Morphology Control of Two dimensional (2D) Metal–Organic Framework Nanosheets via 1 nm Thick Hexagonal Hydroxide Nanosheets through Variable Reaction Parameters

Two different shaped Ni‐based metal–organic framework (MOF) nanosheets are prepared from hexagonal nickel hydroxide monolayer nanosheets under the variable reaction conditions. The hexagonal shaped MOF with similar lateral size to Ni(OH)2 nanosheet is obtained on flat Si substrate but the thickness increased from 1 to 2 nm. Another rectangular MOF nanosheet is obtained when a similar reaction is carried out in a solution in which Ni(OH)2 is dispersed. The difference in the shape of the Ni‐MOF nanosheets is related to the degree of dissolution of the Ni(OH)2 nanosheets during the reaction.


DOI: 10.1002/admi.202300335
MOF/PCPs has been reported 10 years ago, [6] a precise modulation in morphology and thickness of 2D MOFs have started to be explored for new or more efficient functions. Using conventional top-down methods [1b] a process has been reported to adjust the thickness of 2D MOFs as an effect of the organic ligands of the 3D MOFs, but the control over the morphology has not been achieved. [7] On the other hand, another conventional bottom-up method have been succeeded in obtaining uniformly shaped 2D MOF nanoplates. [8] However, such bottomup methods are difficult to obtain nanosheets with homogeneous thickness at the nano-level and often require small molecule, capping agent, and surfactants. [9] It is important to note that the presence of such modulators and their removal after synthesis can disturb the shape and properties of 2D MOFs. Therefore, ideally, the shape and morphology of 2D-MOF nanosheet should be precisely controlled in absence of modulators.
Here, we present a hypothesis to modulate shape or morphology of 2D MOF nanosheets starting from same metal node precursor. Regarding the method via hydroxides, B Wang et al., reported a protocol to convert metal hydroxide nanoplates into quasi 1D nanoribbons, but modulation of shape was not included in their study. [10] In this study, we report 2D-MOF nanosheets with controlled thickness and shape, synthesized using hexagonal metal hydroxide monolayer nanosheets with a uniform thickness of 1 nm as a precursor. Two different morphologies were obtained depending on the orientation of the metal hydroxide monolayer nanosheets (random or fixed) and the diffusion rate of the metal ions during the reaction. The detailed structure of the 2D-MOFs was further confirmed by high resolution transmission electron microscope (HR-TEM) and selective area electron diffraction (SAED).

Results and Discussion
The monolayer Ni(OH) 2 nanosheet was used as a precursor of Ni species in Ni-MOF. The nanosheets were obtained by following previous work [11] (See Supporting Information for detailed process). Layered nickel hydroxide was exfoliated in formamide to obtain monolayer nickel hydroxide nanosheets with a thickness of 1 nm (1 mg solid in 1 mL solvent). Monolayer thickness and other characterizations were confirmed through atomic force microscopy (AFM), HR-TEM, and SAED ( Figure  S1, Supporting Information). 2D Ni-MOF of chemical composition Ni 3 (BDC) 2 (OH) 2 (H 2 O) 4 (referred as Ni-BDC-MOF) (CCDC no. 638866) was synthesized by mixing the 10 mL DMF dispersion containing 0.16 mmol Ni(OH) 2 monolayer nanosheet with 10 mL DMF/water (7:3) solution of 0.13 mmol number of 1,4benzenedicarboxylic acid (BDC) and heating at 90°C for 0.5-12 h under air-tight condition (Elaborated process in Supporting Information). Resultant product was characterized through powder Xray (PXRD), HR-TEM, SAED, AFM, field emission scanning electron microscope (FE-SEM), and thermogravimetric (TG) analysis. MOFs obtaining in this method will be now addressed as P-NiMOF-t-T (P-NiMOF denotes powder species obtained in solution method; t = time and T = temperature of reaction). The PXRD pattern of P-NiMOF-12-90 has matched with the simulated pattern of CCDC no 638866 [12] (Figure 1a). P-NiMOF-12-90 showed rectangular sheet like homogeneous morphology, observed under FE-SEM (Figure 1b). Its crystal structure was further confirmed from SAED (from HR-TEM). SAED has a perfect match with the predicted pattern generated from (010) miller plane of Ni-BDC-MOF (Figure 1c,d).
To check the effect of temperature on MOF formation, reactions were performed under various temperatures. This was carried out by allowing the similar amount reaction mixture as mentioned above at different temp, 60 to 120°C for 12 h. PXRD pattern indicates formation of MOF formation over 60°C ( Figure  S2a, Supporting Information). It suggested that 60°C was not effective for formation of P-NiMOF. Over threshold temperature (60°C) there was no effect on morphology (observed through FE-SEM; Figure S2b, Supporting Information).
To observe the effect of total reaction time at a fixed temperature, solution phase synthesis was performed for various time; 0.5-24 h at 90°C. Figure 2 shows PXRD and FT-IR of the obtained solids at each step. In Figure 2a, the lowest angle signal at 7.66°of Ni(OH) 2 was observed in the PXRD patterns before reaction. This peak gradually shifts right as the reaction proceeds further in the direction of P-NiMOF formation. More specifically, the signal at 7.66°has shifted to right and eventually the signal at 9.24°appeared. This 9.24°signal is signature of (100) miller plane of P-NiMOF-12-90. New signals at around 18.2°and 23.84°w ere started to appear after the reaction of 4 h. The complete conversion of Ni(OH) 2 nanosheets into P-NiMOF was observed after the reaction of 8 h. Moreover, the significant signal of Ni(OH) 2 around 33.7°for (100) face of Ni(OH) 2 disappeared after the reaction of 8 h ( Figure S2c, Supporting Information). P-NiMOF-t-T with a 2D structure (2D-Ni-MOF) formation also has been observed through FT-IR analysis ( Figure 2b). Infrared frequency of metal carboxylate bond (symmetric and asymmetric stretching) has started to appear when conversion of nickel hydroxide to P-NiMOF begins. COsym and COasym for Ni-BDC MOF has appeared on 1580 and 1370 cm −1 (Δ = 190 cm −1 ). [13,14] This signal started to be strong after the reaction time of 4 h. This observation is in strong agreement with PXRD data. Shape of all resultant products (P-NiMOF) was observed by FE-SEM ( Figure S2d, Supporting Information). Further confirmation of crystallinity was obtained from HR-TEM (See Figure S2e-h, Supporting Information). Height of stacked rectangular nanosheets was determined through AFM measurements ( Figure S2i(a), Supporting Information). Thermal analysis was conducted through TG analysis ( Figure S2i(b), Supporting Information). BET surface area of P-NiMOF is 29 m 2 g −1 which tells us that it is a non-porous material ( Figure S2i(c), Supporting Information). Noteworthy feature is that the transformation of Ni(OH) 2 nanosheets into MOF is different from the usual MOF synthesis using free Ni 2+ ion species. In general, metal salts dissolve completely at the first step of reaction. Here, Ni(OH) 2 nanosheets are not being dissolved in water/DMF solution with BDC ligand at the very beginning. As the temperature increases, the Ni(OH) 2 nanosheets begin to react with the BDC ligands.
So, it is necessary to observe the effect of BDC concentration on the proceeding of the reaction. An experiment was conducted where Ni(OH) 2 :BDC molar ratio was taken as 10:1, 5:1, 1:1, and 0.5:1, respectively. All reactions were carried out in air-tight vessel at 90°C for 12 h (experimental details in Supporting Information). PXRD pattern ( Figure S2j, Supporting Information) of the ratio 10:1 at 90°C for 12 h was similar with the product obtained only after reaction of 6 h at 90°C with ratio 1:1 (P-NiMOF-6-90). From these observations, it has been cleared that higher concentration of ligand accelerates the reaction and results plate like structure when BDC is excess ( Figure S2k,l, Supporting Information).
As a proof of applicability of this process of converting metal hydroxide nanosheets into MOF nanosheets, Co(OH) 2 nanosheets' monolayer dispersion was converted into Co-BDC MOF following same process as Ni-MOF synthesis. Details of the process and results have been described in Supporting Information ( Figure S2m, Supporting Information).
In the substrate assisted synthesis, the reaction was conducted on Si wafer and TEM grid (experimental details in Supporting Information). It is named as S-NiMOF. A hexagonal morphology was obtained in this case. The proceeding of the reaction was also monitored by comparing the height profile of nano sheets on Si-wafer before and after the reaction through AFM (Figure 3a,b), where the same nickel hydroxide nanosheets were assessed by AFM before and after the reaction. Surprisingly, the hexagonal shape of Ni(OH) 2 nanosheet was maintained after reaction, while the thickness was increased from 1.05 to 2.00 nm. A statistical measurement of heights of nanosheets also shows around 1 nm height difference between before and after reaction with BDC ligand (Figure S3a Figure S3a(d-f), Supporting Information). Thinner area where hexagonal morphology could be clearly seen through HR-TEM did weakly diffract (Figure 3c,d). However, closer miller plane like (200)'s ED was observed in thicker area's SAED (Figure 3e,f). This electron diffraction pattern (Figure 3f) has matched with the predicted ED pattern generated from 2 nm thick (001) miller plane of Ni 3 (BDC) 2 (OH) 2 (H 2 O) 4 as shown in Figure 4a,b. However, for more concluding evidence, similar substrate reaction on Si-wafer was conducted with high amount of loading of Ni(OH) 2 NS (See Supporting Information, Figure S3b,c, Supporting Information). It was named as S-NiMOF-high. S-NiMOFhigh gives very broad PXRD pattern ( Figure S3c(d), Supporting Information, out of plane measurement condition). ED of exfoliated product of Si-wafer surface gives evidence of formation of MOF like crystal arrangement ( Figure S3c(a,b), Supporting Information).
With regard to the reaction of Ni(OH) 2 nanosheets on Si wafer and TEM grid, it is assumed that the movement of nickel hydroxide nanosheets is restricted, which slow down the dissolution rate of Ni 2+ ions during the reaction. As soon as Ni 2+ ions are being released, they are becoming the part of Ni-BDC-MOF by reacting with BDC ligand. If the dissolution/diffusion rate of nickel hydroxide nanosheet on Si substrate is faster, the coordination sites of the metal centers cannot be blocked during the conversion reaction to the 2D-MOF structure and the hexagonal morphology may be destroyed. To confirm this hypothesis, two other similar reactions were carried out. First, the BDC ligand solution was placed in a stirred state ( Figure S3d, Supporting Information). Under these strong stirring conditions, the diffusion rate of Ni(OH) 2 nanosheets into solution is faster than the rate of 2D-MOF formation. Therefore, under such conditions, the hexagonal morphology of nickel hydroxide is destroyed. Actually, in AFM image, former hexagonal sheets were appeared as a bunch of particles ( Figure S3e, Supporting Information, and FE-SEM image in Figure S3f   nanosheets was happening due to stirring condition. It was confirmed through a controlled reaction (Supporting Figure S3g,h, Supporting Information). Another observation is worth to mention that Ni(OH) 2 nanosheets in dispersion also are not being soluble in the DMF/water solvent in the absence of BDC ligand.
Second, BDC concentration was ten times increased than previous condition, but reaction was allowed under static condition without further addition of any mechanical or magnetic stirring condition. This higher concentration made the dissolution rate of Ni 2+ ion into solution as well as rate of reaction faster. This time also, hexagonal morphology was not found, in accordance with AFM observation ( Figure S3i-k, Supporting Information). Now, assuming that these two results on the reactivity of nickel hydroxide on Si substrates also apply to reactions in solution, in reactions with a 1:1 ratio of BDC to Ni(OH) 2 nanosheets (corresponding to the reaction conditions to synthesize the P-NiMOFs shown in Figures 1 and 2), the shape of the resulting 2D-Ni-MOF should retain the initial hexagonal morphology of Ni(OH) 2 nanosheet. However, as mentioned above, the FE-SEM showed rectangular shape instead (Figure 1b and Figure S2b, Supporting Information). A possible explanation of this anomaly might be the fact that as reaction proceeds, stability of nanosheets dispersion decreases and at the time 0.5 h, aggregations of nanosheets happens ( Figure S2b, Supporting Information). In this case, the aggregation is thought to result in the proximity of nanosheets and intermediate species, and promote secondary nucleation based on dissolution and reprecipitation, which leads to the formation of rectangular sheet-like structures as shown in Scheme 1. In addition, because of the ease of nanosheet migration in solution-phase synthesis the volume expansion of the nanosheets during the conversion of hydroxide nanosheets into MOFs occurs without steric limitation, accelerating the transformation of the hexagonal structure of nickel hydroxide. SAED of P-NiMOF-12-90 matches with the predicted ED pattern of (010) plane of Ni-BDC-MOF, Ni 3 (BDC) 2 (OH) 2 (H 2 O) 4 (Figure 4d,e). In the Ni-MOF structure corresponding to the (010) plane shown in Figure 4d, the BDC and Ni 2+ layers appear alternately in a specific in-plane direction. In the 2D-MOF formation reaction, the alternating growth of BDC and Ni 2+ layers from the edges of the hexagonal nanosheets by the dissolution-reprecipitation reaction might lead to the elongation of the hexagonal nanosheets in a specific direction. So, the final product becomes rectangular nanosheet instead of hexagonal nanosheet and very few elongated hexagons also observed ( Figure S2k,l, Supporting Information). On the other hand, the reaction in which nickel hydroxide fixed on Si substrates is converted to Ni-MOF leads to the retention of morphology, as Ni 2+ diffusion and growth direction are restricted.
These two conditions were achieved in the case of surface synthesis and hexagonal MOF nanosheets were found. Although it could not be denied that crystallinity is very week where hexagonal shaped MOF nanosheets were observed (Figure 3d), the ED pattern of the thicker area was matched with the predicted ED pattern of (100) plane of Ni-BDC-MOF (Figure 3f). This indicates that hexagonal nanosheet has a structure similar to that of Ni-BDC-MOF oriented in the (001) plane as shown in Figure 4a. The hexagonal 2D-Ni-MOF nanosheet shown in Figure 3b maintains the hexagonal structure of the starting nickel hydroxide, while the thickness is increased by a factor of about two. In other words, they expanded in the longitudinal direction, indicating that BDC ligand could come proximity to Ni metal center along the +c axis. In this reaction, Si-wafer makes inaccessible the other side of Ni-metal center for BDC ligand except for access from the longitudinal direction (c axis). Therefore, volume expansion is occurring along c axis (for a visual explanation, please refer to the Scheme 1). Also, it is a noteworthy fact that secondary building unit of Ni-BDC MOF is Ni-OH unit and the crystal parameter of Ni-OH unit, for example, Ni-Ni distance, perfectly matches with Ni-Ni distance of Ni(OH) 2 (Figure 4c,f) distance is 0.31 nm. [15] This suggests that Ni(OH) 2 can be transformed into Ni-BDC-MOF structure by keeping the hexagonal morphology intact when BDC ligand is approaching to Ni center along the c axis.
For a comparison purpose, similar crystal structured MOF (CCDC No. 638866) was synthesized from Ni(NO 3 ) 2 by following the reported protocol [12] and this did not produce nanosheets ( Figure S4a,b, Supporting Information). So, it also can be stated that 2D MOF nanosheets were realized as an effect of shape, size, and thickness of metal precursor.
As a proof of concept, 0.26 mm thick free-standing thin film was successfully prepared from P-NiMOF. This ultrathin MOF film could be used as proton conducting film. It showed 4 × 10 −5 S cm −1 proton conductance at 85°C and 100% relative humidity (RH) with activation energy of 0.85 eV (from Arrhenious plot) . Unlike this, Ni-BDC MOF (produced from Ni(NO 3 ) 2 ) could www.advancedsciencenews.com www.advmatinterfaces.de not be used to produce ultrathin free standing film. In this case, only a minimum thickness of 0.55 mm could be prepared. Wide nanosheet (large aspect ratio) morphology of P-NiMOF facilitates ultrathin thickness [16] (Figure 5 and Figure S5, Supporting Information).

Conclusion
In conclusion, from the same nickel hydroxide nanosheet precursor, two different morphologies of the same 2D MOF were achieved by varying reaction conditions. A rectangular morphology had been obtained when the movement of nickel hydroxide nanosheets is faster (in solution state) and hexagonal morphology is found, when the movement of hexagonal nanosheets as well as Ni 2+ ion dissolution and diffusion rate both were slower (on the surface reaction). Controlling thickness of rectangular nanosheets is difficult. However, reactions at the surface have made it possible to control thickness and morphology at the nano level. In this work, uniform 2D-MOF nanosheets with lateral size as big as 1 μm and thickness as small as 2 nm were synthesized through simple one pot reaction at temperature as low as 90°C without any chemical modulator. Realization of hexagonally shaped 2D-Ni-BDC MOF nanosheets with a thickness of around 2 nm is expected to lead to the discovery of new physical properties and applications in nanostructured materials. Although the discovery of such properties and their application to nanostructured materials is a future challenge, this process of converting nanomaterials from other nanomaterials may provide a new platform for the synthesis of 2D MOFs.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.