhao, W. Feng, F. Li, Chem. Rev. 113 (2013) 192–270. [14] K.M.L. Taylor-Pashow, J. Della Rocca, R.C. Huxford, W. Lin, Chem. Commun. 46 (2010) 5832–5849. [15] Z. Wei, Z.-Y. Gu, R.K. Arvapally, Y.-P. Chen, R.N. McDougald, J.F. Ivy, A.A. Yakovenko, D. Feng, M.A. Omary, H.-C. Zhou, J. Am. Chem. Soc. 136 (2014) 8269–8276. [16] W.P. Lustig, S. Mukherjee, N.D. Rudd, A.V. Desai, J. Li, S.K. Ghosh, Chem. Soc. Rev. 46 (2017) 3242–3285. [17] M.D. Allendorf, C.A. Bauer, R.K. Bhakta, R.J.T. Houk, Chem. Soc. Rev. 38 (2009) 1330–1352. [18] Y. Cui, Y. Yue, G. Qian, B. Chen, Chem. Rev. (Washington, DC, U. S.) 112 (2012) 1126–1162. [19] J. Heine, K. Muller-Buschbaum, Chem. Soc. Rev. 42 (2013) 9232–9242. [20] V. Stavila, A.A. Talin, M.D. Allendorf, Chem. Soc. Rev. 43 (2014) 5994–6010. [21] A. Ablet, S.-M. Li, W. Cao, X.-J. Zheng, W.-T. Wong, L.-P. Jin, Chem. - Asian J. 8 (2013) 95–100. [22] L. Chen, C. Yan, M. Pan, H.-P. Wang, Y.-N. Fan, C.-Y. Su, Eur. J. Inorg. Chem. 2016 (2016) 2676–2680. [23] Y. Cui, T. Song, J. Yu, Y. Yang, Z. Wang, G. Qian, Adv. Funct. Mater. 25 (2015) 4796–4802. [24] P. Deria, J. Yu, T. Smith, R.P. Balaraman, J. Am. Chem. Soc. 139 (2017) 5973– 5983. [25] R.F. D’Vries, S. Alvarez-Garcia, N. Snejko, L.E. Bausa, E. Gutierrez-Puebla, A. de Andres, M.A. Monge, J. Mater. Chem. C 1 (2013) 6316–6324. [26] R.-P. Ye, L. Lin, Y.-G. Yao, Inorg. Chim. Acta 453 (2016) 8–15. [27] P.R. Matthes, J. Eyley, J.H. Klein, A. Kuzmanoski, C. Lambert, C. Feldmann, K. Mueller-Buschbaum, Eur. J. Inorg. Chem. 2015 (2015) 826–836. [28] Q. Zhu, C. Shen, C. Tan, T. Sheng, S. Hu, X. Wu, Chem. Commun. (Cambridge, U. K.) 48 (2012) 531–533. [29] R.-P. Ye, X. Zhang, J.-Q. Zhai, Y.-Y. Qin, L. Zhang, Y.-G. Yao, J. Zhang, CrystEngComm 17 (2015) 9155–9166. [30] M. Gutierrez, F. Sanchez, A. Douhal, Phys. Chem. Chem. Phys. 18 (2016) 5112– 5120. [31] F. Wang, W. Liu, S.J. Teat, F. Xu, H. Wang, X. Wang, L. An, J. Li, Chem. Commun. (Cambridge, U. K.) 52 (2016) 10249–10252. [32] J. An, C.M. Shade, D.A. Chengelis-Czegan, S. Petoud, N.L. Rosi, J. Am. Chem. Soc. 133 (2011) 1220–1223. [33] Y. Gai, F. Jiang, L. Chen, M. Wu, K. Su, J. Pan, X. Wan, M. Hong, Cryst. Growth Des. 14 (2014) 1010–1017. [34] S. Freslon, Y. Luo, C. Daiguebonne, G. Calvez, K. Bernot, O. Guillou, Inorg. Chem. 55 (2016) 794–802. [35] D.T. De Lill, A. De Bettencourt-Dias, C.L. Cahill, Inorg. Chem. 46 (2007) 3960– 3965.

31

[36] L. Jia, Y.-C. Hui, Z. Li, H.-L. Sun, Z. Wang, CrystEngComm 16 (2014) 6483– 6490. [37] X. Liu, S. Akerboom, S.H.C. Askes, I. Mutikainen, E. Bouwman, Inorg. Chem. Commun. 61 (2015) 60–63. [38] M. Sobieray, J. Gode, C. Seidel, M. Poss, C. Feldmann, U. Ruschewitz, Dalton Trans. 44 (2015) 6249–6259. [39] J.-X. Ma, X.-F. Huang, X.-Q. Song, W.-S. Liu, Chem. Eur. J. 19 (2013) 3590– 3595. [40] S. Zou, Q. Li, S. Du, RSC Adv. 5 (2015) 34936–34941. [41] L. Ma, O.R. Evans, B.M. Foxman, W. Lin, Inorg. Chem. 38 (1999) 5837–5840. [42] F. Wang, Z. Zhou, W. Liu, L. Zhou, L. Chen, J. Li, Dalton Trans. 46 (2017) 956– 961. [43] B.J. Deibert, E. Velasco, W. Liu, S.J. Teat, W.P. Lustig, J. Li, Cryst. Growth Des. 16 (2016) 4178–4182. [44] Z. Hu, G. Huang, W.P. Lustig, F. Wang, H. Wang, S.J. Teat, D. Banerjee, D. Zhang, J. Li, Chem. Commun. (Cambridge, U. K.) 51 (2015) 3045–3048. [45] Q. Gong, Z. Hu, B.J. Deibert, T.J. Emge, S.J. Teat, D. Banerjee, B. Mussman, N.D. Rudd, J. Li, J. Am. Chem. Soc. 136 (2014) 16724–16727. [46] O. Toma, M. Allain, F. Meinardi, A. Forni, C. Botta, N. Mercier, Angew. Chem., Int. Ed. 55 (2016) 7998–8002. [47] Y. Xu, L. Li, S. Zhang, S. Zhao, J. Luo, Cryst. Growth Des. 16 (2016) 406–411. [48] D.F. Sava Gallis, L.E.S. Rohwer, M.A. Rodriguez, T.M. Nenoff, Chem. Mater. 26 (2014) 2943–2951. [49] W. Xu, C.-J. Zhang, H. Wang, Y. Wang, J. Cluster Sci. 28 (2017) 2005–2015. [50] M. Yamamoto, T. Nakanishi, Y. Kitagawa, K. Fushimi, Y. Hasegawa, Mater. Lett. 167 (2016) 183–187. [51] Y.-N. Gong, X.-W. Wei, L.-H. He, Y.-Q. Wen, Y.-W. Ren, Y.-R. Xie, D.-C. Zhong, Dalton Trans. 46 (2017) 7403–7407. [52] R.-S. Liu, V. Drozd, N. Bagkar, C.-C. Shen, I. Baginskiy, C.-H. Chen, C.H. Tan, J. Electrochem. Soc. 155 (2008) P71–P73. [53] C.-Y. Sun, X.-L. Wang, X. Zhang, C. Qin, P. Li, Z.-M. Su, D.-X. Zhu, G.-G. Shan, K.Z. Shao, H. Wu, J. Li, Nat. Commun. 4 (2013), 3717/3711–3717/3718. [54] M.-S. Wang, S.-P. Guo, Y. Li, L.-Z. Cai, J.-P. Zou, G. Xu, W.-W. Zhou, F.-K. Zheng, G.-C. Guo, J. Am. Chem. Soc. 131 (2009) 13572–13573. [55] J. Chen, Q. Zhang, Z.-F. Liu, S.-H. Wang, Y. Xiao, R. Li, J.-G. Xu, Y.-P. Zhao, F.-K. Zheng, G.-C. Guo, Dalton Trans. 44 (2015) 10089–10096. [56] W. Xie, J.-S. Qin, W.-W. He, K.-Z. Shao, Z.-M. Su, D.-Y. Du, S.-L. Li, Y.-Q. Lan, Inorg. Chem. Front. 4 (2017) 547–552. [57] X. Gao, S. Chang, H. Liu, Z. Liu, Eur. J. Inorg. Chem. 2016 (2016) 2837–2842. [58] Q. Tang, S. Liu, Y. Liu, D. He, J. Miao, X. Wang, Y. Ji, Z. Zheng, Inorg. Chem. 53 (2014) 289–293. [59] J. Roncali, Macromol. Rapid Commun. 28 (2007) 1761–1775. [60] B. Gui, N. Yu, Y. Meng, F. Hu, C. Wang, J. Polym. Sci., Part A: Polym. Chem. 55 (2017) 1809–1817. [61] L.V. Meyer, F. Schoenfeld, K. Mueller-Buschbaum, Chem. Commun. (Cambridge, U. K.) 50 (2014) 8093–8108. [62] W. Ying, Y. Mao, X. Wang, Y. Guo, H. He, Z. Ye, S.-T. Lee, X. Peng, ChemSusChem 10 (2017) 1346–1350. [63] B. Zhao, N. Li, X. Wang, Z. Chang, X.-H. Bu, A.C.S. Appl, Mater. Interfaces 9 (2017) 2662–2668. [64] C.-X. Chen, Q.-K. Liu, J.-P. Ma, Y.-B. Dong, J. Mater. Chem. 22 (2012) 9027– 9033. [65] W.P. Lustig, F. Wang, S.J. Teat, Z. Hu, Q. Gong, J. Li, Inorg. Chem. 55 (2016) 7250–7256. [66] D. Farrusseng, Metal-Organic Frameworks: Applications from Catalysis to Gas Storage, Wiley-VCH, 2011. [67] S. Kaskel, The Chemistry of Metal-Organic Frameworks: Synthesis, Characterization, and Applications, Wiley-VCH, 2016. [68] G.-P. Yong, Y.-Z. Li, C.-F. Li, Y.-M. Zhang, W.-L. She, Dalton Trans. 40 (2011) 4131–4139. [69] N.C. Burtch, H. Jasuja, K.S. Walton, Chem. Rev. 114 (2014) 10575–10612. [70] J. Canivet, A. Fateeva, Y. Guo, B. Coasne, D. Farrusseng, Chem. Soc. Rev. 43 (2014) 5594–5617. [71] C. Wang, X. Liu, N. Keser Demir, J.P. Chen, K. Li, Chem. Soc. Rev. 45 (2016) 5107–5134. [72] K. Binnemans, Coord. Chem. Rev. 295 (2015) 1–45. [73] B. Grobelna, A. Synak, P. Bojarski, Opt. Appl. 42 (2012) 337–344. [74] M. Hatanaka, S. Yabushita, Theor. Chem. Acc. 133 (2014) 1–15. [75] F.J. Steemers, W. Verboom, D.N. Reinhoudt, E.B. van der Tol, J.W. Verhoeven, J. Am. Chem. Soc. 117 (1995) 9408–9414. [76] S. Sivakumar, M.L.P. Reddy, A.H. Cowley, R.R. Butorac, Inorg. Chem. 50 (2011) 4882–4891. [77] L. Shi, B. Hong, W. Guan, Z. Wu, Z. Su, J. Phys. Chem. A 114 (2010) 6559–6564. [78] S.-M. Li, X.-J. Zheng, D.-Q. Yuan, A. Ablet, L.-P. Jin, Inorg. Chem. 51 (2012) 1201–1203. [79] Z.-F. Wu, B. Tan, Z.-L. Xie, J.-J. Fu, X.-Y. Huang, J. Mater. Chem. C 4 (2016) 2438–2441. [80] R. Li, S.-H. Wang, Z.-F. Liu, X.-X. Chen, Y. Xiao, F.-K. Zheng, G.-C. Guo, Cryst. Growth Des. 16 (2016) 3969–3975. [81] D.T. de Lill, N.S. Gunning, C.L. Cahill, Inorg. Chem. 44 (2005) 258–266. [82] A.R. Ramya, D. Sharma, S. Natarajan, M.L.P. Reddy, Inorg. Chem. 51 (2012) 8818–8826. [83] Q. Tang, Y. Liu, S. Liu, D. He, J. Miao, X. Wang, G. Yang, Z. Shi, Z. Zheng, J. Am. Chem. Soc. 136 (2014) 12444–12449. [84] Y.-S. Wei, X.-P. Hu, Z. Han, X.-Y. Dong, S.-Q. Zang, T.C.W. Mak, J. Am. Chem. Soc. 139 (2017) 3505–3512.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017

32

W.P. Lustig, J. Li / Coordination Chemistry Reviews xxx (2017) xxx–xxx

[85] H.-N. Li, H.-Y. Li, L.-K. Li, L. Xu, K. Hou, S.-Q. Zang, T.C.W. Mak, Cryst. Growth Des. 15 (2015) 4331–4340. [86] J. Cepeda, S. Perez-Yanez, G. Beobide, O. Castillo, J.A. Garcia, M. Lanchas, A. Luque, Dalton Trans. 44 (2015) 6972–6986. [87] B. Chen, Y. Yang, F. Zapata, G. Qian, Y. Luo, J. Zhang, E.B. Lobkovsky, Inorg. Chem. 45 (2006) 8882–8886. [88] X. Rao, Q. Huang, X. Yang, Y. Cui, Y. Yang, C. Wu, B. Chen, G. Qian, J. Mater. Chem. 22 (2012) 3210–3214. [89] S. Dang, J.-H. Zhang, Z.-M. Sun, J. Mater. Chem. 22 (2012) 8868–8873. [90] J. Li, J.-G. Li, S. Liu, X. Li, X. Sun, Y. Sakka, J. Mater. Chem. C 1 (2013) 7614– 7622. [91] Y. Liu, M. Pan, Q.-Y. Yang, L. Fu, K. Li, S.-C. Wei, C.-Y. Su, Chem. Mater. 24 (2012) 1954–1960. [92] Y. Zhou, B. Yan, Nanoscale 7 (2015) 4063–4069. [93] R. Zhao, L. Mei, K.-Q. Hu, L. Wang, Z.-F. Chai, W.-Q. Shi, Eur. J. Inorg. Chem. 2017 (2017) 1472–1477. [94] P.O. Adelani, P.C. Burns, Inorg. Chem. 51 (2012) 11177–11183. [95] W. Yang, H. Wang, W.-G. Tian, J. Li, Z.-M. Sun, Eur. J. Inorg. Chem. 2014 (2014) 5378–5384. [96] Y. Li, Z. Weng, Y. Wang, L. Chen, D. Sheng, Y. Liu, J. Diwu, Z. Chai, T.E. AlbrechtSchmitt, S. Wang, Dalton Trans. 44 (2015) 20867–20873. [97] D.L. Reger, A.P. Leitner, M.D. Smith, Cryst. Growth Des. 16 (2016) 527–536. [98] H.-Y. Wu, R.-X. Wang, W. Yang, J. Chen, Z.-M. Sun, J. Li, H. Zhang, Inorg. Chem. 51 (2012) 3103–3107. [99] Z.-F. Chen, R.-G. Xiong, J. Zhang, X.-T. Chen, Z.-L. Xue, X.-Z. You, Inorg. Chem. 40 (2001) 4075–4077. [100] C.A. Bauer, T.V. Timofeeva, T.B. Settersten, B.D. Patterson, V.H. Liu, B.A. Simmons, M.D. Allendorf, J. Am. Chem. Soc. 129 (2007) 7136–7144. [101] N.B. Shustova, B.D. McCarthy, M. Dinca, J. Am. Chem. Soc. 133 (2011) 20126– 20129.

[102] Z.-F. Wu, B. Tan, Z.-H. Deng, Z.-L. Xie, J.-J. Fu, N.-N. Shen, X.-Y. Huang, Chem. Eur. J. 22 (2016) 1334–1339. [103] J. He, M. Zeller, A.D. Hunter, Z. Xu, J. Am. Chem. Soc. 134 (2012) 1553–1559. [104] F. Luo, M.-S. Wang, M.-B. Luo, G.-M. Sun, Y.-M. Song, P.-X. Li, G.-C. Guo, Chem. Commun. (Cambridge, U. K.) 48 (2012) 5989–5991. [105] Y. Wang, Z. Xue, Z. Sun, C. Tan, Y. Wen, S. Hu, Q. Zhu, T. Sheng, X. Wu, Dalton Trans. 44 (2015) 2217–2222. [106] C. Xu, A.M. Kirillov, Y. Shu, Y. Liu, L. Guo, L. Yang, W. Dou, W. Liu, C. Chen, X. Huang, J. Zhang, W. Liu, CrystEngComm 18 (2016) 1190–1199. [107] A.K. Chaudhari, M.R. Ryder, J.-C. Tan, Nanoscale 8 (2016) 6851–6859. [108] X.-Y. Xu, B. Yan, Dalton Trans. 44 (2015) 1178–1185. [109] W. Xie, W.-W. He, D.-Y. Du, S.-L. Li, J.-S. Qin, Z.-M. Su, C.-Y. Sun, Y.-Q. Lan, Chem. Commun. (Cambridge, U. K.) 52 (2016) 3288–3291. [110] E. Feijo de Melo, N. da C. Santana, K.G. Bezerra Alves, G.F. de Sa, C. Pintode Melo, M.O. Rodrigues, S.A. Junior, J. Mater. Chem. C 1 (2013) 7574–7581. [111] J. Wang, W. Dou, A.M. Kirillov, W. Liu, C. Xu, R. Fang, L. Yang, Dalton Trans. 45 (2016) 18610–18621. [112] V.A. Montes, G.V. Zyryanov, E. Danilov, N. Agarwal, M.A. Palacios, P. Anzenbacher, J. Am. Chem. Soc. 131 (2009) 1787–1795. [113] G. Haider, M. Usman, T.-P. Chen, P. Perumal, K.-L. Lu, Y.-F. Chen, ACS Nano 10 (2016) 8366–8375. [114] D. Chen, H. Xing, Z. Su, C. Wang, Chem. Commun. (Cambridge, U. K.) 52 (2016) 2019–2022. [115] P.A. Julien, K. Uzˇarevic´, A.D. Katsenis, S.A.J. Kimber, T. Wang, O.K. Farha, Y. Zhang, J. Casaban, L.S. Germann, M. Etter, R.E. Dinnebier, S.L. James, I. Halasz, T. Frišcˇic´, J. Am. Chem. Soc. 138 (2016) 2929–2932. [116] A.K. Das, R.S. Vemuri, I. Kutnyakov, B.P. McGrail, R.K. Motkuri, Sci. Rep. 6 (2016) 28050.

Please cite this article in press as: W.P. Lustig, J. Li, Luminescent metal–organic frameworks and coordination polymers as alternative phosphors for energy efficient lighting devices, Coord. Chem. Rev. (2017), https://doi.org/10.1016/j.ccr.2017.09.017