High-efficient liquid exfoliation of 2D metal-organic framework using deep-eutectic solvents.

The exfoliation of bulk two-dimensional metal-organic framework (MOF) into few-layered nanosheets has attracted much attention recently. In this work, an environmental-friendly route has been developed for layered-MOF (MAMS-1) delamination using deep eutectic solvent (DES), which is more sustainable and efficient alternative than conventional organic solvents for MOF nanosheet preparation. Under sonication condition, DES as solvents, the highest exfoliation rate of MAMS-1 is up to 70% with two host layers via poly(vinylpyrrolidone) (PVP) surfactant-assisted method. The presence of tert-butyl exteriors and the atomically thickness endow the MOF nanosheets stable suspension for at least one month. Due to the 2D structure and excellent stability, MAMS-1 nanosheet (MAMS-1-NS) was chosen as a good candidate to encapsulate Eu3+ cations. The obtained Eu3+@MAMS-1-NS acts as a multi-responsive luminescent sensor through fluorescence quenching, and can specifically recognize Fe3+ (LOD = 0.40 μM, KSV = 1.05 × 105 M-l), Hg2+ (LOD = 0.038 μM, KSV = 5.78 × 106 M-l), Cr2O72- (LOD = 0.33 μM, KSV = 1.55 × 105 M-l) and MnO4- (LOD = 0.088 μM, KSV = 4.49 × 105 M-l). Compared with bulk Eu3+@MAMS-1, the sensitivity of Eu3+@MAMS-1-NS is greatly improved owing to its ultrathin nanosheet morphology and highly accessible active sites on the surface.

Introduction

Since the discovery of graphene, two-dimensional (2D) nanomaterials [1], [2], [3], [4], [5], [6], such as boron nitride, metal hydroxides, transition metal chalcogenides, and black phosphorus, have attracted much attention due to their unique electronic structure and fascinating properties. However, 2D materials mentioned above are mostly inorganic compounds with simple composition and lack of structural diversity. Metal-organic frameworks (MOFs) nanosheet comprising organic linkers and metallic nodes, as a new kind of 2D nanomaterials, appeals increasing interest recently [7], [8], [9], [10], [11], [12], [13]. Due to the ultrathin thickness, numerous structural possibilities, easy access to active sites and large surface areas, 2D MOF nanosheets show promising applications in molecular sieving [14], [15], energy storage and conversion [16], [17], catalysis [18], [9], [19], [20], [21], and luminescent sensing [22], [23]. Although there are multitudinous 2D MOF structures, only a few studies on ultrathin 2D MOF nanosheets have been reported owing to the poor stability and the limitations of production. It is more difficult for 2D MOF exfoliation than traditional 2D materials, because hydrogen bonds, π–π stacking, multiple interactions between the wrinkle surface all result in the interlayer interactions not always weaker than the intralayer coordination bonds of layered 2D MOFs. Therefore, developing new strategies for ultrathin MOF nanosheet construction is of great significance.

Similar to other 2D materials preparation, both top-down exfoliation and bottom-up assembly can be adopted to synthesis 2D MOF nanosheets [24]. Especially, “top-down” approaches, for example, liquid-phase sonication exfoliation [25], [26], ball-milling [27], [28] and chemical intercalation [29], are readily achieved and scalable production pathway. Solvent-mediated sonication exfoliation of bulk layered MOFs into 2D nanosheet is very attractive for its general applicability and accessibility. Solvents play important roles during exfoliation. The good wettability, suitable Hasen solubility parameter and surface tension, intercalation, strong interactions and stabilization effects of solvents all contribute significantly for bulk MOFs stripping ultrasonically. Unfortunately, solvents used for layered MOFs peeling off, for example N,N-Dimethylformamide, tetrahydrofuran, hexane and acetone, are generally volatile, toxic or not eco-friendly, and not efficient enough [30]. Therefore, approaches to exfoliate layered MOFs in a more efficient and environment-friendly manner are highly demanded.

Deep eutectic solvent (DES), constructed from H-bond donor and acceptor via hydrogen bonding, is a kind of green solvents similar to ionic liquids with a melting point lower than 100 °C [31], [32], [33], [34]. DES can be synthesized 100% atom efficient without any solvent used and components are adjustable, biodegradable and non-toxic. Compared with ionic liquids which are efficient green solvents for 2D materials preparation, DES are considered as more easily available and cost-effectively green solvents. Considering the surface tensions of DES based on ethylene glycol or glycerol around 40–47 mJ m−2, DES has already been used for 2D material construction under sonication, such as graphene [35], [36], [37] and metal hydroxides [59]. The adjustable composition, low vapor pressure, suitable surface tension and abundant hydrogen interactions endow DES great potential to produce 2D MOF nanosheet with crystal integrality and high production by ultrasonic exfoliation. Nevertheless, there is still no report for MOF nanosheet preparation in DES.

Chemical sensor is an effective method for trace pollutant determination in water. MOFs are good candidates for metal cations and anions sensing due to their tunable topological structures and high porosity [38], [39], [40]. Particularly, lanthanide based MOFs (Ln-MOFs) exhibit brilliant performance as luminescent probes owing to their special 4f electronic configurations [41], [42]. However, the high coordination numbers and flexible coordination geometry of lanthanide made the desired Ln-MOFs design and synthesize challenging. Moreover, the current used Ln-MOFs are mainly in bulk-size which limits their interactions with the analytes. With large lateral size and ultrathin thickness, 2D Ln-MOF nanosheet is in favor of more accessible active sites on their surface and beneficial to realize highly sensitive luminescent sensing. Post-synthetic modification (PSM) provides convenient opportunities to prepare 2D Ln-MOFs as luminescent sensors. In this work, bulk MOFs named as MAMS-1 was successfully exfoliated into layered nanosheets (MAMS-1-NS). PSM was utilized as a facile yet versatile approach to fabricate a 2D luminescent Ln-MOF by encapsulating Eu3+ cations into MAMS-1-NS. The obtained Eu3+@MAMS-1-NS exhibits multi-responsive behavior towards Hg2+, Fe3+ and Cr2O72−, MnO4− with high quenching efficiency.

Experimental section

Bulk MAMS-1 crystals synthesis

MAMS-1 was synthesized as previously reported [43]. Briefly, 0.34 mmol 5-tert-Butyl-1,3-benzenedicarboxylic acid (H2(bbdc)) and 0.51 mmol Ni(NO3)2·6H2O were dissolved in 7.5 mL H2O/ethylene glycol (4:1 v/v,) and sealed in an autoclave. The reaction proceeded at 210 °C for 24 h and light-green crystals were obtained.

Ultrasonic exfoliation in DES

Firstly, DES was prepared by heating the mixture of choline chloride (ChCl) and ethylene glycol (EG) with a molar ration 1:2 at 80 °C. Then, the MAMS-1 crystals were dispersed in DES with the help of surfactants. The concentration of MAMS-1 dispersion is 3 mg/g in DES. Surfactants such as PVP, SDS and CTAB are added into DES with a concentration of 0.5 mg g−1. Finally, the dispersion was ultrasonically treated in an ultrasonic processor for 60 min. After centrifugation at 5,000 r.p.m. for 30 min, the colloidal MAMS-1 nanosheet (MAMS-1-NS) suspension was collected. MAMS-1-NS in solid state was obtained after a 15,000 r.p.m centrifugation and vacuum drying. Controlled experiment was done without surfactants at the same sonication conditions. Moreover, MAMS-1 crystals dispersed in DES with PVP under mechanical agitation for an hour was also investigated to certify the function of sonication.

Preparation of Eu3+@MAMS-1-NS

Eu3+@MAMS-1-NS was synthesized by soaking 100 mg MAMS-1-NS into 10 mL Eu(NO3)3·6H2O (0.2 mmol ml−1) ethanol solution for 48 h. The resulting Eu3+@MAMS-1-NS was obtained after centrifugation and washing with ethanol for several times. Eu3+@MAMS-1-NS suspension used for luminescent sensing was prepared by introducing 1 mg Eu3+@MAMS-1-NS power into 10 mL ethanol under ultrasonication.

Luminescence sensing of Eu3+@MAMS-1-NS

The luminescent sensing of Eu3+@MAMS-1-NS was explored by treating 3 mL MOF suspension (0.1 mg/mL) with 30 μL aqueous solution of M(NO3) × (10 mM; M = Na+, Ag+, Mg2+, Co2+, Ca2+, Ni2+, Zn2+, Pb2+, Cd2+, Cu2+, Cr3+ and Fe3+) or KyX (0.1 mM; X = Br−, Cl−, NO3−, SO42−, CH3COO−, ClO4−, F−, I−, SO32−, Cr2O72− and MnO4−) at room temperature. The mixtures were then used for luminescence measurements. Quinine sulphate (0.1 M in H2SO4) was used as reference for relative photoluminescence quantum yield measurement. The absorbance of Eu3+@MAMS-1-NS dispersion was detected with the same procedure for quantum yield calculation. The fluorescence lifetime is determined by time-resolved fluorescence.

Characterization

A Nicolet 6700 FTIR spectrometer is used for FTIR spectra detection in the range 4000–400 cm−1. PXRD patterns were collected from 5° to 50° (2θ) with a scanning rate 4° min−1 on an X-ray powder diffractometer (Rigaku MiniFlex 600). Transmission electron microscope (TEM, JEM-2100plus) and scanning electron microscope (SEM, Hitachi S-4800) equipped with an energy-dispersive X-ray spectrometer are applied to characterize the morphology of sensors used in this work. Excitation and emission spectra were taken on a Cary Eclipse spectro-fluorimeter. The atomic force microscopy (AFM, MuLtimode 8) is carried out using a tapping-mode. X-ray photoelectron spectroscopy (XPS) is recorded on an ESCALab220i-XL electron spectrometer with 300 W Al Kα radiation.

Results and discussion

Exfoliation of bulk MAMS-1

A layered 2D MOF named as MAMS-1 is a hydrothermal stable molecular sieve with an interlayer distance 1.9 nm [44]. Herein, we conducted the first study on bulk MAMS-1 exfoliation in DES. The preparation procedure of MAMS-1-NS is illustrated in Scheme 1. DES was selected as an appropriate solvent for liquid exfoliation. Conventional DES composed of ChCl and EG with a molar ratio 1:2 was certified to be effective. With the help of sonication and surfactants, bulk MAMS-1 was exfoliated. Upon exposing layered MAMS-1 to an acoustic cavitation process, single or few-layer MAMS-1-NS can be produced, attributing to the high energy supplied via bubbles creation and collapsion under sonication in solvents. Without sonication, MAMS-1-NS cannot be obtained (Fig. S2). Sonication is essential and leads to mechanical deformation of bulk MAMS-1 at the edge, followed by sequential insertion of ions or molecules from DES solution. This process was eco-friendly without volatile organic solvents participation.

Scheme 1

The exfoliation of MAMS-1 into few-layer MAMS-1-NS in DES.

Before exfoliation, bulk MAMS-1 with obvious lamellar structure can be observed from SEM and TEM images in Fig. 1(A). The bulk MAMS-1 presents a rectangular morphology with the length 50 μm and width 1.5 μm. After stripping, nanosheets were obtained and a colloidal suspension with good dispersion stability can be prepared. Due to light scattering, Tyndall effect can be observable (the insert of Fig. 1(B)). From the SEM and TEM images in Fig. 1(B), bulk MAMS-1 chunks have disappeared and curly nanosheets presented. Ultrasonic cavitation remarkably reduced the thicknesses along the c-axis of MAMS-1-NS during the exfoliations. Inevitably, the MOF size along the plane direction was also diminished under ultrasonication. XRD patterns of MAMS-1-NS presented in Fig. 1(C) showed no obvious peak position changes compared with the original MAMS-1. The main diffraction peaks are observed at 8.28°, 10.56°, 14.07°, 18.18°, and 28.17°. However, the peak intensity of (0 1 1) located at 8.28° become weaker in MAMS-1-NS, indicating that MAMS-1 stacks along the (0 1 1) direction and MAMS-1-NS was successfully exfoliated from bulk MAMS-1 via ultrasonication . Additionally, the relative peak intensities of I(0 1 3)/I(0 1 1) and I(0 1 5)/I(0 1 1) for MAMS-1-NS (0.26, 0.30) are noticeable greater than those of MAMS-1 (0.17, 0.14). It is concluded that more lattice plane were exposed, such as (0 1 3) and (0 1 5) shown in Fig. 1(C), which may offer abundant active sites and be beneficial to ions transmission [45].

Fig. 1

(A) SEM image of layered MAMS-1 (the insert is the TEM image of enlarged MAMS-1); (B) SEM and TEM images of lamellate MAMS-1-NS (the insert is the digital photograph of MAMS-1-NS dispersion in ethanol); (C) XRD patterns of MAMS-1 before and after exfoliation.

The morphologies and surface topographies of finely dispersed MAMS-1-NS were further characterized by AFM. MAMS-1-NS was deposited onto a silica substrate for AFM observation. AFM images revealed flaky MAMS-1-NS with several lamellas were fully exfoliated (Fig. 2). MAMS-1-NS obtained exhibit relatively large lateral dimensions, some of which are up to 5 × 1 μm2. Their thicknesses are typically around 4 nm corresponding to the thickness of tow elementary host layers. The nanometre-sized thickness accompanying with broad lateral dimension distinctly certified the successful exfoliation. Bulk MAMS-1 crystals composed of 2D layers via weak van der Waals stacking, are deliberately broken under sonication in DES here. Accordingly, bulk MAMS-1 could be efficiently exfoliated into ultrathin nanosheets with about two-layer overlapped stacks in a facile and green manner. The successful exfoliation may be attributed to the ultrasonication, befitting surface tension of solvents, hydrogen bonding and hydrophobic interactions between DES and surface ligands of MAMS-1, and the stabilization of surfactants [46].

Fig. 2

(A) An AFM overview of MAMS-1-NS with different host layers, (B) and (C) the enlarged images of MAMS-1-NS with two layers.

Surface tension of solvents plays an important role during the ultrasound-assisted exfoliation of 2D materials [47]. Using water and glycerol as probe liquids, the surface energy of MAMS-1 was determined 43 mJ m−2 (Surface energy Determination of MAMS-1 in supporting information). The reported surface tension of DES composed of ChCl/EG (1:2) is 48 mN m−1 [48], and a little larger than MAMS-1. After the introduction of surfactants, the surface tension of DES decreased and was closer to the MAMS-1 surface energy. Different surfactants such as PVP, SDS, and CTAB were investigated in this work. MAMS-1 can be peeled in DES without surfactants under ultrasonication with only 23% efficiency. With the addition of surfactants, the exfoliation efficiencies of MAMS-1 were enhanced to 73%, 80% and 75% in PVP, SDS, and CTAB solutions, which are much higher than pure DES. It can be attributed to surface tension suitability for DES solution after surfactant introduction. With a surfactant PVP, SDS and CTAB concentration 0.5 mg g−1 in DES, the surface tensions of solvents are 46 mN m−1, 39 mN m−1 and 34 mN m−1 respectively. Obviously, the highest efficiency was achieved in SDS solution which can be explained by the best matched surface tension with MAMS-1. Therefore, surface tension compatibility is conducive to the improvement of exfoliation efficiency.

Furthermore, the obtained MAMS-1-NS assisted without or with surfactants was studied by SEM in Fig. 3. After ultrasonic treatment, the original rectangular bulk MAMS-1 was broken and MAMS-1-NS with a thickness in nanometers can be observed. The lateral dimension of MAMS-1-NS is about several micrometers. However, there are still some swollen MAMS-1 appeared in Fig. 3. Obviously, PVP exhibits the most effective role in assisting the stripping and dispersion of MAMS-1-NS. A stable MAMS-1-NS dispersion with a high concentration of 2.2 mg g−1 could be generated within an hour and no obvious precipitation can be observed after standing more than one month. Non-ionic surfactant tends to exhibit better performance than ionic ones. It has been reported that the higher molecular weight of PVP than SDS and CTAB results in stronger steric repulsion effect and better suspending ability for 2D nanosheets [49].

Fig. 3

The SEM images of exfoliated MAMS-1 in DES solutions (A) without surfactant, and with different surfactants (B) SDS, and (C) CTAB.

Luminescence and sensing properties

MAMS-1-NS is composed of a hydrophilic octanickel cluster layer sandwiched by two hydrophobic carboxylate layers with a 2D porous structure [43], providing a platform for post-modification and functionalization. Therefore, MAMS-1-NS is a good candidate for luminescent sensing. Herein, a simple post-synthetic pathway was adopted for luminescent Eu3+-doped MAMS-1-NS preparation. XRD patterns of the as-synthesized Eu3+@MAMS-1-NS showed in Fig. S3 are in good agreement with that of the original MAMS-1-NS, suggesting that the crystal structure still remains unchanged after Eu3+ incorporation. The supported Eu3+ on MAMS-1-NS can be directly observed by the SEM element mapping shown in Fig. 4. The morphology of MAMS-1-NS was well preserved after Eu3+ adsorption as revealed in Fig. 4(A). Eu3+ was dispersed uniformly on MAMS-1-NS (Fig. 4(E)).

Fig. 4

(A) SEM image of Eu3+@MAMS-1-NS and (B) elemental mapping of Eu3+@MAMS-1-NS, (C) C, (D) O, (E) Eu and (F) Ni (all scale bar dimensions are 2 μm).

XPS analysis proves that the Eu3+ ions were successfully coordinated with MAMS-1-NS. The peak of Eu 3d5 at 1135.05 eV observed in Fig. 5(A) confirmed the presence of Eu3+ in MAMS-1-NS. The peak of O 1s is broaden and binding energy moved from 531.34 to 531.76 eV after Eu3+ introduction. These changes imply the successful coordination between free carboxylic acid and Eu3+ (Fig. 5(B)) [50]. The highly porous feature and exposed active surface containing uncoordinated carboxyl groups of MAMS-1-NS were proved to be suitable platforms for Ln(III) ions encapsulation to form novel Ln-doped luminescent materials.

Fig. 5

(A) XPS spectra and (B) O 1s spectra of MAMS-1 and Eu3+@MAMS-1-NS.

Successful doping Eu3+ ion onto MAMS-1-NS was further certified by the photoluminescence study. Due to the π*-π transition of ligands, MAMS-1-NS displays weak luminescence with a wide band centered around 450 nm when it is excited by 295 nm (Fig. S4). After the integration of Eu3+ ion, the ligand centered emissions of MAMS-1-NS were remarkably depressed. The excitation and emission spectroscopy of Eu3+@MAMS-1-NS was demonstrated in Fig. 6(A). As excited at 290 nm, the characteristic emission bands of Eu3+ at 562, 594, 616, 650, and 700 nm can be observed, which can be attributed to 5D0 → 7FJ (J = 0–4) transitions of Eu3+ ions [51]. There is a direct correlation between transitions and the inversion symmetry around the Eu3+ ion. Magnetic dipole transition from 5D0 to 7F1 suggests an inversion location. Conversely, the electric dipole transition from 5D0 to 7F2 reveals Eu3+ ion in the noninversion site [52]. The intensity proportion between 5D0 → 7F2 to 5D0 → 7F1 of Eu3+@MAMS-1-NS was up to 3 and proved the Eu3+ ions low-symmetry coordination site locations [53]. The reduced ligand centered emission and the strong luminescent peaks of Eu3+@MAMS-1-NS demonstrate a so-called “antenna effect” occurs and energy transfer from ligands to Eu3+ ions within the framework has happened. Therefore, MAMS-1-NS is an efficient scaffold for lanthanide sensitization. The photophysical behaviors of Eu3+@MAMS-1-NS were also studied. To evaluate the luminescent efficiency, relative photoluminescence quantum yield referred to quinine sulfate (54%) was calculated and an 8.8% quantum yield achieved. The exciton recombination dynamic was investigated by time-resolved PL spectra. The fluorescent lifetime value was calculated through fitting with the decay curve (Fig. 6 B). Eu3+@MAMS-1-NS has an average lifetime 3.73 ns.

Fig. 6

(A) The excitation and emission spectra of Eu3+@MAMS-1-NS (B) Time-resolved PL decay of Eu3+@MAMS-1-NS and the corresponding fitting curve.

Based on the excellent luminescence of Eu3+@MAMS-1-NS, we examine the feasibility of Eu3+@MAMS-1-NS as a fluorescent-based sensor for cations and anions in water systems. The as-prepared Eu3+@MAMS-1-NS suspension of 0.1 mg/mL was used for luminescent detection with various metal ions (Ag+, Ca2+, Cd2+, Co2+, Mg2+, Na+, Ni2+, Pb2+, Zn2+, Cu2+, Cr3+, Fe3+, and Hg2+, 0.1 mM) and different anions (Br−, CH3COO−, Cl−, ClO4−, F−, I−, NO3−, SO32−, SO42−, Cr2O72−, and MnO4−) respectively. As shown in Fig. 7 (A) and (B), under excitation at 290 nm, various metal ions and anions have markedly different effects on the luminescence of Eu3+ ions. The Mg2+, Na+ ions have little influence in the luminescence intensity of Eu3+, while Ag+, Ca2+, Cd2+, Co2+, Ni2+, Pb2+, Zn2+, Cu2+, and Cr3+ ions could weakly decrease the Eu3+ emission intensity. However, the luminescence intensity at 616 nm is remarkably restrained with a quenching efficiency up to 90.2% and 96.1% when Fe3+ and Hg2+ are involved respectively, indicating that Eu3+@MAMS-1-NS can be a promising luminescent sensor for detecting Fe3+ and Hg2+ ions. Furthermore, the effects of Fe3+ and Hg2+ respectively mixed with other metal ions (Ag+, Ca2+, Cd2+, Co2+, Mg2+, Na+, Ni2+, Pb2+, Zn2+, Cu2+, and Cr3+ ions) on the emission of Eu3+@MAMS-1-NS were investigated. Although different metal ions were introduced, the quenching effects of Fe3+ and Hg2+ on the emission intensity of Eu3+@MAMS-1-NS were almost undisturbed (Fig. 7 C and D). Therefore, Eu3+@MAMS-1-NS is an ideal luminescent probe for Fe3+ and Hg2+ detecting.

Fig. 7

Luminescent intensity at 616 nm of Eu3+@MAMS-1-NS treated with 0.1 × 10−3 M various (A) cations and (B) anions; Comparisons of the luminescence intensities of Eu3+@MAMS-1-NS of different cations with (C) Fe3+, (D) Hg2+.

Fluorescence titration experiments were performed by the gradual addition of concentrated Fe3+ and Hg2+ into the Eu3+@MAMS-1-NS dispersion. As shown in Fig. 8 A and B, the emission intensity of Eu3+@MAMS-1-NS at 616 nm obviously decreased with the increasing Fe3+ and Hg2+ concentration. With Fe3+ concentration 100 μM and Hg2+ 3 μM, the emission of Eu3+@MAMS-1-NS was almost completely quenched. The relationships between luminescence intensity and the concentration of Fe3+ and Hg2+ were studied and the results were showed in Fig. 8 (C) and (D). For Fe3+ concentrations within 10 μM, the calculated detection limit (LOD) of 0.40 μM is obtained at a signal-to-noise ratio for 10 times. The legal provision of Fe3+ ions in drinking water (no more than 5.4 μM) is covered in our detection range [54]. For Hg2+ concentrations within 1.0 μM, the detection limit is 38.16 nM. The Health Organization (WHO) prescribed maximum permissible level of mercury concentration in drinking water is 190 nM [55]. Therefore, Eu3+@MAMS-1-NS exhibit good-performances in Fe3+ and Hg2+ sensing.

Fig. 8

The luminescent spectra of Eu3+@MAMS-1-NS in presence of (A) Fe3+ and (B) Hg2+ ions with different concentrations. Insets: the linear correlation for the plot of I/I0 vs. concentration of Fe3+ and Hg2+, respectively. The corresponding linear relationships between the emission intensity and (C) Fe3+, (D) Hg2+ concentration.

Furthermore, the quenching efficiency was rationalized by Stern-Volmer equation quantitatively [56]:where I and I are the luminescence intensity of Eu3+@MAMS-1-NS suspension without and with metal ion addition. K represents the quenching constant; [M] is the metal ion concentration. Good linear correlations are observed for the plots of I0/I vs. [Fe3+] and [Hg2+] over the concentration range from 0 to 45 and 1.6 to 2.8 μM (the insert of Fig. 8A and B). The K of Fe3+ and Hg2+ were calculated to be 1.05 × 105 M−l (R2 = 0.993) and 5.78 × 106 M−l (R2 = 0.992). The linear variation at the low concentration region is mainly ascribed to the static quenching [57]. Low detection limits (0.40 and 0.038 μM) and strong quenching constants (1.05 × 105 and 5.78 × 106 L mol−1) denote that Eu3+@MAMS-1-NS is a promising fluorescent sensor for the practical detection of Fe3+ and Hg2+.

The fluorescence responses of Eu3+@MAMS-1-NS towards typical halide anions (F−, Cl−, Br− and I−) and inorganic acid radicals (CH3COO−, ClO4−, NO3−, SO32−, SO42−, Cr2O72− and MnO4−) were also explored. As illustrated in Fig. 5B, Br−, Cl−, ClO4−, F−, and NO3− have negligible effects on Eu3+@MAMS-1-NS luminescence intensity; CH3COO−, I−, SO32−, and SO42− ions can weaken the luminescent intensity. However, the fluorescence intensities of Eu3+@MAMS-1-NS towards Cr2O72− and MnO4− were sharply depressed with a quenching efficiency up to 83.4% and 91.5%. Anti-interference experiments showed in Fig. 9 (A) and (B) indicating the good sensing selectivity for Cr2O72− and MnO4− ions. The detection limits obtained from Fig. 9 (C) and (D) for Cr2O72− and MnO4− ions were 0.33 μM (in the range of 0 to 10 μM) and 0.088 μM (in the range of 0 to 2.0 μM) respectively. The K of Cr2O72− and MnO4− were calculated to be 1.55 × 105 M−l (R2 = 0.995) and 4.49 × 105 M−l (R2 = 0.995) (Fig. 9 E and F). Consequently, Eu3+@MAMS-1-NS shows highly selectivity and sensitivity for Cr2O72− and MnO4− ion detection.

Fig. 9

Comparisons of the luminescence intensities (616 nm) of Eu3+@MAMS-1-NS of different anions (A) Cr2O72−, (B) MnO4−. The luminescent spectra of Eu3+@MAMS-1-NS in presence of (C) Cr2O72− and (D) MnO4− ions with different concentrations. Insets: the linear correlation for the plot of I/I0 vs. concentration of Cr2O72− and MnO4−, respectively. The corresponding linear relationships between the emission intensity and (E) Cr2O72−, (F) MnO4− concentration.

As a comparison, the sensing ability of bulk Eu3+@MAMS-1 was also studied (Fig. 10). The detection limits of bulk Eu3+@MAMS-1 for Fe3+, Hg2+, Cr2O72− and MnO4− sensing were 0.87 μM, 0.11 μM, 0.49 μM and 0.20 μM respectively (Fig. S5). Obvious enhancements for ion detection can be observed after the bulk MAMS-1 delaminated into nanosheets. The K of bulk Eu3+@MAMS-1 are 4.75 × 104 M−l, 1.32 × 106 M−l, 7.57 × 104 M−l and 2.14 × 105 M−l for Fe3+, Hg2+, Cr2O72− and MnO4− respectively (the insert of Fig. 10). Therefore, the nanosheet morphology is much more sensitive than the bulk ones. Furthermore, the comparison of detection limit (LOD) and K value was summarized in Table S1 for different ion sensing using Eu-MOF-based fluorescent materials. Eu3+@MAMS-1-NS demonstrates higher sensitivity and selectivity for ion sensing than most previously reported Eu-MOFs and has the ability to simultaneously detect a variety of cations and anions. The nanosheet morphology of Eu3+@MAMS-1 endows this fluorescent-based sensor with highly accessible active sites on the surface and facilitates the close contact between ions and Eu3+@MAMS-1. Thus 2D Eu3+@MAMS-1-NS exhibits a much lower detection limit and higher sensitivity compared with bulky MOFs.

Fig. 10

Luminescent intensity at 616 nm of Eu3+@MAMS-1-NS treated with 0.1 × 10−3 M various cations (A) and anions (B), and the luminescent spectra of Eu3+@MAMS-1-NS in the presence of Fe3+ (C), Hg2+ (D), Cr2O72− (E) and MnO4− (F) ions with different concentrations. Insets: the linear correlation for the plot of I/I0 vs concentration of Fe3+, Hg2+, Cr2O72− and MnO4− ions, respectively.

Sensing mechanism

To gain deep insight into the luminescence quenching effect induced by Fe3+, Hg2+, Cr2O72− and MnO4− ions, further characterizations including PXRD patterns and UV–visible absorption spectra were carried out. The powder XRD of Eu3+@MAMS-1-NS after soaking in Fe3+, Hg2+, Cr2O72− and MnO4− solutions remained unchanged (Fig. S6), indicating its good stability as luminescent detectors and no skeletal collapse during the fluorescence quenching. The UV–vis absorption spectra of Fe3+, Hg2+, Cr2O72− and MnO4− exhibit maximum overlap with the excitation spectrum of Eu3+@MAMS-1-NS between 250 and 350 nm (Fig. 11), clearly indicating a competitive absorption for the light source energy between Eu3+@MAMS-1-NS and Fe3+, Hg2+, Cr2O72−, MnO4−. Thus, energy transfer from ligand to Eu3+ is hindered on account of the competition light absorption between the Eu3+@MAMS-1-NS and the quencher ions, leading to the prominent fluorescence quenching [58]. As a consequence, the efficient energy transfer allows Fe3+, Hg2+, Cr2O72−, MnO4− to have a much higher fluorescence quenching effect compared with other tested cations and anions.

Fig. 11

(A) The UV spectra of metal ions in water and the fluorescence excitation spectra of Eu3+@MAMS-1-NS dispersion; (B) The UV spectra of anions in water and the fluorescence excitation spectra of Eu3+@MAMS-1-NS dispersion.

Conclusions

In conclusion, we have prepared 2D MOF nanosheets via an efficient top-down exfoliation pathway in DES under sonication. The suitable surface tension, abundant hydrogen bonding and hydrophobic interactions endow DES great efficiency to MAMS-1 stripping. The obtained MAMS-1-NS was doped with Eu3+ by a PSM method and showed relative selectivity and high sensitivity to Fe3+, Hg2+, Cr2O72−, MnO4− in solution due to its highly accessible active sites on the surface. According to the quenching response of Eu3+@MAMS-1-NS induced by ions, lower LODs (0.40 μM, 0.038 μM, 0.33 μM and 0.088 μM for Fe3+, Hg2+, Cr2O72− and MnO4−) and larger quenching constants (1.05 × 105 M−l, 5.78 × 106 M−l, 1.55 × 105 M−l and 4.49 × 105 M−l for Fe3+, Hg2+, Cr2O72− and MnO4−) are achieved compared the bulk Eu3+@MAMS-1. This interesting work provides a good example of designing LnIII-MOFs nanosheets with excellent multiple-responsive fluorescence-sensing abilities.

Declaration of Competing Interest

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