Colored polymer-reinforced metal-organic framework microparticles with high charge-to-mass ratio for electrophoretic display
Jiamin Cheng, Mian Qin, Wenhao Wang, Jingxing Zhang, Yao Wang, Pengfei Bai, Hao Li, Guofu Zhou
TL;DR
Researchers developed colorful microparticles using metal-organic frameworks to improve electrophoretic displays with faster response and better stability.
Contribution
Colored MOF microparticles reinforced with PEI offer superior electrophoretic performance and stability for color displays.
Findings
MOF microparticles with PEI modification showed high Zeta potential and electrophoretic mobility.
Response and recovery times were under 2.3 and 5.9 seconds under low field strength.
Colored MOF microparticles outperformed traditional pigments in stability and display quality.
Abstract
Besides high porosity and controllable structure, metal-organic framework (MOF) has some natural advantages for color electrophoretic particles: easy modification, controllable morphology, low and adjustable density, and high charge density, as well as rich, vivid, and stable colors. Therefore, we first integrated the four colored MOF microparticles and polyelectrolytes into the blue, reddish-brown, green, and purple electrophoretic particles to construct the two-color stable dispersions in nonpolar isododecane as the electrophoretic inks. Here, the surface modification of polyethyleneimine (PEI) chains based on non-covalent interaction rendered these MOF microparticles fully reinforced to ~+30 mV in Zeta potential and over 3.6 × 10−10 m2 V−1 s−1 in the electrophoretic mobility. Under the ultralow field strength of 0.02 V μm−1, all the response time and recovery time were no more than…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9- —Startup Foundation from SCNU
- —https://doi.org/10.13039/100012541Guangdong Innovative and Entrepreneurial Research Team Program
- —Guangdong Provincial Key Laboratory of Optical Information Materials and Technology (No. 2023B1212060065), MOE International Laboratory for Optical Information Technologies, and the 111 Project.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsConducting polymers and applications · Covalent Organic Framework Applications · TiO2 Photocatalysis and Solar Cells
Introduction
As a promising type of porous crystalline hybrids, metal-organic frameworks (MOFs) are assembled by coordination between inorganic metal ions and organic ligands^1,2^. The variation of the two key components enables MOFs to be infinitely developed into diverse chemical and physical properties^3,4^. So besides high specific surface area and porosity, MOFs feature both controllable structure and function, and have been widely used in the fields of catalysis, adsorption, separation, sensing, drug delivery, and so on^5,6^.
In particular, MOFs exhibit unique charge characteristics of inorganic-organic hybrids. The overall charge of the framework can be tuned with the oxidation state of metal ions within the ligand^7^, ZIF-8, ZIF-67, MIL-101(Cr), and MOF-801 were proved to be positively charged^8^. From the perspective of electrophoretic particle, MOFs have few natural advantages: (1) more options for organic functional groups to allow exquisite modification^2,9^; (2) good morphology controllability, including wide variation of sizes ranging from micrometer to nanometer scale, and structures changing from three to zero dimensional^10^; (3) low and adjustable density, for example, the proven density of HKUST-1 changing from 0.883, 1.07, to 2.8 g cm^−1^ (refs. ^3,11,12^). The third virtue is very beneficial for improving the migration rate and multiparticle resolution^13^. In addition, rich, vivid, and stable colors of MOFs provide more possibilities for fine patterning and color electrophoretic display (EPD). Undoubtedly, MOFs are much lighter and more colorful than other inorganic electrophoretic particles (e.g. metal ion-doping TiO_2_ nanoparticle)^14,15^, and much easier to be modified, and more stable in color than common organic ones (e.g. color dye/pigment particles)^13,16,17^. But MOFs still need less aggregation, higher charge density, better stability and compatibility, to meet the requirements of high-performance display^17^. In general, a good way is polymer modification on the MOF surface via. covalent anchoring or physical adsorption^8,18^.
In our work, 1,3,5-benzenetricarboxylic acid (BTC) as a common organic ligand was coordinated with Cu^2+^, Fe^3+^, Ni^2+^, and Co^2+^ ion by solvothermal method, to form the blue, reddish-brown, green, and purple MOF (M-BTC) microparticles (i.e., Cu-BTC, Fe-BTC, Ni-BTC and Co-BTC), respectively (see Fig. 1). And these resulting M-BTC microparticles were further reinforced by physically adsorbed polyethyleneimine (PEI) for stronger electrophoretic performances. Here, PEI with a high amine density is easily localized on the surface of the MOF microparticle through electrostatic interaction and hydrogen bonding, and endows the MOF surface with abundant positively-charged amino groups for higher charge density. In addition, isododecane, an oil phase with ultralow polarity was chosen as electrophoresis medium, to effectively avoid possible MOF decomposition/transformation.Fig. 1. Preparation process of EPD sample.Schematic diagrams on the preparation routes of the colored PEI-reinforced MOF microparticles (a), and the fabrication process of the electrophoretic display cell (b)
Results
As a truly mainstream display with a reflective paper-like effect, EPD has some distinct advantages, e.g. quick response, bistable state, high reliability, wide viewing angle, and low cost and power consumption^19–21^. Currently, black-and-white EPD has gained widespread recognition and commercialization, but it faces significant limitations in terms of color variety and update time. The two latest color products of E-Ink’s, Spectra^TM^ 6 and Gallery^TM^ 3, have not yet achieved practical video effects^22^. Although it is technically easy to colorize EPD using color filters, it comes at the expense of display brightness and resolution, as well as device thickness^23^. Therefore, the major challenge in further developing color EPDs lies in precisely controlling multiple colors, charges, and motions of electrophoretic particles under a low electric field ^17^.
A series of innovational colored electrophoretic particles has been developed for the first time by integrating polyelectrolyte and lightweight, colorful MOF microparticles into charged colloid particles. These MOFs represent four metals widely used in MOF synthesis: Cu, Fe, Ni, Co. Scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images (Fig. 2 and Fig. S1) demonstrated the morphology of M-BTC and M-BTC-PEI. As shown in Fig. 2a, b, those crude Cu-BTC microparticles were regular octahedra with uniform size of about 700 nm, and those PEI-modified Cu-BTC-PEI microparticles presented in similar but adherent morphology. In Fig. 2c, d, the blank Fe-BTC microparticles took a large agglomerate look comprising many irregular polyhedrons with a size of about 100 nm, and did not change much after PEI modification. It can be seen from in Fig. 2e, f that, those original Ni-BTC microparticles appeared in rough microsphere with an average size of about 2 µm, and those PEI-modified ones almost remained unchanged. By comparison, both the Co-BTC and the Co-BTC-PEI microparticles were nano/micron disordered nanosheets in Fig. 2g, h. Evidently, all the above-mentioned MOF microparticles were completely consistent with what has been previously reported^24–27^. And there was almost no change in morphology before and after PEI modification, indicating that the adsorbed PEI chains were limited. This was also confirmed by the energy dispersive spectroscopy (EDS) of Cu-BTC & Cu-BTC-PEI (see Fig. 2i, j) and Ni-BTC & Ni-BTC-PEI (see Fig. S2). Hereinto, a slight increase in nitrogen content showed the successful PEI modification on these M-BTC microparticles.Fig. 2. Morphological characterization of electrophoretic particles.SEM images of the resulting Cu-BTC (a), Fe-BTC (c), Ni-BTC (e), and Co-BTC microparticles (g), and the corresponding PEI-reinforced Cu-BTC-PEI (b), Fe-BTC-PEI (d), Ni-BTC-PEI (f), and Co-BTC-PEI (h) microparticles; EDS elemental maps of the resulting Cu-BTC (i) and Cu-BTC-PEI sample (j). Notes: PEI M.W., 10,000 g mol^−1^; PEI feed ratio, 30 wt.%
The x-ray powder diffraction (XRD) patterns of the resulting M-BTC and M-BTC-PEI microparticles in the 2θ angle region at 5–70◦ were illustrated in Fig. 3a. Apparently, all the diffractograms before and after PEI modification are almost identical and in good agreement with the results previously reported in the literature, showing good crystalline^27–29^. These major diffraction peaks were listed as below: Cu-BTC (CCDC: 846570), 6.8° (200), 9.5° (220), 11.63° (222), 13.53° (400), 14.7° (331), 15.03° (420), 16.47° (422), 17.52° (333), 19.11° (440), 20.26° (600), 24.11° (711), 26.10° (553), 29.37° (662), and 35.3° (951); Fe-BTC (CCDC: 640536), 7.1° (440), 10.3° (660), 11.07° (428), 12.65° (1022), 14.3° (088), 20.16°(4814), 24.1° (6618) and 27.78°(9321); Ni-BTC (CCDC: 1990438), 12.2° (2-10), 15.33° (002), 18.0° (300), 20.9°, 22.32°, 23.88°, 26.13° (2-13), and 27.9° (5-10); Co-BTC (CCDC: 1274034), 2θ value of 10.92° (200), 14.6° (001), 18.8° (111), 20.1° (021), 24.69° (13-1), 25.36° (42-1), 27.19° (20-2), 33.9° (62-1), 35.43◦ (440).Fig. 3. Characterization on chemical structures and compositions of the electrophoretic particles.Powder XRD patterns (a), FT-IR spectra (b), and ATR-FTIR spectra (100 °C) of the resulting M-BTC and M-BTC-PEI microparticles (c–f); high-resolution XPS spectra (g) and the corresponding O 1s spectra of the resulting Cu-BTC and Cu-BTC-PEI microparticles (h). Notes: PEI M.W., 10,000 g mol^−1^; PEI feed ratio, 30 wt.%
The Fig. 3b exhibited the characteristic functional groups of different M-BTC and M-BTC-PEI composites. In these Fourier-transform infrared (FT-IR) spectra, a few similar peaks at 750, 1360–1450, and 1550–1670 cm^−1^, were attributed to symmetric, asymmetric stretching vibrations of C=C, C–O, and C=O bonds in the BTC ligands, respectively. And the representative absorption peaks at 725, 615, 720, and 720 cm^−1^ corresponded to the vibrational peaks of the intrinsic coordination bonds between oxygen atom (O) and metal ion (i.e., Cu^2+^, Fe^3+^, Ni^2+^, and Co^2+^). But significant difference before and after PEI modification cannot be found, which may be attributed to the absorption overlap of the amino groups (N-H; 3100–3500 cm^−1^) with the hydroxy groups (−OH) of residual alcohol or water (300–3500 cm^−1^). High-temperature tests were performed by using variable temperature attenuated total reflectance-Fourier transform infrared spectroscopy (ATR-FTIR), to eliminate possible interference from the residual^30^. Just as shown in Fig. 3c–f, all the N-H peaks of the resulting M-BTC-PEI microparticles still emerged clearly in the range from 3100 to 3400 cm^−1^, indicating successful PEI adsorption on the surface of M-BTC. In addition, Fig. 3g, h further exhibited the chemical states and compositions of Cu-BTC and Cu-BTC-PEI microparticles. Obviously, the C 1s, O 1s, and Cu 2p peak kept constant, and the N 1s peak was enhanced after PEI modification, indicating the successful bonding of PEI chains. Accordingly, the interaction between –NH group and Cu^2+^ ion reduced the binding energy of Cu 2p (see Fig. S3), similar to the previously reported results^31^. In addition, the binding energies of carbonyl (C=O) and carbon-oxygen bonds (C–O) in the BTC ligands rose by 0.59 eV. This may be attributed to the newly formed hydrogen bonds between oxygen atom and amino groups of PEI chain (N–H ∙ ∙ ∙ O), as well as another coordination interaction between the O atoms and Cu^2+^ (refs. ^18,32^).
The thermogravimetric (TGA) curves of the resulting M-BTC and M-BTC-PEI microparticles were recorded in Fig. 4a–d. Apparently, all the M-BTC microparticles before and after PEI modification, behaved with almost the same residual masses but the different thermolytic behaviors. The first stage of weight loss below 150 °C undoubtedly originates from the removal of the residual solvent (e.g. water, ethanol, and DMF). The differences in thermolysis are mainly manifested in the subsequent process (see Fig. S4). It is well-known that the decomposition range of the crude BTC molecules is normally between 300 and 350 °C, and broadened to higher temperatures after coordination with metal ions^33–35^. This is embodied well in the TGA curves of these M-BTC microparticles above 300 °C. The thermolysis of PEI also occurs in the temperature range between 300 and 350 °C (refs. ^36–38^). Hence, in the main stage of weight loss between 150 and 500 °C, those modified PEI chains and their resulting heat resistance visibly made M-BTC-PEI microparticles decompose slightly faster than the crude M-BTC microparticles. Although the accurate mass of PEI in the M-BTC-PEI microparticles was very difficult to determine, its existence can be confirmed.Fig. 4TGA curves of the electrophoretic particles.TGA curves of the resulting M-BTC and M-BTC-PEI microparticles in oxygen atmospheres: Cu-BTC and Cu-BTC-PEI (a), Fe-BTC and Fe-BTC-PEI (b), Ni-BTC and Ni-BTC-PEI (c), and Co-BTC and Co-BTC-PEI (d). Notes: PEI M.W., 10,000 g mol^−1^; PEI feed ratio, 30 wt.%
As the key element of electrophoretic particle above all else, Zeta potential is defined as the potential drop in the diffuse layer of ions surrounding the charged particle^17^, and mainly depends on surface composition of charged particle, used medium, surfactant type, and concentration^39^. In the case of Zeta potential greater than +30 mV or less than −30 mV, charged particle is considered stable. And the magnitude of Zeta potential is positively correlated with electrophoretic mobility^13,39^. Based on the above research basis, we systematically investigated the Zeta potential and electrophoretic mobility of the resulting M-BTC and M-BTC-PEI microparticles. Of course, the reinforcing effect of PEI modification is more reflected from the colloidal nature of the M-BTC-PEI microparticle. In Fig. 5a–d and S5, all the Zeta potentials of the M-BTC microparticles were around −10 mV, which was caused by the exposed carboxyl groups on the MOF surface. To M-BTC-PEI microparticles, the abundant protonated amino groups of the linked PEI chains inevitably enabled these microparticles to convert into being positively charged. At the same PEI feed of 20 wt.%, different molecular weights of PEI brought about different Zeta potentials. Hereinto, PEI with the molecular weight of 10,000 g mol^−1^ was viewed as the best option, whose corresponding Zeta potential was slightly higher than the ones’ with the molecular weight of 25,000 g mol^−1^. It seems that, larger random coil of the latter may be unfavorable for PEI bonding and adsorption. As the molecular weights of PEI was settled as 10,000 g mol^−1^, all the maximum Zeta potential of the M-BTC-PEI microparticles emerged around +30 mV at the same PEI feed of 30 wt.%. Similarly, higher feed above the saturated bonding point may not be helpful for improving surface potential. Meanwhile, such variation in Zeta potential also appeared in mobility (see Table 1). Naturally, this surface PEI decoration consequently made these M-BTC microparticles bigger with wider size distribution, as shown in Fig. S6. The optimized parameters of PEI modification for those electrophoretic M-BTC-PEI microparticles were chosen as: the molecular weight of PEI was 10,000 g mol^−1^, and the feed of PEI was 30 wt.%.Fig. 5. Characterization on surface potential changes of the electrophoretic particles.Zeta potentials and electrophoretic mobilities of the crude M-BTC microparticles and the reinforced M-BTC-PEI microparticles with different molecular weights and feed ratios of PEI in isododecane: Cu-BTC and Cu-BTC-PEI (a), Fe-BTC and Fe-BTC-PEI (b), Ni-BTC and Ni-BTC-PEI (c), and Co-BTC and Co-BTC-PEI (d); AFM (1) and KPFM images on the silicon wafer (2), corresponding line-scanning surface height (3) and potential (4) of the resulting Cu-BTC (e) and Cu-BTC-PEI microparticles (f)Table 1. Particle characteristics of blank PEI chains, M-BTC and M-BTC-PEI microparticles in isododecaneSampleZeta potential (mV)Mobility (×10^−8^m^2^ V^−1^ s^−1^)Effective Size (nm)PolydispersityPEI 180010.31 ± 1.070.014 ± 0.0013.51 ± 0.210.525 ± 0.086PEI 1 W13.51 ± 1.32−0.018 ± 0.0019.06 ± 0.030.225 ± 0.003PEI 2.5 W15.41 ± 1.270.020 ± 0.0019.41 ± 0.040.221 ± 0.024Cu-BTC−14.89 ± 1.026−0.010 ± 0.001738.36 ± 9.880.265 ± 0.025Cu-BTC-PEI 30%27.51 ± 1.830.036 ± 0.003782.44 ± 12.960.282 ± 0.031Fe-BTC−9.50 ± 1.19−0.010 ± 0.001734.84 ± 6.520.268 ± 0.026Fe-BTC-PEI 30%30.57 ± 2.480.041 ± 0.002806.87 ± 13.640.330 ± 0.034Ni-BTC−13.08 ± 1.20−0.016 ± 0.0031890.94 ± 29.560.351 ± 0.021Ni-BTC-PEI 30%36.11 ± 2.900.047 ± 0.0032018.78 ± 74.410.385 ± 0.067Co-BTC−3.91 ± 1.93−0.009 ± 0.0011270.70 ± 10.220.296 ± 0.031Co-BTC-PEI 30%35.54 ± 1.690.047 ± 0.011307.40 ± 22.400.321 ± 0.013Note: In M-BTC-PEI 30%, PEI denotes the bonded PEI chain with the molecular weight of 10,000 g mol^−1^, and 30% did the feed ratio of PEI
To obtain more visual evidence, the surface potential changes in the resulting Cu-BTC and Cu-BTC-PEI microparticles were measured using atomic force microscopy (AFM) and Kelvin probe force microscopy (KPFM)^40^. Just shown in Fig. 5e1, e3, f1, and f3, the average height differences were 321 nm (Cu-BTC; Ra: 53 nm) and 298 nm (Cu-BTC-PEI; Ra: 60 nm), respectively, indicating no significant changes in morphology or size before and after PEI loading of Cu-BTC microparticle. In Fig. 5e2, e4, f2, and f4, the tip potential was considered as the local contact potential difference (CPD) reference electrode, to enable relative CPD comparisons recorded on the two different functionalized regions of the sample (ΔCPD)^40,41^. Herein, the ΔCPD value of the modified Cu-BTC-PEI microparticles (26.9 mV) was clearly greater than that of the crude Cu-BTC microparticles (21.5 mV), which must be ascribed to the reinforcing effect of surface PEI loading.
In addition, we attempted to utilize negatively charged polyacrylic acid (PAA) and uncharged polyethylene glycol (PEG) to modify the surface of the M-BTC microparticles using the same non-covalent bonding. As shown in Fig. S7, plentiful carboxyl groups of the bonded PAA chains further rendered those resulting M-BTC-PAA microparticles more negatively charged to about −30 mV. Similarly, plentiful hydroxyl groups of the bonded PEG chains only led to a slight decrement of the Zeta potential of those resulting M-BTC-PEG microparticles. It is a sign that this non-covalent “adsorption” (e.g. hydrogen bonding and electrostatic interaction) is suitable for various polymers, and applicable for adjusting the charge properties of the MOF surface^42^.
Subsequently, these PEI-reinforced MOF microparticles were dispersed with an acidic charge control agent (polyisobutenyl succinic anhydrides, PIBSA) into isododecane to formulate a series of colored electrophoretic inks. Figure 6a showed that all the reflectance spectra of the M-BTC-PEI inks agreed well with the real colors of the M-BTC-PEI microparticles. We further calculated their tristimulus values (i.e., X, Y, and Z) according to the following formula (Eq. 1–5) and converted into chrominance coordinates^43^.
\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$X={\rm{\kappa }}\mathop{\sum }\limits_{\lambda }{{\rm{\varphi }}}_{{\rm{\lambda }}}\left({\rm{\lambda }}\right)\bar{{\rm{\chi }}}\left({\rm{\lambda }}\right)\Delta {\rm{\lambda }}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Y={\rm{\kappa }}\mathop{\sum }\limits_{\lambda }{{\rm{\varphi }}}_{{\rm{\lambda }}}\left({\rm{\lambda }}\right)\bar{{\mathcal{Y}}}\left({\rm{\lambda }}\right)\Delta {\rm{\lambda }}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Z={\rm{\kappa }}\mathop{\sum }\limits_{\lambda }{{\rm{\varphi }}}_{{\rm{\lambda }}}\left({\rm{\lambda }}\right)\bar{{\mathcal{Z}}}\left({\rm{\lambda }}\right)\Delta {\rm{\lambda }}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\rm{\kappa }}=\frac{100}{\mathop{\sum }\limits_{\lambda }{\rm{S}}\left({\rm{\lambda }}\right)\bar{{\mathcal{Y}}}\left({\rm{\lambda }}\right)\Delta {\rm{\lambda }}}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\rm{\varphi }}\left({\rm{\lambda }}\right)={\rm{R}}\left({\rm{\lambda }}\right){\rm{S}}\left({\rm{\lambda }}\right)$$\end{document}where k is the normalization constant, Δλ is the wavelength interval (1 nm), S(λ) is the relative spectral distribution of the light source (D65), R(λ) is the reflectance of the sample under this light source, φ(λ) and φλ(λ) are the relative color stimulus function and the color stimulus function, respectively, and φ(λ) can be used instead of φλ(λ) for the reflected object color. So, we calculated out k and XYZ based on the standard value of S(λ) and the measured R(λ), and obtained the color matching functions of the CIE 1931 standard chromaticity observer (i.e., x, y, and z) by the corresponding formula, x = X/X + Y + Z, y = Y/X + Y + Z, and z = Z/X + Y + Z. All the results, including the detailed CIE coordinate parameters, were finally plotted in Fig. 6b, which also corresponds well to the real colors of the M-BTC-PEI microparticles.Fig. 6. Optical properties of the electrophoretic particles.Reflectance curves (a) and color chromaticity diagrams of the M-BTC isododecane solution (M-BTC concentration: 100 mg mL^−1^) encapsulated inside the electrophoretic display cell (b; the object figures of the used M-BTC microparticles at the top right); UV–vis spectra of the Cu-BTC-PEI isododecane solution (Cu-BTC concentration: 10 mg mL^−1^) with the settled PIBSA concentration of 50 mg mL^−1^ (c) and the one without any PIBSA (d), and corresponding object figures (e; left, containing PIBSA; right, no PIBSA). Notes: PEI M.W., 10,000 g mol^−1^; PEI feed ratio, 30 wt.%
Given many polar groups on the surface of the M-BTC-PEI microparticles, a right amount of PIBSA as both dispersant and charge control agent, was added into the nonpolar isododecane dispersion against possible aggregation. In particular, the long and flexible carbon chain of PIBSA has a spatial site-blocking effect to keep these electrophoretic microparticles away from each other. Just shown in Fig. 6c, d, the ultraviolet-visible spectra (UV–Vis) absorbance of the Cu-BTC-PEI dispersion decreased slowly within 1 h in the presence of PIBSA, but the one without PIBSA cannot keep in a good dispersion even in 5 min. In this Cu-BTC-PEI/PIBSA formula, those colloid particles clearly exhibited good dispersibility in isododecane, especially low apparent density close to that of isododecane. This is also reflected in the corresponding object figures (see Fig. 6e).
Next, we verified the single particle electrophoresis system of M-BTC-PEI microparticles in PIBSA/isododecane for the first time. At a driving direct current (DC) voltage of 20 V, those M-BTC-PEI microparticles were visibly fixed on the side of the negative ITO electrode, and the other part became completely transparent while the electrophoresis plate tilted (see Fig. S8). Apparently, these electrophoretic microparticles were positively charged. And we integrated the modified TiO_2_ nanoparticles as representative white electrophoretic particles into this dispersion, to form a two-color electrophoretic ink^44^. As shown in Fig. S9 and Table S1, the applied TiO_2_ nanoparticles had few basic features: particle size, ~200 nm; Zeta potential, + 55.57 mV; electrophoretic mobility, 4.1 × 10^−10^ m^2^ V^−1^ s^−1^). Here, the Zeta potential was higher than that of the M-BTC-PEI microparticles, and surely makes electrophoresis migration of the modified TiO_2_ nanoparticles faster under the same driving situation. It may be effective to avoid possible adsorption, agglomeration, and precipitation of positive and negative particles during multiple electrophoresis to some extent.
This two-color electrophoretic ink was poured into electrophoretic display cell, and then driven at an alternating DC voltage of ±20 V (see Fig. 7a) in daylight. Herein, response time (Ton) and shutdown time of electrophoretic display (Toff; also known as recovery time) were defined as the required time from baseline to 90% of reflectance maximum, and the one from 90% of reflectance maximum to baseline, respectively. Figure 7b–e showed that all the Ton of the M-BTC-PEI/TiO_2_ inks were less than 2.3 s, and all the Toff were less than 5.9 s under the low field strength of 0.02 V μm^−1^. Here, high charge-to-mass ratio of the M-BTC-PEI microparticles based on high Zeta potential and low apparent density was the main reason for the rapid response. Very clearly, the Co-BTC-PEI/TiO_2_ formula was the most effective: Ton, 1.10 s; Toff, 4.82 s. Because Co-BTC-PEI microparticle took the lead in charge density per unit volume and mobility: Zeta potential, 35.54 ± 1.69 mV; effective size, 1307.40 ± 22.40 nm; mobility, 4.7 ± 1 × 10^−8^ m^2^ V^−1^ s^−1^.Fig. 7. Reflectance performances of the two-color EPD samples.Alternating waveform design of the applied driving DC voltage (a; ±20 V); two-color device reflectance of the M-BTC-PEI/TiO_2_ isododecane solutions (mass ratio of the resulting M-BTC-PEI microparticle to the modified TiO_2_ nanoparticle, 2:1) with the settled PIBSA concentration of 50 mg mL^−1^ encapsulated inside the electrophoretic display cell (b–e). Notes: PEI M.W., 10,000 g mol^−1^; PEI feed ratio, 30 wt.%
As well known, contrast ratio (CR) is the ratio of the white reflectance value (Rw) to the black reflectance value (Rb), namely CR = Rw/Rb. In this study, the CR value is defined as the ratio of Rw to the color reflectance value, and named as CRCu-T, CRFe-T, CRNi-T, and CRCo-T according to the corresponding M-BTC-PEI/TiO_2_ ink, respectively. It was drawn from Fig. 7b–e that, CRFe-T (1.37) > CRCu-T (1.35) > CRNi-T (1.23) > CRCo-T (1.16). By the comprehensive comparison of the four key parameters, i.e. Ton, Toff, CR, and color, the Cu-BTC-PEI/TiO_2_ ink is viewed as the best option, while the Co-BTC-PEI/TiO_2_ ink has the fastest response but the lowest CR value.
In actual electrophoresis scenes, the real-time variations in chromaticity values of the M-BTC-PEI/TiO_2_ inks were monitored using a colorimeter. As we can see from Fig. 8a~d, Video S1, S2, S3, and S4, the blue-white Cu-BTC-PEI/TiO_2_ and the reddish-brown-white Fe-BTC-PEI/TiO_2_ systems with high Rw values were undoubtedly the best represented in the display effect. First, the two-color coordinates were far away from the white point (0.33, 0.33). Second, both the Cu-BTC-PEI and Fe-BTC-PEI microparticles showed significantly better behavior with small size, regular morphology, and high color saturation (see Fig. 2). Third, both of them had high charge density per unit volume and mobility in the core electrophoretic index (see Table 1). By comparison, the purple-white Co-BTC-PEI/TiO_2_ system did not perform well, which was mainly ascribed to its large size and irregular morphology^45^.Fig. 8. Display effects of the two-color EPD samples.Color EPD tests of the two-color M-BTC-PEI/TiO_2_ electrophoretic inks (mass ratio of the resulting M-BTC-PEI microparticle to the modified TiO_2_ nanoparticle, 2:1; PIBSA concentration, 50 mg mL^−1^) encapsulated inside the electrophoretic display cell at ±20 V for 20 s, and corresponding color coordinates: Cu-BTC-PEI/TiO_2_ (a), Fe-BTC-PEI/TiO_2_ (b), Ni-BTC-PEI/TiO_2_ (c), and Co-BTC-PEI/TiO_2_ (d). Notes: PEI M.W., 10,000 g mol^−1^; PEI feed ratio, 30 wt.%
Moreover, we collected and characterized those charged M-BTC-PEI microparticles on the electrophoretic plate after continuous power over 1 h. In Fig. S10, the characteristic peaks of those MOF microparticles still remained, indicating the good structural stability after multiple electrophoresis in a low dielectric liquid. We also tested the four two-color M-BTC-PEI/TiO_2_ electrophoretic systems after multiple driving cycles with a duration of 1200 s (see Fig. S11). All the M-BTC-PEI/TiO_2_ systems experienced an inevitable decrease in reflectance of 10.61% to 27.25% after continuous multiple for 1200 s. This can be attributed to the interwinding of the grafted PEI chains between electrophoretic particles, as well as resulting particle agglomeration. It also revealed that PEI modification can substantially improve the surface potential of the MOF particle, but surely affects the driving lifetime.
Compared with the other color electrophoretic particles, these PEI-reinforced MOF microparticles have distinct advantages in driving field strength and response time, as shown in Table 2 (refs. ^46–55^). And their low cost, facile preparation, high and high stability in color, particle and display, were also obviously superior to the other organic color pigments and inorganic particles.Table 2. Comparison of the electrophoretic characteristics of the reported color electrophoretic inksMaterialsZeta potential (mV)Electrophoretic mobility (10^−8^ m^2^ V^−1^ s^−1^)Average diameter (nm)Driving Voltage (V)/Spacing (µm)Ton (s)ReferenceFe/Co/Al-doped TiO2−102.14−0.1048286.830/~1.121^46^CuPcCl−26.8–12030/10001.5^47^PY110/PS−60 to −70–50015/10002.0^48^PY181/PE-30 to 4030015/10004.048BAM-MPS-PLMA−63.73–70030 /~0.17^49^RYB-poly(styrene-co-4-VP)–40−2.89 × 10^−2^80020 /100–^50^PY13−3.45−8.269 × 10^−3^217.7100 /10001.0^51^PR254−31.83−7.631 × 10^−2^155.2100/10000.851PB15−0.45−1.076 × 10^−3^80.3100/10000.551TiO2-OTS/P(4-VP-co-LA21)–3.6 ± 1−2.5 × 10^−2^435 ± 1650 /1000.416^52^PB-IL235+71.46+1.64263 ~ 438––^53^Fe3O4@SiO2−55.9–101 ~ 1772.5 /500.472^54^PIL/SCAs+4.13+6.17 × 10^−2^18820/~0.165^55^Cu-BTC-PEI+27.51+3.6 × 10^−2^782.4420/10002.16This workFe-BTC-PEI+30.57+4.1 × 10^−2^806.8720/10001.14This workNi-BTC-PEI+36.11+4.7 × 10^−2^2018.7820/10002.28This workCo-BTC-PEI+35.54+4.7 × 10^−2^1307.4020/10001.10This workNote: To M-BTC-PEI, PEI was the bonded PEI chains with the molecular weight of 10,000 g mol^−1^, and the feed ratio of 30 wt.%
Discussion
In general, charged particles suspended in a dielectric medium are subjected to the four forces under an applied electric field: electrophoretic force, buoyancy, gravity and retarding viscous forces^39^. The Helmholtz-Smoluchowski Eq. 6 is used to describe the electrophoretic velocity (V) of a charged particle^17^. In this equation, ε, ξEP, U and μ denote the dielectric constant of the liquid, the Zeta potential of the particle, the applied electric field and the mobility of the particle, respectively. (Note: For ease of comparison, electrophoresis velocity, applied electric field, and viscosity coefficient are uniformly represented by the symbols V, U, and η, respectively, without using other symbols):
\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V=\frac{\varepsilon {\xi }_{{EP}}{\rm{U}}}{\mu }$$\end{document}According to the electrophoretic theory and Stoke’s theorem, Eqs. 7 and 8 can also be used to describe the V of the charged particles^56,57^. Here, U is the applied electric field, q is the charge, l is the distance between the electrode plates, η is the viscosity coefficient, and m is the particle mass:
\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\frac{{Uq}}{l}-6\pi \eta R=m\frac{{\rm{d}}V}{{\rm{d}}t}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V=\frac{{Uq}}{6\pi l\eta R}\left(1-{e}^{-\frac{6\pi \eta R}{m}t}\right)$$\end{document}When the electrophoretic force is greater than the retarding viscous forces, a charged particle can be driven. The equation based on Stokes’ law for the adhesion force on the particle is expressed in Eq. 9 (ref. ^58^):
\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${F}_{d}=6\pi \eta {Vd}$$\end{document}where Fd is the drag force of the fluid on a particle, and d is the diameter of the particle^47^.
Based on Eqs. 6, 7, 8, and 9, it is found that the electrophoretic velocity of the particle is influenced by particle (e.g. Zeta potential, mobility, charge, and mass), dielectric liquid (e.g. dielectric constant and viscosity coefficient), applied electric field, and electrode plate spacing. In Eq. 8, the V of a charged particle is proportional to the q of the particle and inversely proportional to the m. So, a high charge-to-mass ratio (q/m)^59^ surely helps improve the electrophoretic velocity of the charged particles. Apparently, M-BTC-PEI microparticles have a huge advantage in charge-to-mass ratio for electrophoretic particles.
Moreover, given gravity and buoyancy forces, a small density difference between charged particles and dielectric liquids is also beneficial to high electrophoretic velocity. In our study, it is particularly noteworthy that the density of the resulting M-BTC particles can be controlled^12^. For example, nano-sized Cu-BTC can be prepared with a density of 0.88 g/cm³, which is very close to the density of the isododecane used (0.75 g/cm³).
In summary, we integrated the four colored MOF microparticles and polyelectrolytes into the blue, reddish-brown, green, and purple electrophoretic particles for the first time. In particular, the surface modification of PEI chains based on non-covalent interaction rendered these M-BTC microparticles reinforced fully: the Zeta potential rose from ~−10 to ~+30 mV; the electrophoretic mobility grew up from ~1.0 × 10^−10^ to over 3.6 × 10^−10^ m^2^ V^−1^ s^−1^. The colored M-BTC-PEI microparticles were further paired with the representative white TiO_2_ nanoparticle to form the two-color stable dispersions in nonpolar PIBSA/isododecane as the electrophoretic inks. Under the ultralow field strength of 0.02 V μm^−1^, all the response time and recovery time were no more than 2.3 and 5.9 s, respectively. Even after long-time or multiple driving, these MOF microparticles and their reflectance value still remained constant to some extent. Hereinto, the blue-white Cu-BTC-PEI/TiO_2_ system with high CR values behaved with the best represented in the display effect. By comparison, these colored PEI-reinforced MOF microparticles are superior to the other organic color pigments and inorganic particles, in response capability, charge-to-mass ratio, preparation method, production cost, and stability in color, particle, and display. It is anticipated to provide an innovative and promising technical path for color electrophoresis display.
Materials and methods
Materials
Copper (II) nitrate trihydrate (Cu(NO_3_)2·3H_2_O), cobalt (II) nitrate hexahydrate (Co(NO_3_)2·6H_2_O), nickel (II) nitrate hexahydrate (Ni(NO_3_)2·6H_2_O), iron(III) nitrate nonahydrate (Fe(NO_3_)3·9H_2_O), 1,3,5-benzenetricarboxylic acid (BTC), polyvinylpyrrolidone (PVP; mean molecular weight (M.W.): ~58,000 g mol^−1^.), polyethyleneimine (PEI; M.W.: 1,800, 10,000, and 25,000 g mol^−1^), polyacrylic acid (PAA; M.W.: ~2000 g mol^−1^), polyethylene glycol (PEG; M.W.: ~2000 g mol^−1^.), sorbitan oleate (Span 80), anatase titanium dioxide (TiO_2_; mean size: ~200 nm) naoparticles were purchased from Shanghai Macklin Biochemical Co. Other reagents and organic solvents were of analytical grade and were used directly without further purification.
Moreover, polyisobutenyl succinic anhydrides (PIBSA; industrial grade) as acidic charge control agent were provided by Jinzhou Gadorun Materials Technology Co., LTD. Isododecane was supplied by Weng Jiang Regent Co. Indium tin oxide (ITO) glass plates (thickness: 1.1 mm; area resistance: 30 Ω/mm^2^; transmittance 550 nm: 85–87%) were bought from Wuhu Token Sciences Co., LTD (Wuhu, China), and completely rinsed and dried prior to use.
Syntheses and characterizations of polymer-reinforced MOF microparticles
The whole synthetic routes were systematically shown in Fig. 1a, according to the previously reported solvothermal method^24–27^.
Syntheses of colored MOF microparticles constructed by coordination of metal ion (M) and BTC (M-BTC; M = Cu, Fe, Ni, or Co)
Cu-BTC: Typically, 5 mL of ethanol solution of PVP (10 g L^−1^), 25 mL of dimethyl formamide (DMF), 25 mL of ethanol solution of BTC (0.10 M), and 25 mL of aqueous solution of Cu(NO_3_)2·3H_2_O (0.02 M), were added in turn every 10 min, and thoroughly stirred. Subsequently, the resulting mixture was poured into a stainless-steel vessel and then heated up in the oven at 80 °C for 24 h^24^. After cooling overnight, the product was washed with anhydrous ethanol three times by centrifugation, and dried in vacuum at 70 °C to obtain a blue powder.
Fe-BTC: Typically, 14.544 g of Fe(NO_3_)3·9H_2_O and 5.04 g of BTC were codissolved in 36 mL of deoxidized ultrapure water, and magnetically stirred at ambient temperature for 1 h. Subsequently, the mixture was poured into a stainless-steel vessel and then heated up to 160 °C in the oven for 12 h. After cooling overnight, the solid products were collected by centrifugation, and then mixed with the right amount of ultrapure water and anhydrous ethanol at 70 °C for 3 h^25^. Finally, the reddish-brown powders were obtained by centrifugation and vacuum drying at 60 °C.
Ni-BTC: Typically, 3 g of PVP, 0.3 g of BTC, 0.864 g of Ni(NO_3_)2·6H_2_O were added into the mixed solvent (H_2_O: ethanol: DMF = 20: 20: 20 mL) in turn, and magnetically stirred for 30 min. Subsequently, the resulting mixture was transferred to a 100 mL stainless-steel vessel, and heated up to 150 °C in the oven for 10 h^26^. Finally, the green products were washed three times with methanol by centrifugation, and dried in a vacuum at 60 °C for 10 h.
Co-BTC: Typically, 1.2 g of Co(NO_3_)2·6H_2_O and 0.3 g of BTC were added into 30 mL of DMF in turn, and magnetically stirred for 0.5 h. Subsequently, the mixed solution was transferred to a 50 mL stainless-steel vessel, and heated up to 120 °C in the oven for 15 h^27^. Finally, the purple products were cooled, filtered, washed with DMF and ethanol, and dried in a vacuum at 80 °C overnight.
Syntheses of colored polymer-modified MOF microparticles (M-BTC-polymer)
PEI (M.W.: 1800, 10,000, and 25,000 g mol^−1^), PAA (M.W.: 2000 g mol^−1^), PEG (M.W.: 2000 g mol^−1^) were ultrasonically dissolved in anhydrous ethanol with the settled concentration of 10 mg mL^−1^, respectively. At the same time, M-BTC microparticles were dispersed in anhydrous ethanol with a settled concentration of 1.5mg mL^−1^. Subsequently, 250 μL of polymer solution was dropwise added into 8 mL of M-BTC dispersion, and then ultrasonicated for 15 min, followed by magnetic stirring for 15 min. Next, the mixture was collected by centrifugation (8000 rpm, 10 min) and washed twice with anhydrous ethanol. The final product was obtained by vacuum drying overnight at 50 °C. Prior to characterization, each sample was centrifuged and washed three times with anhydrous ethanol for full removal of the residual polymers. Here, the resulting product was recorded as M-BTC-PEI (or PAA, PEG; 20%), respectively, according to the mass injection ratio of PEI (or PAA, PEG) to M-BTC (1:5).
Scanning electron microscopy (SEM)
The microscopic morphologies of the resulting MOF microparticles were photographed by SEM (Ultra 55, Zeiss, Germany). The related elemental compositions were analyzed by energy dispersive spectroscopy (Dual QUANTAX 200, Bruker, Germany) with an XFlash6 dual-probe (Bruker, Germany). All the samples were the ethanol dispersion solutions of the resulting MOF microparticles with the fixed concentration of 1 mg mL^−1^. Each 10 μL of the sample dispersion was taken out on the silicon wafer till dried out, and then were sprayed with platinum particles (thickness: 15 nm).
Transmission electron microscopy (TEM)
The configurational morphologies of the resulting MOF microparticles were observed by TEM (JEM-1400 plus, JEOL, Japan). All the samples were also the ethanol dispersion solutions with the fixed MOF concentration of 1 mg mL^−1^.
Fourier transform infrared spectra (FT-IR) and attenuated total reflectance-Fourier and transform infrared spectra (ATR-FTIR)
The surface variations between the crude and modified MOF microparticles were characterized by FT-IR in the spectral range from 4000 to 400 cm^−1^ with a resolution of 2 cm^−1^ (Vertex 70, Bruker, Germany) using potassium bromide (KBr) tableting. The rich functional groups on the surfaces of the resulting MOF microparticles were further determined by ATR-FTIR in the spectral range from 4000 to 800 cm^−1^ and the temperature range from 30 to 100 °C.
Thermogravimetric analyzer (TGA)
The thermogravimetric analyses of the resulting MOF microparticles were performed using TGA (TGA2 Metler-Toledo, Switzerland). Each sample (~10 mg) was measured within a temperature range from 30 to 800 °C with the settled heating rate of 10 °C/min, in oxygen flow with the fixed rate of 20 mL/min.
Ultrapurple-visible spectra (UV-Vis)
The electrophoretic particles were characterized with a wavelength range from 300 to 500 nm using UV–Vis spectrophotometer (LAMBDA 950, Perkin-elmer, USA). Here, the sample was the isododecane dispersion solution of the resulting MOF microparticles with a fixed concentration of 10 mg mL^−1^.
X-ray photoelectron spectra (XPS)
The surface elemental compositions of the resulting MOF microparticles XPS spectra were determined by an X-ray photoelectron spectrometer (AXIS SUPRA, Shimadzu, Japan) equipped with monochromatic Al Kα (E = 1486.6 eV) radiation. The measured XPS energies were collected using the C 1s peak of the C–C bond at 284.8 eV.
Powder X-ray diffraction (XRD)
The crystal phase analyses of the resulting MOF microparticles were carried out at room temperature using X-ray diffractometer (D8 ADVANCE, Bruker, Germany) with a generator voltage of 40 kV and a 2θ range of 5–70°. Here, the scanning speed was set as 5°/min.
Dynamic light scattering (DLS) and Zeta potential analysis
The colloidal properties of the resulting MOF microparticles were measured using a nanoparticle size and Zeta potential analyzer (NanoBrook 90 plus PALS, Brookhaven, U.S.A.). Here, the sample was the isododecane dispersion solution of the resulting MOF microparticles with a fixed concentration of 0.02 mg mL^−1^. Typically, the resulting autocorrelation functions were analyzed using built-in software to extract hydrodynamic dimensions and polydispersity. And the auto-balanced voltage values and default parameters were used to obtain the Zeta potential.
Atomic force microscopy (AFM) and Kelvin probe force microscopy (KPFM)
AFM and KPFM measurements were performed by a Multimode 8 (Bruker, Germany) instrument, using the SCM-PIT-V2 test probe. KPFM technology provides the surface potential of samples (SP) in air. Here, the SP value of the sample was calculated by the relation, SP = Φsample-Φtip, where Φtip and Φsample are the work functions of the sample and the tip, respectively^40,41^. And the local contact potential difference (CPD) of the sample was calculated according to the formula, CPD = (Φtip - Φsample)/e.
Electrophoretic characterizations of polymer-reinforced MOF microparticles
Preparation of colored electrophoretic inks based on polymer-reinforced MOF microparticles
Typically, 0.1 g of the resulting M-BTC-PEI (M=Cu, Fe, Ni, Co) microparticle was ultrasonically dispersed into 1 mL of the isododecane solution of PIBSA (50 mg mL^−1^) to obtain positively charged colored electrophoretic ink.
Preparation of white electrophoretic ink
Typically, 0.1 g of the modified TiO_2_ nanoparticles was ultrasonically dispersed in 1 mL of the isododecane solution of PIBSA (50 mg mL^−1^) to obtain a positively charged white electrophoretic ink.
Fabrication of two-color electrophoretic display cell
Just as shown in Fig. 1b, the adopted display cell comprised two parallel ITO glass plates (3 cm × 3 cm) with the inner face-to-face ITO layers and the fixed spacing distance (100 μm) determined by a standard double-sided tape. As the electrophoretic fluid, 1 mL of the colored electrophoretic ink and 0.5 mL of the white electrophoretic ink were ultrasonically mixed to obtain the two-color electrophoretic ink, and then injected into the cell.
Effect test of electrophoretic display
During the process of electrophoretic display, a predetermined driving voltage of ±20 V was applied through a direct current (DC) power supply (CE0400010T, Earthworm Electronics, China). Here, the display effect was evaluated by reflectance, response time, and color coordinates using a high-speed reflectometer (Admesy, Netherlands), according to CIE (Commission Internationale de I’ Éclairage) color space standard.
Supplementary information
Revised Supporting Informantion (no marks) 20250926 Video S1 Video S2 Video S3 Video S4