PROCESS FOR CONDITIONING METAL-ORGANIC FRAMEWORKS BY MEANS OF MEMBRANE FILTRATION

Abstract

The present invention relates to a process for conditioning a raw suspension comprising at least one metal-organic framework and at least one suspension medium by means of at least one membrane filtration to obtain a product suspension. The invention relates also to a method, wherein said product suspension is coated to at least part of the surface of a substrate.

Claims

1-15. (canceled)

16. Process for conditioning a raw suspension SR comprising at least one metal-organic framework and at least one suspension medium SM1, wherein the at least one metal-organic framework comprises at least one at least bidentate organic compound coordinated to at least one metal ion and wherein the at least one metal-organic framework is Al-fumarate, wherein the raw suspension SR is conditioned by means of at least one membrane filtration, to obtain a product suspension SP, comprising the at least one metal-organic framework and at least one suspension medium SM2, wherein the at least one membrane filtration comprises at least one purification step, or at least concentration step, or at least one purification step and at least one concentration step and wherein the at least one purification step and/or the at least one concentration step is carried out as diafiltration.

17. The process according to claim 16, wherein the at least one membrane filtration comprises at least one purification step and at least concentration step.

18. The process according to claim 17, wherein the diafiltration medium is selected from the group consisting of water, methanol, ethanol, i-propanol, n-butanol, i-butanol, sec-butanol or a mixture of two or more thereof.

19. The process according to claim 16, wherein the separation layers of the membrane has a pore size in the range from 50 to 800 nm.

20. The process according to claim 16, wherein suspension medium SM1 and suspension medium SM2 are independently from each other selected from water, methanol, ethanol, i-propanol, n-butanol, i-butanol, sec-butanol or a mixture of two or more thereof.

21. The process according to claim 16, wherein suspension medium SM1 and suspension medium SM2 are equal.

22. The process according to claim 16, wherein the raw suspension SR is obtained by a method for the preparation of the at least one metal-organic framework, comprising reacting at least one metal salt with at least one at least bidentate organic compound, in a reaction medium, wherein the reaction medium corresponds to the at least one suspension medium SM1.

23. Method for coating at least part of a surface of a substrate with an active layer, comprising at least one metal-organic framework, comprising a1) a process according to claim 16 for conditioning a raw suspension SR to obtain a product suspension SP, a2) bringing at least part of the surface of the substrate into contact with a coating composition, wherein the coating composition comprises the product suspension SP obtained according to step a1) and the coating composition further comprising at least one binder.

Description

EXAMPLES

[0198] FIG. 1: Schematic of the Membrane Unit

[0199] FIG. 1 shows a sketch of a filter unit. The filter unit comprises a pumped circuit comprising a feed vessel, a pump 1, a thermostat and a membrane module. A pressure meter (PI) and a flow meter (FI) are integrated into the circuit upstream of the membrane filter. (LI) represents a level indicator for the feed vessel. In addition to the pumped circuit, the filter comprises a reservoir for the diafiltration medium or permeate, which can be metered into the feed vessel by means of pump 2. Permeate obtained can be discharged into a permeate container and weighed by means of a scale.

[0200] A. Preparation of MOF

[0201] Example of Al Fumarate Synthesis (1 h, 60° C.)

[0202] Raw Materials:

TABLE-US-00002 Fumaric acid 0.211 mol 24.47 g 24.47 g Sodium hydroxide 0.633 mol 25.32 g 25.32 g Aluminum sulfate * 0.105 mol 70.0 g 70.0 g 18 mol water Water 36.8 mol 661.7 g 661.77 g

[0203] Aluminum sulfate was dissolved in 300 g of water in a beaker, and heated up to 60° C. To this solution a mixture of 411.5 g of solution (at 60° C.) of fumaric acid, sodium hydroxide and water was added during the 58 minutes. During the pumping time a white precipitate appeared. Once the solutions was exhausted, white precipitate was filtered off and washed 3 times with 500 ml of demineralized water. Obtained filter cake was dried in the air at 100° C. until the material appears as dry powder, and transferred to a vacuum dried at 130° C. for the 16 h to remove residual water.

[0204] Preparation yields approximately 33.08 g of material.

[0205] B. Microfiltration Experiments

[0206] 1. General Experimental Setups

[0207] The membrane filtration experiments followed the basic schematic of FIG. 1 and were conducted in batch operation mode. The slurry comprising the metal-organic framework was filled into a feed vessel and circulated continuously over a membrane module, from which a permeate stream was removed. To concentrate/purify (“diafiltrate”) the solution, either additional MOF-containing slurry or VE water was added into the feed vessel. Further concentration of the MOF was enabled by having no feed flow, i.e. by reducing the total volume of the slurry in the filtration loop. During the experiment, all pressures (pressures before and after the membrane module), temperature and cross-flow velocity were controlled. Moreover, the conductivity in the feed vessel was measured; giving an indication about the progress of the purification.

[0208] Characteristic values for a membrane separation experiment are the total mass concentration factor (CF) and the diafiltration factor (DF).

[0209] CF relates the total permeate mass to the initial mass in the feed vessel.

CF=(m.sub.permeate+m.sub.feed,initial)/m.sub.feed,initial.

[0210] By that definition, assuming a 100% retention of the MOF, the Concentration Factor CF represents the relative concentration of the Basolite A520 compared to its initial concentration (e.g. CF=1 in the beginning, CF=2 when m.sub.permeate=m.sub.feed,initial and the solid concentration is doubled).

[0211] DF represents the “washing factor”, i.e. relating the amount of added VE water to the batch size. By this definition, DF is equal to the total permeate mass divided by the initial feed.

DF=m.sub.permeate/m.sub.feed,initial

[0212] Further characteristics of the units are summarized in Table 1 below. It is important to note, that in membrane unit 2, a membrane pump that enables a gentler conveyance of the medium was applied, as compared to a centrifugal pump in membrane unit 1.

TABLE-US-00003 TABLE 1 Characteristics of the two membrane units applied for the lab-scale membrane experiments Criterion Membrane unit 1 Membrane unit 2 Type of pump Impeller-type Membrane pump centrifugal pump Max. pumping 600 600 flow (L/h) Applied for Lab-scale modules Lab-scale and experiments (single-channel technical-scale ceramic) modules (multi- channel ceramic) Automation Manual operation Fully automated

[0213] 2. Experimental Plan and Process Evaluation

[0214] In total, two batches with Basolite A520 slurry were applied. In Table 2 below, the analytical properties of these two batches are summarized. As can be seen, both batches showed Basolite A520 concentrations of approximately 4.5 wt.-% and Na.sub.2SO.sub.4 concentrations of approximately 5-6 wt.-%, based on the total weight of the batch. The concentrations were measured after insertion into the respective membrane units which lead to a slight dilution of the slurry.

TABLE-US-00004 TABLE 2 Properties of the initial Basolite A520 batches Parameter Batch 1 Batch 2 Total solid concentration (wt.-%) 10.0-10.1 10.0 Na.sub.2SO.sub.4 5.2-5.6 5.2 concentration (wt.-%) Basolite A520 concentration (wt.-%) 4.5-4.8 4.8

[0215] The parameters of the conducted experiments are summarized in Table 3. In total, three different experiments were conducted, MF1″, MF2″ and MF3″; aiming at a variation of the type of membrane and aiming at a maximum concentration of the Basolite A520 slurry (using a maximum attainable CF of 5 in MF3). The maximum CF was determined by the experimental setup (minimum holdup, pumps). In all experiments, similar—non-optimized—process conditions based on experience were applied.

TABLE-US-00005 TABLE 3 Parameters applied for purification and concentration of Basolite A520 Parameter MF1 MF2 MF3 Membrane unit Unit 1 Unit 1 Unit 2 Membrane Atech 0.2 Atech 0.05 Atech 0.05 μm (MF μm (UF μm (UF 20n) 50A) 50A) Module geometry Single-channel, 6 mm inner diameter, 1 m length (0.019 m.sup.2) Basolite A520 Batch Batch 1 Batch 1 Batch 2 Temperature (° C.) 40 40 60 TMP (bar) 1 1   1*** Cross-flow velocity (m/s) 4 Diafiltration factor (DF) 8 8  9 Concentration factor (CF) 2.5* 2.5*   5** *limited by minimum hold-up in membrane unit 1, **limited by viscosity of the concentrated Basolite A520 slurry. ***the TMP was increased at a later stage of the experiment to increase permeate fluxes.

[0216] To evaluate the concentrations of Na.sub.2SO.sub.4 and Basolite A520, analytics of total solids content analysis using evaporation and sulfate elemental analysis at GMC for determination of Na.sub.2SO.sub.4 concentrations were applied. The main assumption in determining the concentrations of Basolite A520 is that the slurry only consists of Na.sub.2SO.sub.4, Basolite A520 and water. By that, the concentration of Basolite A520 can be calculated by the following formula:

c(Basolite A520)=c(total solids)−c(Na.sub.2SO.sub.4).

[0217] Thus; it is possible to calculate the retention of Basolite A520 by the membrane, given by the formula

R(Basolite A520)=1−c(Basolite A520,permeate)/c(Basolite A520,retentate)

indicating 100% retention in case no Basolite A520 is found in the permeate and 0% retention of both concentrations c(Basolite,A520,permeate) and c(Basolite,A520,retentate) are the same. The same calculations can be done for the retention of Na.sub.2SO.sub.4. Samples in both the retentate and the permeate were collected at defined DF and CF.

[0218] The MF experiments were evaluated by three criteria: [0219] (1) Permeate flux as function of DF/CF. [0220] (2) Retention of Basolite A520. [0221] (3) Na.sub.2SO.sub.4 removing capability as function of DF (indicated by a decrease in conductivity and Na.sub.2SO.sub.4 concentration in the retentate loop).

[0222] Moreover, water flux measurements before and after the experiments were conducted at a TMP of 0.5 bar for evaluating the magnitude of membrane fouling effects. During the experiments, back flushing with permeate was applied at certain intervals to check its effect on membrane performance. As it was observed that in the end a lot of material was sticking to the walls (e.g. on the feed glass vessel), it was washed (cleaned) twice with VE water and both samples were analyzed for total solids concentration (which in the end was almost equal to the concentration of Basolite A520 as almost no salts were present).

[0223] 3. Experimental Results

[0224] In FIGS. 2 and 3, the MF permeate fluxes (left axes) and retentate conductivities (right axes) are shown as functions of the DF and CF respectively. Conductivity has been determined by measuring 100 mL of wash water with the standard conductivity measuring electrode.

[0225] The performance of both membranes (0.2 μm and 0.05 μm) was very similar. This indicates that, as expected, a boundary layer controlled permeation mechanism occurred. Both membranes reached very high permeate fluxes between 200 and 550 kg/m.sup.2 h during the diafiltration step, which was then decreasing to values of about 50 kg/m.sup.2 h during concentration by a factor of approx. 2.5. Concerning conductivities, both membranes lead to final conductivities well below 500 μS/cm; with an extensive decline in the beginning of the experiments but a lower washing out effect at the end of the experiments (note the logarithmic scaling for conductivity). The decrease in conductivities during the course of the last DFs was very low which may either be explained by a blocking effect at elevated washing factors (very unlikely: also with a pre-concentration by a factor of 2 in MF3, conductivities decreased significantly during diafiltration) or by the conductivity of the Basolite A520 itself.

[0226] The water flux measurements at TMP=0.5 bar before and after the Basolite A520 filtration experiments, showed the following: Water fluxes were 700 kg/m.sup.2 h before and 1280 kg/m.sup.2 h after the filtration in MF1, 574 kg/m.sup.2 h and 555 kg/m.sup.2 h in MF2, and 565 kg/m.sup.2 h and 562 kg/m.sup.2 h in MF3; thus indicating a very low magnitude of membrane fouling.

[0227] In Detail,

[0228] FIG. 2 shows permeate fluxes and conductivities in the retentates as function of DF or CF in experiments 1 and 2. In both experiments, first a diafiltration followed by a concentration were conducted. During both experiments, the effect of back flushing on the permeate flux was tested. It was shown that this effect was minor or only visible for a small time interval (see spikes in the graphs).

[0229] FIG. 3 shows permeate fluxes and conductivities in the retentates as function of DF or CF in experiment 3. In this experiment, an initial concentration of the Basolite A520 slurry was followed by a diafiltration and by a final concentration. The conductivities shown are the ones during the diafiltration (a manual measurement device had to be used due to plugging of the online conductivity measurement device during the first concentration).

[0230] In Table 4, the resulting properties of the final retentates of the three experiments are summarized. As can be observed, the difference between retentate MF1 and MF2 was very low; each reaching a final Basolite A520 concentration of approx. 9.7 wt.-% and very low amounts of Na.sub.2SO.sub.4.

TABLE-US-00006 TABLE 4 Resulting Basolite A520 slurries after diafiltration and concentration Final Final Final Parameter retentate MF1 retentate MF2 retentate MF3 Total solid 9.85 9.81 19.07 concentration (wt.-%) Conductivity (μS/cm) 413 291 391 Na.sub.2SO.sub.4 0.12 0.15 0.15 concentration (wt.-%) Basolite A520 9.73 9.66 18.85 concentration (wt.-%)

[0231] 5. Conclusion

[0232] The conducted experiments show, that it was possible to purify Basolite A520 slurries using DI water by means of membrane filtration. As can be seen in above tables, the concentration of Na.sub.2SO.sub.4 could be reduced form about 5.5% by weight to values below 0.15% by weight. The needed diafiltration volumes for that is 8-9 times the initial volume.

[0233] Furthermore, the experiments show that the concentration of Basolite A520 in the slurries could be increased from 4.5% respectively 4.8% by weight up to values of about 10% by weight respectively about 19% by weight.

[0234] C. Coating Experiments

[0235] The aim of the slurry characterization and coating experiments was to show a general proof of principle and set the process in relation to other processing alternatives such as conventional spray drying and dispersing the filter cake.

[0236] 1. Experimental

[0237] a) Materials:

[0238] Substrate [0239] Substrate 1: Nippon foil (1N30) 120mm×20 μm was used as substrates for peel test and film density measurements.

[0240] Three different substrates were used to test dip coating of complex geometries: [0241] Substrate 2: Segment of finned tube heat exchanger, aluminum plates with tube fitting, 390 g/m.sup.2, fin pitch 2 mm. [0242] Substrate 3: Single finned tube heat exchanger plate, 311 g/m.sup.2. [0243] Substrate 4: Polypropylene corrugated plastic, 287 g/m.sup.2.

[0244] MOF-Materials: [0245] Aluminum fumarate samples “Final Retentate of MF1 , 2 and 3” as described above, hereafter referred to as MF1-MOF, MF2-MOF and MF3-MOF [0246] conventionally filtrated and spray dried material (Basolite A520; BASF SE), hereafter referred to as SPD-MOF. [0247] Laboratory sample of dispersed filter cake (“WFC”, Aluminum fumarate). hereafter referred to as WFC-MOF

[0248] Binder:

[0249] A water based polyacrylate binder (Joncryl 3030; BASF SE, M.sub.w>200000) was used at a solid content of 19.6 wt. % with regard to the total dry mass of the coating.

[0250] Primer

[0251] Polyethyleneimine (Lupasol PS; BASF SE, about M.sub.w750000) was used as primer. If not stated explicitly coatings were applied without primer layer.

[0252] b) Preparation of the Coating Compositions

[0253] Water was deposited into a polypropylene mixing container and SPD-MOF powder was added under manual stirring. Subsequently the binder dispersion was added and dispersed for 3 min at 2000 rpm using a centrifugal mixer (ARE310; Thinky inc).

[0254] The Polyethyleneimine solution (Lupasol PS; BASF SE) was diluted from its original solid content of 32.76 wt. % to 6.55 wt. % with demineralized water.

[0255] To give an indication whether the slurry is stable under process conditions, it was stirred by a magnetic rod in a 300 ml beaker at 100 rpm. Storage stability was tested by aging the slurry in a closed container, without stirring at ambient temperature (˜20-25° C.). In both cases viscosity was measured in time intervals after preparation.

[0256] c) Coating

[0257] The particular coating compositions were coated onto the substrate using a knife coating system (ZAA2300; ZUA2000; Zehntner GmbH) with a 200 μm gap setting at 150 mm/s and a deposition volume of 8 ml.

[0258] The primer solution (6.55wt. % Lupasol PS) was coated onto the substrate using a 25 μm wire bar applicator that was pulled over the substrate at 30 mm/s and a deposition volume of 2 ml.

[0259] d) Drying

[0260] The coating was dried by contact drying for 10 min at 60° C. using an electrically temperature controlled vacuum plate (ACC188.230; Zehntner GmbH) and subsequently for >1 h at 110° C. in a drying cabinet (UN110Plus; Memmert).

[0261] e) Analytics:

[0262] Solid content was measured using a dry-weight-scale (setting 120° C.; HB43-S; Mettler Toledo).

[0263] Viscosity was measured using a cone plate rheometer (MCR102, Anton Paar; PP50; 400 μm gap; 25° C.) at shear rates from 1−/s to 1000−/s.

[0264] Adhesion on a reference substrate was measured by a 90° peel test according to DIN 28510 with 50 mm/min pull rate, 25° C. at a relative humidity of ˜40% rH.

[0265] Density was calculated from a measurement of coating weight (Excellence Plus, 10.sup.−4 g; Mettler Toledo) and thickness (ID-H0530, 10.sup.−4 mm; Mitutoyo Corp) on a 50 cm.sup.2 film sample. The sample was cut out of a coated aluminum substrate using a circular knife cutter (150805; Karl Schroder KG).

[0266] f) Nomenclature:

[0267] A dispersion of Basolite A520 and water (as prepared by membrane filtration or dispersion of SPD-MOF powder or WFC-MOF in water) will be referred to as “MOF premix”. After binder addition and dilution, the final coating formulation will be referred to as “MOF/binder slurry”.

[0268] Properties are determined by the processing route but can be tuned by changing process parameters as well (i.e. feed rate during crystallization, temperature during spray drying). Whether the changes observed here are due to processing route only or may be optimized has not been checked in detail and is outside the scope of this study.

[0269] 2. Properties of the Coating Compositions

[0270] a) Viscosity

[0271] Ideally, coating compositions should have a low viscosity at medium to high shear-rates (100-1000−/s), a high viscosity at low shear rates <10/s, a high yield point, and should be stable under process and storage conditions.

[0272] With respect to a coating application, both the fluid dynamic properties of the coating formulation as well as the properties of the dried film need to be considered. The results presented here give an indication whether the material produced with the membrane filtration process described above is suitable for a coating application but does not represent a thorough characterization for coating process development.

[0273] A coating composition is composed of a solvent (here water), active material (here Aluminum Fumarate (A520)), a binder (here Joncryl 3030) and processing additives. To keep the comparison as simple as possible, binder type, binder content and as far as possible solid content were kept constant. No processing additives were used.

[0274] Ideally, coating composition should exhibit a low viscosity at medium to high shear-rates (100-1000−/s), a high viscosity at low shear rates <10/s, a high yield point, and should be stable under process and storage conditions. Rheological yield point was not measured explicitly but may be estimated to be proportional to zero shear viscosity for similar compositions.

[0275] For a dip coating process, viscosity is limited by two characteristic boundaries. Too high viscosity will lead to blocking due to incomplete draining of the fluid from the coated structure. Too low viscosity will lead to inhomogeneous coating as the fluid forms rivulets running down the substrate. A dip coating process aiming for high coating weights will typically operate close to the upper viscosity limit since this allows for a higher solid content (higher solid content implies that higher dry film thicknesses can be realized and less water needs to be evaporated.) Viscosity may change under process or storage conditions. A change in viscosity under process conditions may indicate destabilization, structure formation or incomplete dispersion and is not acceptable in a technical coating process.

TABLE-US-00007 TABLE 5 Comparison of viscosity flow curve of Aluminium fumerate prepared by the membrane filtration process (MF1-MOF) and spray drying (SD6-MOF) process. Measurement was done at 25° C. and a solid content of 7.5 wt.-%. Material Shear rate [1/s] Viscosity [Pas] MF1-MOF 1 1 MF1-MOF 2 0.9 MF1-MOF 5 1 MF1-MOF 10 0.7 MF1-MOF 20 0.5 MF1-MOF 50 0.2 MF1-MOF 100 0.09 MF1-MOF 200 0.04 MF1-MOF 500 0.02 MF1-MOF 1000 0.013 SD6-MOF 1 0.2 SD6-MOF 2 0.2 SD6-MOF 5 0.2 SD6-MOF 10 0.1 SD6-MOF 20 0.05 SD6-MOF 50 0.025 SD6-MOF 100 0.013 SD6-MOF 200 0.009 SD6-MOF 500 0.006 SD6-MOF 1000 0.005

[0276] A reliable measurement of a premix from re-dispersed WFC-MOF was not possible because the slurry became too low viscous and subsequently unstable during dilution to 7.5% by weight.

[0277] b) Process Stability

[0278] To evaluate process stability, the MOF premix from MF3-MOF and the MOF/binder slurry (MF3-MOF with binder) were measured at different time intervals after end of membrane filtration. No change in viscosity could be observed within the first 5 hours. The slurry can thus be assumed to be stable for process relevant timespans.

[0279] After binder addition no change in viscosity was observed within process relevant time spans as long as the dispersion was continually stirred, as well.

[0280] 3. Results and Discussion

[0281] Since the purification and concentration method can be expected to change particle morphology, it is important to check the impact on film properties. Accordingly, mechanical properties such as cracking, film density and adhesion were evaluated.

[0282] a) Crack Formation

[0283] Shrinkage during drying can cause cracking of a coated film. The tendency of a film to crack depends primarily on composition (e.g. binder type and content, solvent type and surface tension), particle size distribution and morphology and film thickness. Here composition, drying temperature and film thickness were kept constant to show the impact of particle size and morphology.

[0284] Crack formation of three different dried coating compositions, comprising MOF (washed MOF-WFC, Membrane filtrated MF1-MOF and SPD-MOF) were compared.

[0285] It could be observed, that the amount of cracking increased dorm almost none for the WFC-MOF, slight crack formation form the membrane-filtrated MF1-MOF and severe for the SPD-MOF.

[0286] The MF1-MOF thus shows more cracking than the MOF-WFC material but less than the SPD-MOF. The critical cracking thickness for this composition was determined for wet filter cake material to be 120 μm. While cracking can be suppressed e.g. by lower coating thickness or higher binder content, the spray dried material used in this study can be seen as unsuitable for a high capacity coating application.

[0287] b) Film Density

[0288] The density of a coated film is an important parameter for a sorption heat pump coating. As coating thickness will be limited by module geometry and required minimal adhesion for a given composition, density is proportional to the power density of the overall system.

TABLE-US-00008 TABLE 5 Film density of Aluminum-Fumarate films prepared by different processes Material Density [g/cm.sup.3] SPD- MOF 0.34 WFC-MOF* 0.42 MF1-MOF und MF3-MOF** 0.59 *Some air entrainment defects were observed for the WFC-MOF and the actual bulk density may be slightly higher **The value for membrane filtration is an average of coatings produced from MF1-MOF and MF3-MOF material with less than 5% relative deviation.

[0289] c) Adhesion

[0290] The adhesion of a coating depends primarily on composition, drying conditions, film thickness, particle size distribution and morphology, and surface properties. Here, composition and drying conditions were kept constant. Since adhesion cannot be measured reproducibly on complex surface of a heat exchanger fin, measurements were conducted on aluminum foil (Nippon) as reference substrate using a standardized 90°-peel test (see experimental section for details). The required minimal adhesion depends on the operating conditions in the final application but values of less than 5 N/m can be considered insufficient for most technical applications.

[0291] Adhesion of MOF coatings from material produced in experiment 3 by membrane filtration and by re-dispersing filter cake were evaluated.

TABLE-US-00009 TABLE 6 Film adhesion as measured by a 90° peel test as a function of film thickness Material Film thickness [μm] Adhesion [N/m] WFC-MOF 30 25 WFC-MOF 81 14 MF3-MOF 61 43 MF3-MOF 93 24 MF3-MOF 115 17 MF3-MOF 192 14

[0292] As expected the adhesion decreases with coating weight (corresponds to film thickness).

[0293] Adhesion of the MOF-MF3-coatings is lower compared to the WFC MF3-MOF-coatings. However, all measured values are higher than 10 N/m indicating that a binder content of 19.6 wt.-% is sufficient for most applications.

[0294] For SPD-MOF prepared by spray drying and MF-MOF prepared by membrane filtration, an increase in adhesion was observed when a polyethyleneimine primer layer has been applied to the substrates surface (PCT/EP2017/071096). This effect was reproduced with the membrane filtrated material MF3-MOF as summarized in Table 8.

TABLE-US-00010 TABLE 7 Impact of a polyethyleneimine primer coating and a water bath test Adhesion [N/m] No primer, as prepared 14 No primer, after water bath test 11 With primer, as prepared 24 With primer, after water bath test 23.5

[0295] Even though the primer is water based, adhesion remains high after immersion in boiling water for 3 min (denoted as “water bath test”). The water bath test emulates the harsh operating conditions in a sorption heat pump by immersing the coating in boiling water for 3 min.

[0296] d) Dip Coating of Complex Geometries

[0297] Depending on the application, MOF-coatings need to be applied onto different substrates. Substrates may differ with respect to surface material and geometry. Complex geometries may lead to highly inhomogeneous coating weight as fluid can accumulate in recesses or completely block narrow gaps in a tube finned heat exchanger. Polymeric surface may lead to de-wetting or low adhesion. The viscosity of the MOF-coating compositions needs to be adapted to the target geometry. Too high viscosity will lead to blocking or too high coating weights while too low viscosity may cause rivulet formation and very low coating weights.

[0298] The substrates were hand dipped, dried at room temperature for 30 minutes and afterwards at 110° C. for more than 1 hour. Three exemplary geometries were coated with Basolite A520 MOF-MF1 mixed with polyacrylic binder and diluted with water to a suitable solid content. The binder mass fraction in solid was kept constant at 19.6% by weight, no processing additives were used and the solid content was varied between 12.5 to 15.5 by weight to adapt viscosity. [0299] Substrate 2: Segment of finned tube heat exchanger, aluminum plates with tube fitting, 390 g/m.sup.2, fin pitch 2 mm, 12.5% by weight [0300] Substrate 3: Single finned tube heat exchanger plate, 311 g/m.sup.2, solid content (MOF) 15% by weight [0301] Substrate 4: Polypropylene corrugated plastic, 287 g/m.sup.2, 12.5% by weight

[0302] Though some cracking can be observed for the highly corrugated polypropylene substrate, all coatings show good adhesion and only minor visible defects.

[0303] 4. Conclusion

[0304] In summary, the MF-MOF-containing coating compositions according to the invention show several improved properties over coating compositions comprising SPD-MOF or WFC-MOF.

[0305] In particular, the coating compositions according to the invention show a reproducible high density of the MOF-layer in combination with sufficient high critical cracking thickness. No significant differences in the properties of coating compositions or MOF-films could be observed between membrane filtration experiments MF1, MF2 and MF3. A compact overview of the results is given in Table 6. Moreover, the energy consumption is significant lower for the inventive coating compositions.

TABLE-US-00011 TABLE 8 Estimate values for important process, slurry and film properties Process Adhesion Crack Slurry energy (90° formation viscosity consumption peel (critical [Pas], (number of Film test cracking 7.5% by drying Density @80 m thickness, weight at steps) [g/cm.sup.3] [N/m] μm) 10/s SPD-MOF (3) 0.33 21 ~20 0.1 WFC-MOF (2) 0.41 34 ~120 <0.04 MF1-MF3 (1) 0.6 14 ~80 0.6