708 Communication Chemie Ingenieur Technik Heat and Mass Transfer Properties of Zeolite Coatings: Comparison of Reactive- and Spray-Coated Systems Rajesh Chanda, Lisha Wang, and Wilhelm Schwieger* DOI: 10.1002/cite.201700150 Coated materials are widely used in the process industry. For the generation of adsorption heat exchangers for heat pumps, coated systems offer compactness, robustness as well as high performance. The heat and mass transfer properties into and through functional coating layers play a prominent role in the overall performance of the whole coated systems. In this contribution, laser flash analysis and sorption measurement methods are applied to compare the heat and mass transfer properties of the common spray coating and reactive coating. The experimental results indicate that the reactive coating technique leads to layers with better heat transfer properties and in contrast to lower mass transfer. Keywords: Coating, Diffusion coefficient, Laser flash analysis, Thermal conductivity, Zeolite Received: October 31, 2017; revised: January 10, 2018; accepted: January 24, 2018 1 Introduction The adsorption heat pump (AHP) is an environmentalfriendly heating and cooling system that utilizes energy sources such as solar or waste heat instead of electrical energy as the driving energy [1, 2]. Therefore, with the raising interest in the recent years, research has focused on optimizing the efficiency of AHPs. The adsorption heat exchanger arrangements play a crucial role in the overall performance of the heat pumps. Among the studied adsorption heat exchanger systems, coating-based heat exchangers offer the best mass and heat transfer performances [1], as the direct application of sorbent materials by coating reduces the thermal contact resistance between the heat exchanger and the adsorbent [3]. Several coating techniques using glue, binder [4], or no binder [5] have been reported. The coating application technique influences the coating layer properties, i.e., packing density and the thickness of the layer, contact between support and coating layer. This plays a significant role in the overall heat and mass transfer behavior [6–8]. Despite the utmost importance, there are not many experimental analysis methods available in the literature to investigate heat and mass transfer properties of coated systems. Zeolites have essential industrial applications owing to their unique microporous network and easily tunable acidity/basicity [9]. Among the zeolites, zeolite Y and SAPO-34 are extensively used in various applications, e.g., as catalyst in FCC [10] and MTO reactions [11] or as molecular sieves [12]. In recent years, both of these materials, due to their high water sorption capacity (~ 0.3 – 0.35 g g–1), are prominently applied as adsorbents for adsorption heat pumps [13]. However, these two have different material properties. www.cit-journal.com Zeolite Y having FAU- morphology is an aluminosilicate [(SixAlYNam)O2] with a silicon framework density (FDSi) of 13.3 T-atoms 1000 Å–3 and an accessible volume of 27.7 %. On the other hand, zeotype SAPO-34 having CHA morphology is an (silico-)aluminophosphate [(SixAlYPz)O2] with an FDsi of 15.1 T-atoms 1000 Å–3 and accessible volume of 17.3 % [14]. Therefore, these materials are used in the present study to compare the influence of the material structural properties on the coating layer heat and mass transfer properties. Two different coating techniques have been applied in the sample preparation. Spray coating using a binder as an additional component is applied as a representative of the common coating method. Reactive coating, based on the partial support transformation (PST) technique, is used to achieve coating layers without the use of any binder. It is known to have inter-twinned nature contact with coating layers and the support, resulting in higher surface contact [5]. In this contribution, laser flash and sorption-kinetic methods are applied to determine the thermal conductivity and diffusion coefficient of zeolite (SAPO-34 and Y) coating layers on aluminum plates to investigate heat and mass transfer properties. – Rajesh Chanda, Lisha Wang, Wilhelm Schwieger wilhelm.schwieger@crt.cbi.uni-erlangen.de Friedrich-Alexander-Universität Erlangen-Nürnberg, Lehrstuhl für Chemische Reaktionstechnik, Egerlandstraße 3, 91058 Erlangen, Germany. ª 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Ing. Tech. 2018, 90, No. 5, 708–712 Chemie Ingenieur Technik 2 709 Communication Materials and Methods 2.1 Sample Preparation and Characterization All the composite samples were prepared using AA8006 alloy aluminum plates with 167 mm thicknesses. Spray composites were prepared using a slurry mixture consisting of 20 wt % solid and 80 wt % ethanol as a mineralizing agent. The solid content was a mixture of 70 wt % zeolite powder and 30 wt % commercial attapulgite (from Clariant AG) as binding material. For the spraying, an airbrush pistol (AFC101A) from Conrad Electronics SE was used as spraying gun and operated at 2 hPa air (as pressurizing fluids). The number of spraying was varied to achieve samples with different coating thicknesses. Reactive composites, Zeo-type SAPO-34 and Zeolite Y samples were prepared using the synthesis recipe of Bauer [5] and our previously reported work [15], respectively. X-ray diffraction (XRD) measurements using CuKa radiation on an X’Pert Pro diffractometer (Philips Analytical) in the 2q range of 2 – 80 and scanning electron microscope (SEM) analysis using a Carl Zeiss ULTRA 55 microscope at a voltage of 2 kV without applying pre-treatment to the samples were used to characterize the fabricated composites. The samples for microscopic analysis were prepared by fixing the samples in polymer resin consisting of EpoFIX resin with EpoFIX as hardner from Struers and the resulting images were used to determine the layer thickness. Therefore, cross-sectional SEM images taken using an Aztec 300 microscope at a voltage of 20 kv for reactive composites and images from LEXT OLS400 laser scanning microscope (LSM) from Olympus for spray composites were used, respectively. The solid packing density (rb) of the coating layer was determined from the water sorption measurements carried out in a dynamic vapor sorption analyzer (DVS), from, surface measurements at 40 C and relative humidity of 23.1: mass of water adsorbed by the composites water sorption capacity of a pure zeolite ðg g1 Þ rb ¼ thickness of the coating layer  geometrycial surface area (1) Mt ¼2 M¥ pffiffiffiffiffi Dt 1 pffiffiffi l p (2) where Mt is the amount of diffusing substance which has entered the material at time t, M¥ the total amount of diffusing substance which has entered the material at equilibrium, D the effective diffusion coefficient, l the coating layer thickness, and t the time. According to Eq. pffiffi(2), the normalized sorption capacity (Mt/M¥) against t =l was plotted. From the slope of the M curve at the region of 0:1 < t < 0:4, the effective diffuM¥ sion coefficient was determined. 2.3 Thermal Conductivity Measurement The laser flash method was used for thermal conductivity determination by the 2-layer method. As thermal conductivity cannot be measured directly, it was calculated using: lðT Þ ¼ aðT Þrb ðT Þcp ðT Þ (3) where a is thermal diffusivity, cp is heat capacity, and rb is the bulk density of the sample at a certain temperature. Thermal diffusivity was measured according to ASTM E1461 by using a Linseis LFA 1000 instrument. It determines the apparent thermal diffusivity of the layered samples based on Parker’s theory [17]. The data reduction was performed using Linesis analysis software with a 2-layer method to determine the thermal diffusivity of the coating layer, where the pre-measured data of the known support layer was used as input. The thermal diffusivity of the support aluminum plate was pre-measured by using different thicker aluminum plates and determined to be 0.79 cm2s–1. The thermal conductivity value of the coating layer was then calculated by Eq. (3), where the heat capacity of the coating material was pre-measured using a differential scanning calorimeter (DSC) from TA Instruments following ASTM E1269. The LFA-measurements were carried out at ambient conditions of 20 C and 1.01 hPa using the parameters listed in Tab. 1. 2.2 Diffusion Coefficient Measurement 3 Effective diffusion coefficients were determined from water sorption kinetic measurements by hanging the one-side coated samples in the sample chamber of the DVS instrument. The water sorption analyses were carried out at 40 C and relative humidity of 23.1 after pre-heating the samples at 250 C for 1 h under N2 flow. Then the effective vapor diffusion coefficient was determined using the water sorption data according to the Crank method [16] for the case of a fast sorption process in thin layers. The composites were characterized to determine the presence of respective materials in the coating layer. In Fig. 1(left), the XRD patterns of the fabricated samples confirmed that the samples a and b contain zeolite Y, and samples c and d contain zeotype SAPO-34 as coating material. From the cross-sectional images of the SAPO-34 samples, it is visible that the reactive coating method results in much denser coating layers compared to spray coating (Fig. 1 (right)). Chem. Ing. Tech. 2018, 90, No. 5, 708–712 Results and Discussion ª 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.cit-journal.com 710 Communication Chemie Ingenieur Technik Table 1. Sample properties used in the 2-layer method of thermal conductivity measurements. Solid packing density rb [g cm–3] Heat Capacity, Cp [J kg–1K–1] 65 0.76 1.31 (1.0:0.5:0.0:0.4) 28 1.31 0.88 SAPO-34 (1.0:4.5:3.6:0) 60 0.85 1.32 Reactive SAPO-34 (1.0:6.8:5.9:0.0) 60 1.55 0.89 S_Re1 Reactive SAPO-34 (1.0:6.4:5.1:0.0) 46 1.48 0.89 S_Re2 Reactive SAPO-34 (1.0: 8.2:7.4:0.0) 135 1.38 0.89 Sample name Coating technique Zeolite type Zeolite composition (atom %) (Si:Al:P:Na) Y_Sp Spray Zeolite Y Commercial* Y_Re Reactive Zeolite Y S_Sp Spray S_Re Layer thickness [mm] *CBV 100 from Zeolyst international (Si/Al = 2.55 with 13 wt % of Na2O). Figure 1. (Left) XRD patterns of the composite samples prepared by reactive coating: (a) Y_Re, & (c) S_Re, and spray coating: (b) Y_Sp & (d) S_Sp; (right) cross-sectional SEM images of (e) S_Re and (f) S-Sp [*indicates the reflexes of metallic aluminum according to JCPDS No. 89-4037] The powder material density was measured with a He pycnometer. The determined value of zeolite-Y and SAPO34 is 2.0 and 1.95, respectively. It means that the reactivecoated method leads to dense layers up to 65 – 80 % of the theoretical compactness. In Fig. 2a, the impact of the coat- ing packing density on water sorption kinetics is shown. Depending on the physical properties of the coating, the composites are classified in two cases. The SAPO-based composites (case 1) have similar coating thickness but different solid packing density. In this case, the reactive Figure 2. (a) water vapor sorption kinetics of composites produced with different techniques; and (b) diffusion coefficient analysis of ‘reactive’ (SAPO)-composites of various layer thickness based on Crank method [16]. www.cit-journal.com ª 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Ing. Tech. 2018, 90, No. 5, 708–712 Chemie Ingenieur Technik 711 Communication composites, which are around two times denser compared to spray samples, need around two times longer for reaching 80 % sorption capacity (Fig. 2a). Zeolite Y-based composites (case 2) have different coating thicknesses as well as solid packing densities. The Y_re composites have high solid packing density with thinner coating thickness compared to Y_Sp ones with lower packing density but higher coating thickness. In this case, both composites need a nearly similar amount of time to reach 80 % sorption capacity (Fig. 2a). From the obtained results, it is evident that the thickness and packing density of the coating layer influence the diffusion. In Fig 2b, a comparative vapor diffusion analysis of coating layers with different thicknesses (reactive SAPO composites with similar packing density) is shown. It is visible that up to 135 mm of layer thickness, the vapor diffusion coefficient increases with increasing layer thickness. The obtained values agree with the investigations from Tatlier and co-workers for zeolite A coatings [7]. In this range of layer thickness, the higher amount of adsorbents can compensate the reduction in the rate of adsorption. The effective diffusion coefficient is evaluated using the water sorption data according to the crank method, and the values are listed in Tab. 2. The effective diffusion coefficient values indicate that spray-coated samples have three times better molecular transfer compared to reactive-coated ones. The open nature of the coating layer from the spray coating is the reason for it. For reactive coating, dense and thicker layers result in slower vapor diffusion into the layer, which leads to a slower sorption process. The measured coating thermal conductivity values show different patterns from the results of diffusion coefficients (Tab. 2). The reactive-coated dense coatings have higher thermal conductivity compared to spray-coated ones. The measured value is around four times higher compared to spray-coated samples. This behavior is due to better contact between zeolite and support as well as the dense nature of the coating layer. The measured value shows an agreement with the reported effective thermal conductivity of zeolites as pelletized form, where an increase of effective thermal conductivity is obtained by increasing the packing density [8, 18, 19]. However, the two types of zeolites prepared by the same coating method show slight differences in their thermal conductivity values. It confirms that the higher thermal conductivity is originating from the coating technique and, thus, the resulting coating physical properties (nature of inter-connectivity, packing density) instead of the zeolite material structure. 4 Conclusion In summary, laser flash analysis and sorption kinetic measurement methods are applied to compare the heat and mass transfer properties of coated systems. From the determined values, it is evident that high solid packing density results in higher thermal conductivity of the coated layer. Based on the findings, coating techniques such as reactive coating are preferred for generating an adsorption heat pump with high power density. In case of mass transfer, although the high solid packing density of the coating layer decreases the diffusion coefficient, an optimum thickness of the coating layer will ensure maximum performance of the adsorption heat pump. Interestingly, the measured properties indicate that the adopted experimental analysis methods are applicable for materials with different chemical composition and porosities. Therefore, it opens the window for applying the proposed experimental methods to characterize the contributions of the coated layer to the overall performance of composites for other crystalline as well as non-crystalline materials. Symbols used [J kg–1K–1] [m] [s] [cm2s–1] cp l t D heat capacity layer thickness time diffusion coefficient Table 2. Measured diffusion coefficient and thermal conductivity of the coating layer. Sample name Y_Sp Y_Re S_Sp S_Re S_Re1 S_Re2 Layer thickness [mm] 65 28 60 60 46 135 Chem. Ing. Tech. 2018, 90, No. 5, 708–712 Solid packing density rb [g cm–3] Effective diffusion coefficient Deff [cm2s–1] (±5 %) Effective thermal conductivity lcoat [W m–1K–1] (±1 %) 0.76 3.97  10–8 0.08 1.31 1.17  10 –8 0.42 3.52  10 –8 0.12 1.24  10 –8 0.39 1.17  10 –8 0.37 –8 0.41 0.85 1.55 1.48 1.38 2.31  10 ª 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.cit-journal.com 712 Communication References Greek symbols a rb lc [cm2s–1] [g cm–3] [W m–1K–1] thermal diffusivity packing density thermal conductivity Subscripts ¥ eff sp re equilibrium effective spray-coated reactive-coated Abbreviations PST XRD SEM DSC DVS LFA Y S partial support transformation x-ray diffractions scanning electron microscope differential scanning calorimeter dynamic vapor sorption laser flash analysis zeolite Y zeolite SAPO-34 www.cit-journal.com Chemie Ingenieur Technik [1] H. Demir, M. Mobedi, S. Ülkü, Renewable Sustainable Energy Rev. 2008, 12, 2381–2403. [2] B. Choudhury, B. B. Saha, P. K. Chatterjee, J. P. Sarkar, Appl. Energy 2013, 104, 554–567. [3] U. Wittstadt, G. Füldner, O. Andersen, R. Herrmann, F. Schmidt, Energies 2015, 8, 8431–8446. [4] A. Freni, A. Frazzica, B. Dawoud, S. Chmielewski, L. Calabrese, L. Bonaccorsi, Appl. Therm. Eng. 2013, 50, 1658–1663. [5] J. Bauer, R. Herrmann, W. Mittelbach, W. Schwieger, Int. J. Energy Res. 2009, 33, 1233–1249. [6] L. Wang, D. Zhu, Y. Tan, Adsorption 1999, 5, 279–286. [7] M. Tatlier, G. Munz, G. Fueldner, S. K. Henninger, Microporous Mesoporous Mater. 2014, 193, 115–121. [8] M. J. Prakash, M. Prasad, K. Srinivasan, Carbon N. Y. 2000, 38, 907–913. [9] V. Van Speybroeck, K. Hemelsoet, L. Joos, M. Waroquier, R. G. Bell, C. R. A. Catlow, Chem. Soc. Rev. 2015, 44, 7044–7111. [10] E. T. C. Vogt, B. M. Weckhuysen, Chem. Soc. Rev. 2015, 44, 7342–7370. [11] J. Zhong, J. Han, Y. Wei, P. Tian, X. Guo, C. Song, Z. Liu, Catal. Sci. Technol. 2017, 7, 4905–4923. [12] E. W. Ping, R. Zhou, H. H. Funke, J. L. Falconer, R. D. Noble, J. Memb. Sci. 2012, 415–416, 770–775. [13] S. K. Henninger, F. P. Schmidt, H. M. Henning, Appl. Therm. Eng. 2010, 30, 1692–1702. [14] Iza-Structure.org. (Accessed on 01/08/2018) [15] R. Chanda, T. Selvam, R. Herrmann, W. Schwieger, Mater. Lett. 2018, 211, 103–106. [16] J. Crank, The Mathematics of Diffusion, 2nd ed., Oxford University Press, Oxford 1975. [17] W. J. Parker, R. J. Jenkins, C. P. Butler, G. L. Abbott, J. Appl. Phys. 1961, 32, 1679–1684. [18] Q. Zhang, H. Wang, Z. Ling, X. Fang, Z. Zhang, Sol. Energy Mater. Sol. Cells 2015, 140, 158–166. [19] L. Borchardt, N. L. Michels, T. Nowak, S. Mitchell, J. Pérez-Ramı́rez, Microporous Mesoporous Mater. 2015, 208, 196–202. ª 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Ing. Tech. 2018, 90, No. 5, 708–712