- Email: [email protected]
Ethanol adsorption onto carbonaceous and composite adsorbents for adsorptive cooling system
Energy 84 (2015) 177e185
Contents lists available at ScienceDirect
Energy journal homepage: www.elsevier.com/locate/energy
Ethanol adsorption onto carbonaceous and composite adsorbents for adsorptive cooling system V. Brancato a, *, A. Frazzica a, A. Sapienza a, L. Gordeeva b, c, A. Freni a a Consiglio Nazionale delle Ricerche (CNR), Istituto di Tecnologie Avanzate per l'Energia “Nicola Giordano” (ITAE), S. Lucia Sopra Contesse 5, 98126 Messina, Italy b Boreskov Institute of Catalysis, Pr. Lavrentieva. 5, Novosibirsk, Russia c Novosibirsk State University, Pirogovastr. 2, Novosibirsk, Russia
a r t i c l e i n f o
a b s t r a c t
Article history: Received 10 November 2014 Received in revised form 18 February 2015 Accepted 23 February 2015 Available online 21 March 2015
The aim of the present paper is the experimental characterization of adsorbent materials suitable for practical applications in adsorption refrigeration systems, employing ethanol as refrigerant. Different commercial activated carbons as well as a properly synthesized porous composite, composed of LiBr inside a silica gel host matrix, have been tested. A complete thermo-physical characterization, comprising nitrogen physi-sorption, specific heat and thermo-gravimetric equilibrium curves of ethanol adsorption over the sorbents, has been carried out. The equilibrium data have been fitted by means of the Dubinin e Astakhov equation. On the basis of the experimental data, a thermodynamic evaluation of the achievable performance of each adsorbent pair has been estimated by calculating the maximum COP (Coefficient of Performance) under typical working boundary conditions for refrigeration and air conditioning applications. The innovative composite material shows the highest thermodynamic performances of 0.64e0.72 for both tested working conditions. Nevertheless, the best carbonaceous material reaches COP value comparable with the synthesized composite. The results have demonstrated the potential of the chosen adsorbents for utilization in adsorption cooling systems. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Activated carbons Adsorption refrigeration Ethanol Porous composite adsorbent
1. Introduction With the rapid development of economy and society, our modern life and industry consume large amounts of energy for cooling. Furthermore, the depletion of fossil fuels and the threat of global warming over the last decades have challenged air conditioning industry to develop new cooling technologies alternative to the conventional vapor compression systems. Heat driven adsorption cooling cycles have the great energy saving potential due to utilization of the solar energy and low grade waste heat, e.g. engine exhaust, industrial waste heat, etc. [1,2]. The interest in adsorption cooling systems first is started to the oil crisis in the 1970s, and then later, in the 1990s, because of ecological problems related to the use of CFCs and HCFCs as refrigerants. Accordingly, systems that can recover waste heat at low
* Corresponding author. Tel.: þ39 090 624 243; fax: þ39 090 624 247. E-mail address: [email protected] (V. Brancato). http://dx.doi.org/10.1016/j.energy.2015.02.077 0360-5442/© 2015 Elsevier Ltd. All rights reserved.
temperature levels e such as adsorption systems e can be an interesting alternative to common vapor compression ones, for a wiser energy management [3]. Sorption refrigeration technologies are thermally driven systems, in which the conventional mechanical compressor of the common vapor compression cycle is replaced by a ‘thermal compressor’ [4]. This technology could be suitable, for instance, for fishing boat using waste heat discharged from the internal combustion engine. A sorption refrigeration system is noise-free, easy controllable, environmental friendly and virtually maintenance-free [3]. However the main drawback of these systems, which hinder their practical dissemination, is their reduced performance in terms of specific cooling capacity and COP, as well as cooling power [5,6]. The adsorption characteristics of the adsorbent/refrigerant pair are one of the most essential parameters that affect the system performance [7e9]. An adsorbent must have the ability to adsorb large quantities of an adsorbate in a narrow range of temperatures and to desorb it easily when temperature rises. Its properties
178
V. Brancato et al. / Energy 84 (2015) 177e185
should not change with age and use (i.e. high hydrothermal stability). Water, ammonia, methanol and ethanol are typical refrigerants of the adsorption refrigerator. While silica gel, zeolites and activated carbons are the most common adsorbents used in these systems. Among the common refrigerants, ethanol is one of the most interesting due to its low freezing point (114 C) and to its ecological compatibility. Respect to methanol, ethanol has a similar saturation pressure, but its latent heat is about 30% lower than methanol. Nevertheless, methanol is characterized by high toxicity and inflammability compared to ethanol and for this reason special care must be used during the phases of storage, handling and use. Methanol also shows the problem of dissociation above 120 C, especially in the presence of copper. The typical ethanol adsorbents used in adsorption applications are activated carbons, in fact these latter have a large surface area and exhibit good affinity towards ethanol. A good number of papers related to ethanol adsorption refrigeration have already been published. The activated carbons are common materials proposed as ethanol adsorbents for adsorption cooling [10]. Accordingly, other authors further investigated the features of this activated carbon as discussed below. El-Sharkawy et al. [11] measured the adsorption isotherms Maxsorb III/ethanol pair. They found that this pair is attractive for solar cooling application. Indeed, a cooling cycle realized with this adsorption pair can achieve a specific cooling effect as high as 420 kJ/kg at evaporator temperature of 7 C in combination with heat source and heat sink of temperatures 80 and 30 C, respectively. Maxsorb III/ethanol as adsorption pair was investigated also by Uddin et al. [12]. They studied the characteristics of ethanol adsorption onto “parent” Maxsorb III and surface treated Maxsorb III with controlled oxygen content. Experimental results show that adsorption capacity improves in H2 treated Maxsorb III, the KOHeH2 treated Maxsorb III highlights faster adsorption kinetics but the isosteric heat of adsorption remains higher in the parent Maxsorb III/ethanol pair. Loh et al. [13] have presented the performance analysis of both ideal single-stage and single-effect double-lift adsorption cooling cycles working at partially evacuated and pressurized conditions of six specimens of adsorbents and refrigerant pairs. They found that the activated carbon fibers (AC/20)/ethanol pair provides the highest value of specific cooling effect among the tested pairs. El-Sharkawy et al. [14] have also investigated the activated carbon fibers/ethanol pairs for adsorption cooling system. They have studied the adsorption capacity and adsorption rate of the ACF A-15 and A-20. They found that the equilibrium adsorption capacity of A-20/ethanol pair is considerably larger than that of A-15/ ethanol pair and that the ACF A-20 seems to be a promising adsorbent as it has large ethanol adsorption capacity related per gram of the adsorbent. However the low packing density of ACF reduces their capacity related per the unit volume. The same authors [15] have also analyzed the adsorption characteristics of ethanol onto two promising adsorbents based on spherical phenol resin treated with different mass ratios of KOH named as KOH4-PR and KOH6-PR. They found that the adsorption uptake of ethanol onto the sorbents are 1.43 and 2 kg/kg, respectively. This result shows that the newly developed adsorbents have promising adsorption characteristics with ethanol that may lead to the development of next generation of adsorption chillers, in particular for air conditioning cycles. Under conditions of ice making cycle, means at evaporation temperature 3 C the ethanol uptake on KOH4-PR and KOH6-PR is much lower 0.3e0.4 g/g. Recently, Rezk et al. and [16] Saha et al. [17] have been presented the experimental and theoretical investigations of adsorption characteristics of
ethanol onto metal organic framework namely MIL-101Cr. The adsorption capacity of MIL-101Cr reaches as high as 1.1 g/g at ethanol relative pressure of 1, while under conditions of the typical air conditioning cycle (evaporation and adsorption temperatures of 10 and 30 C, respectively) it is 0.9 g/g. At ice making cycle (evaporator temperature of 3 C) the uptake falls down but remains quite high e 0.45 g/g. The results demonstrate that the MIL-101Cr e ethanol is promising working pair, however more studies on the durability, thermal properties (namely the heat capacity, thermal conductivity) etc. are required for evaluation of its practical potential for the adsorption cooling applications. Recently, a new adsorbent, namely composites sorbent salt/silica gel, have been proposed for ethanol sorption cooling. Gordeeva et al. [18] found that the composite LiBr/SiO2 is the promising composite ethanol adsorbent due to the large variation in the ethanol uptake for both air conditioning and ice making cycles. Thus, the data available in the literature shows that various carbonaceous materials, MIL-101Cr and composite LiBr/SiO2 are the most promising ethanol adsorbents in the adsorption cooling. However, the major part of the studies are related to the conditions of air conditioning cycle. The materials having the greatest potential in terms of ethanol uptake, namely KOH4-PR, KOH6-PR, and MIL-101Cr are not commercially available, and consequently can hardly be considered for practical application in the adsorption chillers. As regards the traditional ethanol adsorbents, activated carbons, it was shown by Li et al. [19] that an activated carbon e ethanol working pair cannot allow ice production. Therefore, an empty room still exists for the search of ethanol adsorbents suitable for practical application in the adsorption chillers. The goal of the present paper is the investigation of the adsorption characteristics of ethanol on some on-purposely selected commercially available activated carbons, produced from different bases (coal, coconut shell) and having different morphology (grains, pellets, fibers) with the aim to find the most suitable material for a sorption cooling system, in particular, for an ice making cycle. Additionally, the previously mentioned novel composite LiBr/SiO2 was investigated and compared with the selected activated carbons. The ethanol sorption isobars of the selected samples were valued by a thermo-gravimetric system, specifically realized for alcohols adsorption measurement under real adsorption cooling conditions. The measured adsorption data were correlated with the nitrogen physi-sorption results and they were analyzed according to the DubinineAstakhov theory in order to evaluate the equilibrium adsorption coefficients and sorption heat. Calorimetric measurements were carried out for all the samples to evaluate the specific heat. The evaluated experimental data were used to carry out thermodynamic analysis with the aim to estimate the performance of a refrigeration and air conditioning cycle in terms of cooling COP. 2. Experimental 2.1. Selected materials Five different commercial activated carbons, supplied by different companies, have been characterized and evaluated for the use in adsorption refrigeration systems. The samples are chosen with different bases, ranging from coconut shell to coal based, and with different shapes, from grains to fibers. Table 1 reports the list of the tested samples. In addition, the innovative composite sorbent LiBr/silica gel (SG/ LiBr) has been studied. The silica gel KSK (“Salavatnefteorgsyntes”, Russia) with the specific surface area Ssp ¼ 280 m2/g, the pore volume Vp ¼ 1.0 cm3/g, the average pore size dav ¼ 15 nm was used as a porous matrix. The composite sorbent SG/LiBr was prepared
V. Brancato et al. / Energy 84 (2015) 177e185 Table 1 Tested samples.
179
Table 2 Specifications of load cell and of pressure gauges.
Sample
Manufacture
Product form
Based
Particle size (mm)
SRD 1352/3 FR20 AP4 e 60 ATO COC eL1200 SG/LiBr
Chemviron Kuraray Chemviron Eurocarb Carbon Activated e
Grains Fiber Pellets Grains Grains Grains
Coconut shell e Coal Coconut shell Coconut shell Silica gel/LiBr
0.5e2 F 0.01 >4 0.25e0.6 0.42e1 0.7e0.85
following the preparation reported in Ref. [18]. The LiBr content (CLiBr) in the anhydrous composite was 19 wt.%. Anhydrous ethanol, with a purity of 99.9%, has been used as refrigerant. 2.2. Experiments The evaluation of the different tested materials is based on detailed data of thermo-physical properties including surface and pore characteristics, adsorption equilibrium data and specific heat. The texture properties of the activated carbons and of the porous composite were characterized by a Micromeritics ASAP 2020, using N2 as the adsorbate at 77 K. Prior to the measurements, the samples have been dried at 160 C under continuous evacuation for 3 h. Adsorption and desorption isotherms were performed in order to detect possible hysteresis effects. The BET (BrunauereEmmetteTeller) theory was applied to determine the specific surface area of mesoporous sample (SG/LiBr) while Langmuir theory was selected to evaluate the specific surface area of activated carbons. HorvatheKawazoe [20] method was applied to evaluate the cumulative pore volume and the median pore width in the microporous samples, while cumulative volume of pores of the composite material was evaluated from BJH (BarretteJoynereHalenda) desorption theory. In this last case the pore width was calculated as 4V/ABET [21]. Ethanol adsorption measurements were performed by a thermo-gravimetric system specifically designed to this scope. Fig. 1 reports both the system layout and the realized apparatus. The core component of the experimental apparatus is the load cell (see Table 2 for specifications), that measures the variations of the sample mass during the adsorption process. In accordance to Fig. 1a, the testing chamber (2) is submersed into a thermo cryostat (1). An evaporator made by glass (4), connected to a thermo cryostat (5), two pressure gauges (P1 and P2 e see Table 2 for specifications), three valves (V1, V2, V3) and a vacuum pump (3) are the other components of the plant. All physical parameters (pressure, temperature) were acquired every 60s through a display system
Load cell mod. PW4C3-300g supplied by HBM
Maximum capacity (Emax) Minimum LC verification interval (vmin) Sensitivity (Cn) Zero signal Pressure gauge mod. VSC43MA4 Measuring Range supplied by Thyracont Accuracy
g g
300 0.05
mV/V 1.0 ± 0.1 0 ± 0.1 mbar 1-1400 e 0.3% f.s.
and data acquisition by means of a software interface specifically developed in LabView. For each test 7 e 8 g of adsorbent material were used. Prior to all the measurements the samples were dried in an oven at 150 C for at least 12 h to determine their dry mass (reference mass). After drying, the sample was put in a nickel crucible in the testing chamber and then evacuated by a vacuum pump for at least 6 h at 120 C. Then the valve between the test section and the evaporator was opened and adsorption started. The measurements were conducted under isobaric conditions at four different ethanol pressures: 12.6 mbar (3 C), 18.0 mbar (þ2 C), 25.5 mbar (þ7 C) and 35.4 mbar (þ12 C). Seven different equilibrium points were measured between 30 C and 120 C. The specific heat of the samples has been measured by a differential scanning calorimeter, DSC mod. 27 HP (Mettler Toledo). Approximatively, 10e15 mg of sample were dried for 12 h at 150 C to remove moisture before the experiment and then inserted into a sealed crucible, to prevent adsorption of moisture from the air. Subsequently, the measurements were carried out from room temperature up to 160 C, with a heating rate of 2 C/min in air. 3. Results and discussion 3.1. Physi-sorption nitrogen measurements Fig. 2 shows the N2 adsorption/desorption isotherms for activated carbons (Fig. 2a) and for the porous composite material SG/ LiBr (Fig. 2b). Only one isotherm is represented for activated carbons AP-60 and COC L-1200, because these two sorbents showed overlapped curves due to the same texture characteristics. All the tested activated carbons exhibit adsorption isotherm of type I according to IUPAC [22] typical of microporous solids. The very steep region at low p/ps is due to the volume filling of very narrow pores and limiting uptake is dependent on the accessible micropore volume rather than on the internal surface area [23]. It is observed that isotherm curves of carbon SRD1352/3, ATO and COC L-1200 show a little hysteresis at p/ps > 0.45, which may be caused
Fig. 1. a) schema of thermo-gravimetric plant; b) photo of thermo-gravimetric plant.
180
V. Brancato et al. / Energy 84 (2015) 177e185
Fig. 2. Adsorption/desorption isotherm of nitrogen on tested samples: a) activated carbons; b) composite (SG/LiBr).
by capillary condensation, due to the presence of some mesopores in these carbons. On the contrary, the isotherm of the activated carbon fibers (FR 20) is completely reversible in the whole range of p/ps indicating its microporous structure. As illustrated in Fig. 2b, porous composite SG/LiBr exhibits adsorption isotherm of type IV with a strong hysteresis at p/ps ¼ 0.8e0.9 that is typical of mesoporous solids. The calculated surface areas of the activate carbon are in the range from 1412 to 2613 m2 g1 (see Table 3) and it does not depend on the precursor of the carbon. Instead, the evaluated area of the porous composite SG/LiBr is 181 m2 g1. As before demonstrated by Gordeeva [18] no pore blocking of the silica gel by LiBr occurs during the composite preparation. In the activated carbons a large surface area is preferable for ensuring large adsorption capacity. However, large surface area in a limited volume inevitably gives rise to large number of small size pores between adsorption surfaces. The size of pore determines the accessibility of adsorbate molecules to the adsorption surface, so the pore size width is one of the most important properties for characterizing adsorptive of this adsorbent [23]. To value the pore size a series of method based on Kelvin equation have been proposed, but this theory becomes progressively less accurate as the pore size decreases and breaks down for micropores. In the present work, to value the cumulative pore volume and the median pore width in activated carbons the HorvatheKawazoe [20] method was applied, which is a semi-empirical, analytic model of adsorption in micropores that is commonly used for determining the pore size distribution of microporous materials. Instead, the cumulative pore volume and the pore width of the composite material was evaluated from BJH [21] desorption theory because this material is mesoporous. Table 3 shows a comparison of the found values for each tested sample. Among the activated carbon, the fibers FR 20 exhibit the largest pore volume (0.75 cm3/g), while the granular AP4-60 the smallest (0.47 cm3/g) with a pore width of 0.59 nm and 0.64 nm, respectively. A value of 0.73 cm3/g and pore width of 17.2 nm were found for the mesoporous composite sample.
Table 3 Thermophysical properties of tested samples determined by nitrogen adsorption. Sample
Surface area (m2/g)
Pore size (nm)
Cumulative pore volume (cm3/g)
SRD 1352/3 FR20 AP4 e 60 ATO COC e L1200 SG/LiBr
2613 2180 1428 1745 1412 181
0.56 0.59 0.64 0.59 0.59 17.2
0.65 0.75 0.47 0.64 0.49 0.73
Thus, all the carbonaceous materials show high surface area and an essentially microporous structure that promotes a high ethanol adsorption. 3.2. Ethanol adsorption measurements Ethanol adsorption measurements have been performed under isobaric conditions. In particular, four different isobars were measured at different pressure from 12.6 mbar (3 C) to 35.4 mbar (þ12 C). Some measures have been done in adsorption/desorption mode to verify the presence of hysteresis effects. Fig. 3 shows the adsorption e desorption isobars for all the tested samples. All the materials exhibit the presence of hysteresis effect between adsorption and desorption branches even if in different magnitude. The trend of the uptake depends on the nature of the adsorbent. Activated carbon shows a monotonic increase of the uptake with the decrease of temperature. Differently, the porous material SG/LiBr exhibits a critical temperature (e.g. 60 C at 12.6 mbar) under which the ethanol adsorption becomes significant. Gordeeva et al. [24] found that the sorption rise is caused by the reaction between the salt and the ethanol and the sorption ability of the composite depends mainly on the nature of the confined salt, while the matrix acts as a media that disperses the salt and provides efficient heat and mass transfer. 3.3. Dubinin e Astakhov transformation The measured data were transformed into a characteristic curve following the Dubinin formalism that describes the volume filling of micropores [25]. A number of empirical and semi-empirical isotherms like traditional Toth, Frendlich, Sips equations, as well as a new one for multi-types adsorption, which includes the loading, the adsorbent/adsorbate interaction and the surface structural heterogeneity factors has been developed [26,27]. Nevertheless, the DubinineRadusgkevich and Dubinin e Astakhov approaches are still the most employed model, especially to analyze the adsorption equilibrium of microporous carbonaceous materials for which the adsorption mechanism is the volume filling of micropores [28,29]. Therefore, the ethanol adsorption data were fitted to the Dubinin e Astakhov equation [30].
n A A n ¼ W0 exp W ¼ W0 exp E bE0
(1)
in which: W (cm3/g) defines the adsorbed volume for the given working pair; W0 (cm3/g) is the total accessible pore volume, which corresponds to a maximum loading; E (J/mol) is the characteristic energy for the working pair and depends on the sorbent as well as
V. Brancato et al. / Energy 84 (2015) 177e185
181
Fig. 3. Adsorption and desorption isobars: a) porous composite (SG/LiBr); b) activated carbon SRD 1352/3; c) fiber of activated carbon FR20; d) activated carbon AP4 e 60; e) activated carbon ATO; f) activated carbon COC-L1200.
on the adsorptive; b is the coefficient of affinity and depends only on the adsorbent; E0 (J/mol) indicates the characteristic adsorption energy depending only on the adsorbent. The exponent n is empirically determined. Its value depends on the pore structures. High values indicate a material with a homogenous pore structure, while low values indicate a heterogeneous structure. A (J/mol) is the differential adsorption potential, defined as follows:
A ¼ RTln
ps p
(2)
The adsorption potential A represents the difference of the chemical potentials of the adsorbent and the free liquid. Moreover, it is the differential change of free energy in a reversible, isothermal transition from liquid phase to the adsorbate phase. The reference pressure ps is the saturated vapor pressure at T.
The data transformation from gravimetric units (gEtOH/gAds), as reported in Fig. 3, into volumetric unit (cm3EtOH/gAds) has been performed considering ethanol density as a function of temperature, by means of the equation suggested by Ref. [31]. The differential heat of adsorption can be calculated from the adsorption equilibrium. Since the adsorption equilibrium was derived from the characteristic curve, the differential heat of adsorption results from the following equation [32], where the non-ideality of gaseous phase effects is not considered:
DH ¼ L þ A TaW
vDA vW
(3) T
where L (J/mol) is the heat of evaporation (evaluated considering the medium temperature of the working cycle), a is the coefficient
182
V. Brancato et al. / Energy 84 (2015) 177e185
of thermal expansion of the adsorbent and W is the adsorbed volume. The isobaric data were then transformed according to Dubinin e Astakhov approach as illustrated in Fig. 4. As can be seen from the diagram, the transformed data generate a very smooth curve. Results of the characterization including all parameters for DA equation are provided in Table 4. The goodness of the fit of the activated carbons has been evaluated considering the plot of ln(W) vs (A)n at the optimal value of n. Table 4 also reports the values of R2 estimated from the plot. All values are close to 1, so this confirms that DA transformation properly describes the adsorption properties of the investigated working pairs. In addition, the physi-sorption measurements have also confirmed the goodness of the fit because the calculated values of W0 from nitrogen adsorption are in close agreement with those of ethanol sorption.
Table 4 DA parameters for the tested samples. Sample
SG/LiBr SRD 1352/3 FR 20 AP4 e 60 ATO COC e L1200
Ethanol adsorption
Nitrogen adsorption
W0 (cm3/g) E (kJ/mol) n
R2
W0 (cm3/g) E (kJ/mol) n
0.68 0.82 0.75 0.45 0.61 0.44
e 0.988 0.990 0.988 0.978 0.994
e 0.78 0.74 0.44 0.63 0.52
6.9 8.78 13.5 10.6 11.2 13.3
1.8 1.5 2 2 1.7 2
e 18.9 20.3 18.9 20.9 23.9
e 1.7 1.6 1.5 1.4 1.4
Composites “salt inside porous matrix” are mesoporous solids and adsorption of water and alcohols includes different physicalchemical process, namely the adsorption on the active surface centers, the formation of the crystalline solvates “salt e sorbate”
Fig. 4. DA transformation of the equilibrium values for: a) composite (SG/LiBr); b) activated carbon SRD 1352/3; c) fiber of activated carbon FR 20; d) activated carbon AP4 e 60; e) activated carbon ATO; f) activated carbon COC - L1200.
V. Brancato et al. / Energy 84 (2015) 177e185
and, finally, the formation of the solution “ salt e sorbates” inside pores [33]. However, it was shown earlier [34] that, despite mesoporous structure of the composites “salt inside porous matrix” and complex sorption mechanism, the water sorption on these materials can be formally described by DA equation as well. Therefore, for complete discussion of the results, DA equation has been applied also to ethanol adsorption on the composite material SG/ LiBr. As expected, the DA parameter W0 ¼ 0.68 cm3/g calculated from ethanol adsorption for SG/LiBr presents a good correspondence with pore volume (0.73 cm3/g) calculated from nitrogen adsorption. The DA equation is not applicable for nitrogen adsorption on the composite because in this case the adsorption occurs only on the surface and the capillary condensation in mesopores is also present. For calculation of DA parameters of all activated carbons, the value of affinity coefficient b was 0.61 as reported by Lopez-Ramon [35] and by Wood [36]. The maximum loading capacity W0 follows the increasing surface area, starting from the carbon COC L-1200 to the carbon SRD 1352/3. In fact, COC L-1200 exhibits a surface area of 1412 m2/g and a W0 of 0.44 cm3/g, while the SRD 1352/3 shows the highest surface area (2613 m2/g) and the highest adsorbed volume (0.82 cm3/g). However, all samples reveal a poor adsorption capacity for an adsorption potential higher than 450 J/g. The porous composite SG/LiBr shows a smaller sorption at A > 250 J/g and a higher slope in the range 0e250 J/g compared to the curves of all activated carbons. This behavior is attributed to the different mechanism of the ethanol sorption by the composite. 3.4. Specific heat measurements The evaluation of specific heat of the dry adsorbents has been carried out by a differential scanning calorimeter DSC. The measured specific heats for all the samples are within the same range from approximately 0.80 J/g K at 40 C to 1.10 J/g K at 100 C. Table 5 shows the measured values for each tested sample. The activated carbon with coal as precursor (AP4-60) exhibits the lowest specific heat, while the carbon SRD 1352/3, produced from coconut shell, the highest. In addition, the found value for the porous composite SG/LiBr is in close agreement with the value estimated by Gordeeva et al. [24]. 3.5. Estimation of the cooling performance The obtained data allow an estimation of the coefficient of performance COP for both refrigeration and air conditioning cycles for each working pairs, considering the following characteristic temperature: 90 C as maximum desorption temperature (Tdes) Table 5 Experimental specific heat for tested samples. Sample
T ( C)
Cp (J/g∙K)
Sample
T ( C)
Cp (J/g∙K)
ATO
40 60 80 100 40 60 80 100 40 60 80 100
0.94 0.97 1.04 1.10 1.05 1.07 1.14 1.21 0.73 0.77 0.82 0.86
COC e L1200
40 60 80 100 40 60 80 100 40 60 80 100
0.85 0.88 0.96 1.03 0.80 0.82 0.90 0.95 0.81 0.82 0.84 0.90
SRD 1352/3
AP4 e 60
FR 20
SG/LiBr
183
30 C as minimum adsorption and condensation temperature (Tad-cond); 2 C as evaporation temperature (Tev) for refrigeration cycle and þ7 C as evaporation temperature (Tev) for air conditioning cycle. COP has been calculated as:
COP ¼
Q ev ðQ sens þ Q lat Þ
¼
L$Dw cp eq ðwmax Þ$ðT2 Tads Þ þ cp eq ðwav Þ$ðTdes T2 Þ þ DH$w (4)
where Qev (J/g), represents the useful cooling effect due to the ethanol evaporation, Qsens (J/g), the sensible energy delivered to heat up the adsorbent and Qlat (J/g), the latent heat needed to desorb ethanol from the adsorbent material. In the calculation, the heating up of the ethanol inside the evaporator, due to the flowing of the liquid from the condenser, has been omitted, since its effect can be considered negligible. In particular, Dw (gEtOH/gAds) is the uptake variation between adsorption and desorption; DH (J/g) is the desorption heat; L (J/g) is the latent heat of ethanol, evaluated at the evaporation temperature; cpeq(wmax) (J/g$K), is the equivalent specific heat evaluated during the isosteric heating phase; cpeq(wav) (J/g$K) is the equivalent specific heat evaluated during the desorption phase and finally T2 is the temperature of the end of isosteric heating phase. According to the following equation (5), the equivalent specific heat of the adsorbents saturated with ethanol up to uptake w has been evaluated as linear combination of the specific heat of the dry adsorbent and that of the adsorbed ethanol in the liquid state, as suggested by Ref. [37].
cp eq ¼
cp adsorbent þ cp EtOH $w 1þw
(5)
The equation (5) implies that the specific heat of adsorbed ethanol is considered equal to that for liquid ethanol. The estimated values of variation in ethanol uptake and COP are reported in Table 6. The ethanol uptake variation has been evaluated as difference between the maximum and the minimum uptake calculated starting from the equilibrium values of Dubinin's fit for the considered cycle. Starting from the metrological data of the employed sensor (see Table 2), according to the theory of the uncertainty of measurements, the typical experimental error was calculated by the method of the extended uncertainty where the composed error contribute of each measurement device as well as the effect of the desired confidence level are considered, as reported in Ref. [38]. An uncertainty around 4% has been calculated for COP evaluation. For the refrigeration cycle, the lowest COP value, 0.39, was found for activated carbon COC L-1200. This carbon presents also the
Table 6 COP for tested samples. Sample
SG/LiBr SRD 1352/3 FR 20 AP4 e 60 ATO COC e L1200
Refrigeration cycle
Air conditioning cycle
Dw (gEtOH/gAds)
COP
Dw (gEtOH/gAds)
COP
0.162 0.151 0.103 0.086 0.103 0.062
0.64 0.55 0.47 0.49 0.48 0.39
0.260 0.235 0.143 0.122 0.152 0.086
0.72 0.63 0.53 0.58 0.55 0.45
184
V. Brancato et al. / Energy 84 (2015) 177e185
lowest variation in ethanol uptake, 0.062 g/g. Instead the carbon SRD 1352/3 reveals the highest variation in ethanol uptake (0.151 g/ g) and the highest COP (0.55) among the activated carbons. These results are in close agreement with the others above discussed, indeed the activated carbon SRD 1352/3 has the highest specific surface area and the highest maximum loading capacity evaluated by DA transformation, vice versa the COC L-1200 shows the worst features. The other carbons show an intermediate behavior between COC L-1200 and SRD 1352/3. The activated carbon based on coal (AP4-60) exhibits a low variation in ethanol uptake (0.086 g/g), but it has also the lowest specific heat so the resulting COP is satisfying (0.49). The same considerations can be applied for the obtained results for the air conditioning cycle. All found values are not far from those found by El-Sharkawy [11,39] and by Saha [40]. In both cases, the highest COP value has been obtained for the composite SG/LiBr. Moreover, COP and ethanol uptake variation have been also calculated for a refrigeration cycle with a Tad-cond equal to 25 C, because this is a realistic operative condition for some applications, such as in the fishing boat refrigeration, abovementioned, that employs sea water for condensation. Interestingly, the decreasing of only 5 C of Tad-cond improves the COP value and the Dw that become comparable or even higher than the values found for the air conditioning cycle. The SG/LiBr was again the best adsorbent showing a COP of 0.69 and a Dw of 0.243 g/g. Among activated carbons the SRD 1352/3 exhibited again the highest performances (COP ¼ 0.62 and Dw ¼ 0.231 g/g) while the COC L-1200 the worst behavior (COP ¼ 0.46 and Dw ¼ 0.094 g/g). Fig. 5 shows the COP vs the regeneration temperature of the samples SG/LiBr and SRD 1352/3, reported as instance of the studied materials, for a refrigeration cycle considering the characteristic temperatures above mentioned (Tad-cond ¼ 30 C). The diagram highlights that high regeneration temperatures are preferred for activated carbon SRD 1352/3 while the regeneration temperature has a lower influence on the performance of the composite material SG/LiBr. Indeed, this last shows a slower decrease of the COP with the decrease of the regeneration temperature. On the contrary, regeneration temperatures lower 80 C negatively affect the achievable performance of both samples (activated carbon and composite material). The maximum theoretical COP ¼ 0.3e0.5, reported in Ref. [41] for “activated carbons e methanol” and “ACFs e methanol” pairs at regeneration temperature above 100 C is significantly lower than those estimated for SRD 1352/3 e ethanol and SG/LiBr e ethanol pairs. Some adsorption machines based on zeolites e
methanol, activated carbons e methanol, activated carbons e ammonia, zeolites e water working pairs were realized in actual practice and operated with real COP varying from 0.12 to 0.33 [42,10]. The modeling of a commercially available two-bed silica gel/water adsorption chiller with “MIL-101Cr e ethanol” working pair demonstrated that COP of 0.18 can be achieved in the refrigeration cycle [16]. Thus, the values of the maximum COP estimated here for SRD 1352/3 and SG/LiBr demonstrate their high potential for practical application in adsorption chillers, in particular for refrigeration cycles. 4. Conclusions Nitrogen physi-sorption, thermo-gravimetric and calorimetric measurements were carried out over selected commercially available activated carbons and over the novel composite sorbent SG/ LiBr to estimate the thermo-physical properties, the ethanol adsorption isobars and the specific heat. Isobaric data of ethanol adsorption were fitted by Dubinin e Astakhov equation. Finally, an estimation of the coefficient of performance (COP) of refrigeration and air conditioning cycle was performed for all the sorbents. The results obtained reveal that: 1. all the carbonaceous materials show high surface area and an essentially microporous structure that guarantees a high adsorption, vice versa SG/LiBr is mesoporous; 2. SRD 1352/3 exhibits the best features in terms of superficial area (2613 m2 g1) and adsorption ability (Dw ¼ 0.151 gEtOH/gAds) for a refrigeration cycle; 3. all the materials exhibit the presence of some hysteresis effect between ethanol ad/desorption branches, which however is not pronounced for the most promising sorbents SG/LiBr and SRD 1352/3; 4. for activated carbons the DA parameters evaluated from nitrogen adsorption are in close agreement with those of ethanol sorption; 5. measured specific heats are all within the same range from approximately 0.80 J/g K at 40 C to 1.10 J/g K at 100 C; 6. SG/LiBr composite allows the highest COP of 0.64 and 0.72 for a refrigeration cycle and for an air conditioning cycle, respectively; The characterization performed within this work allows the identification of SRD 1352/3 activated carbon and SG/LiBr composite as the most promising, from thermodynamic point of view, adsorbent materials suitable for the practical application in adsorption refrigeration systems based on ethanol as refrigerant. In order to identify the best working pair, also dynamic performance has to be taken into account. Such an activity will be soon performed by means of a proper experimental kinetic characterization, working under real operating conditions of a refrigerator. Acknowledgment The financial support by “Progetto Bandiera RITMARE SP2-WP3” e La Ricerca Italiana per il Mare e Coordinato dal CNR e finanziato dal MIUR nell'ambito del Programma Nazionale della Ricerca 2011e2013 is gratefully acknowledged. Nomenclature
Fig. 5. COP vs the regeneration temperature of the samples SG/LiBr and SRD 1352/3 for a refrigeration cycle.
A COP Cp DSC
differential adsorption potential, J g1 coefficient of performance specific Heat, J g1 K1 differential scanning calorimetry
V. Brancato et al. / Energy 84 (2015) 177e185
E E0 EER L p Q R T W W0 w
a b DH
characteristic energy of the working pair, J g1 characteristic energy of the adsorbent, J g1 energy efficiency ratio latent heat, J g1 pressure, Pa energy, J g1 universal gas constant, J mol1 K1 temperature, K adsorbed volume, cm3 g1 maximum adsorbed volume, cm3 g1 refrigerant uptake, g g1 coefficient of thermal expansion, K1 coefficient of affinity adsorption enthalpy, J g1
Subscripts 2 end of isosteric heating phase ads adsorption av average des desorption eq equivalent ev evaporation des desorption ish isosteric heating max maximum s saturation References [1] Zheng X, Ge TS, Wang RZ. Recent progress on desiccant materials for solid desiccant cooling system. Energy 2014;74:280e94. [2] Choudhury B, Saha BB, Chatterjee PK, Sarkar JP. An overview of developments in adsorption refrigeration system towards a sustainable way of cooling. Appl Energy 2013;12:554e67. [3] Wang RZ, Oliveira RG. Adsorption refrigeration e an efficient way to make good use of waste heat and solar energy. Prog Energy Combust Sci 2006;32: 424e58. [4] Wang SG, Wang RZ. Recent developments of refrigeration technology in fishing vessels. Renew Energy 2005;30:589e600. [5] Wang LW, Wang RZ, Oliveira RG. A review on adsorption working pairs for refrigeration. Renew Sustain Energy Rev 2009;12:518e34. [6] Fan Y, Luo L, Souyri B. Review of solar sorption refrigeration technologies: development and application. Renew Sustain Energy Rev 2011;11:1758e75. [7] Wang DC, Li YH, Li D, Xia YZ, Zhang JP. A review on adsorption refrigeration technology and adsorption deterioration in physical adsorption systems. Renew Sustain Energy Rev 2010;14:334e53. [8] Wang LW, Wang RZ, Wu JY, Wang K, Wang SG. Adsorption ice makers for fishing boats driven by the exhaust heat from diesel engine: choice of adsorption pair. Energy Convers Manag 2004;45:2043e57. [9] Aristov Y. Challenging offers of material science for adsorption heat transformation: a review. Appl Therm Eng 2013;50:1610e8. [10] Luo L, Feidt M. Comportement transitoire d'une machine frigorifique a adsorption. Etude experimentale du systeme alcool/charbon actif. Rev ne rale Therm 1997;36:159e69. Ge [11] El- Sharkawy II, Saha BB, Koyama S, He J, Ng KC, Yap C. Experimental investigation on activated carbon e ethanol pair for solar powered adsorption cooling applications. Int J Refrig 2008;31:1407e12. [12] Uddin, K., El-Sharkawy, I. I., Miyazaki, T., Saha, B.B.; Koyama, S., Kil, H.-S., et al., Adsorption characteristics of ethanol onto functional activated carbons with controlled oxygen content. Appl Therm Eng, 1e8, Available online May 2014. [13] Loh WS, El-Sharkawy II, Ng KC, Saha BB. Adsorption cooling cycles for alternative adsorbent/adsorbate pairs working at partial vacuum and pressurized conditions. Appl Therm Eng 2009;29(4):793e8. [14] El- Sharkawy II, Kuwahara K, Saha BB, Koyama S, Ng KC. Experimental investigation of activated carbon fibers/ethanol pairs for adsorption cooling system application. Appl Therm Eng 2006;26:859e65.
185
[15] El - Sharkawy II, Uddin K, Miyazaki T, Saha BB, Koyama S, Kil HS, et al. Adsorption of ethanol onto phenol resin based adsorbents for developing next generation cooling systems. Int J Heat Mass Transf 2015;81:171e8. [16] Rezk A, AL-Dadah R, Mahmoud S, Elsayed A. Investigation of ethanol/metal organic frameworks for low temperature adsorption cooling application. Appl Energy 2013;112:1025e31. [17] Saha BB, El-Sharkawy II, Miyazaki T, Koyama S, Henninger SK, Herbst A, et al. Ethanol adsorption onto metal organic framework: theory and experiments. Energy 2015;79:363e70. [18] Gordeeva L, Frazzica A, Sapienza A, Aristov Y, Freni A. Adsorption cooling utilizing the “LiBr/silica - ethanol” working pair: dynamic optimization of the adsorber/heat exchanger unit. Energy 2014;75:390e9. [19] Li M, Huang HB, Wang RZ, Wang LL, Cai WD, Yang WM. Experimental study on adsorbent of activated carbon with refrigerant of methanol and ethanol for solar ice maker. Renew Energy 2004;29:2235e44. [20] Horvath G, Kawazoe M. Method for calculation of effective pore size distribution in molecular sieve carbon. J Chem Eng Jpn 1983;16(5):470e5. [21] Barrett EP, Joyner LG, Halenda PP. The determination of pore volume and area distributions in porous substances. I. Computations from nitrogen isotherms. J Am Chem Soc 1951;73(1):373e80. [22] Ryu Z, Zheng J, Wang M, Zhang B. Characterization of pore size distributions on carbonaceous adsorbents by DFT. Carbon 1999;37:1257e64. [23] Hamamoto Y, Alam KCA, Saha BB, Koyama S, Akisawa A, Kashiwagi T. Study on adsorption refrigeration cycle utilizing activated carbon fibers. Part 1. Adsorption characteristics. Int J Refrig 2006;29:305e14. [24] Gordeeva L, Aristov Y. Novel sorbents of ethanol “salt confined to porous matrix” for adsorptive cooling. Energy 2010;35:2703e8. [25] Dubinin MM. Chemical reviews. Refrigeration 1960;60:235e66. [26] Chakraborty A, Sun B. An adsorption isotherm equation for multi-types adsorption with thermodynamic correctness. Appl Therm Eng 2014;72(2): 190e9. [27] Sun B, Chakraborty A. Thermodynamic formalism of water uptakes on solid porous adsorbents for adsorption cooling applications. Appl Phys Lett 2014;104(20):201901. [28] Aristov Y. Adsorptive transformation of heat: principles of construction of adsorbents database. Appl Therm Eng 2012;42:18e24. [29] Henninger SK, Schicktanza M, Hugenell PPC, Sievers H, Henning H-M. Evaluation of methanol adsorption on activated carbons for thermally driven chillers part I: thermophysical characterisation. Refrigeration 2012;35: 543e53. [30] Dubinin MM, Astakhov VF. Development of the concepts of volume filling of micropores in the adsorption of gases and vapors by microporous adsorbents. Bull Acad Sci USSR Div Chem Sci 1971;20(1):3e7. [31] Perry RH, Green DW. Perry's chemical engineers' handbook. McGraw Hill Companies; 1999 (s. l). [32] Bering BP, Dubinin MM, Serpinsky VV. On thermodynamics of adsorption in micropores. J Colloid Interface Sci 1972;38(1):185e94. [33] Gordeeva LG, Aristov Yu I. Composites “salt inside porous matrix” for adsorption heat transformation: a current state of the art and new trends. Int J Low-Carbon Technol 2012;7(4):288e302. [34] Prokopiev SI, Aristov YI. Selective water sorbents for multiple applications, 9. Temperature independent curves of sorption from theory of volume filling of micropores. React Kinet Catal Lett 1999;67(2):345e52. [35] Lopez-Ramon MV, Stoeckli F, Moreno-Castilla C, Carrasco-Marin F. Specific and non specific interaction between methanol and ethanol and active carbons. Langmuir 2000;16:5967e72. [36] Wood GO. Affinity coefficients of Polanyu/Dubinin adsorption isotherm equations. A review with compilations and correlations. Carbon 2001;39: 343e56. [37] Aristov Y, Tokarev MM, Cacciola G, Restuccia G. Heat capacity and heat conductivity of aqueous solutions of calcium chloride in silica gel pores. Russ J Phys Chem A 1997;71(3):327e30. [38] Doebelin Ernest O. Measurements systems: application and design. 4th ed. Singapore: McGraw e Hill International Editions; 1990. [39] El-Skarkawy II, Hassan M, Saha BB, Koyama S, Nasr MM. Study on adsorption of methanol onto carbon based adsorbents. Int J Refrig 2009;32:1579e86. [40] Saha BB, El-Sharkawy II, Chakraborty A, Koyama S. Study on an activated carbon fiber-ethanol adsorption chiller: part II e performance evaluation. Int J Refrig 2007;30:96e102. [41] Wang RZ, Jia JP, Zhu YH, Teng Y, Wu JY, Cheng J, Wang QB. Study on a new solid adsorption refrigeration pair: activated carbon fiber-methanol. J Sol Energy Eng 1997;119(3):214e8. [42] Dieng AO, Wang RZ. Literature review on solar adsorption technologies for ice-making and air conditioning purposes and recent developments in solar technology. Renew Sustain Energy Rev 2001;5:313e42.
Contents lists available at ScienceDirect
Energy journal homepage: www.elsevier.com/locate/energy
Ethanol adsorption onto carbonaceous and composite adsorbents for adsorptive cooling system V. Brancato a, *, A. Frazzica a, A. Sapienza a, L. Gordeeva b, c, A. Freni a a Consiglio Nazionale delle Ricerche (CNR), Istituto di Tecnologie Avanzate per l'Energia “Nicola Giordano” (ITAE), S. Lucia Sopra Contesse 5, 98126 Messina, Italy b Boreskov Institute of Catalysis, Pr. Lavrentieva. 5, Novosibirsk, Russia c Novosibirsk State University, Pirogovastr. 2, Novosibirsk, Russia
a r t i c l e i n f o
a b s t r a c t
Article history: Received 10 November 2014 Received in revised form 18 February 2015 Accepted 23 February 2015 Available online 21 March 2015
The aim of the present paper is the experimental characterization of adsorbent materials suitable for practical applications in adsorption refrigeration systems, employing ethanol as refrigerant. Different commercial activated carbons as well as a properly synthesized porous composite, composed of LiBr inside a silica gel host matrix, have been tested. A complete thermo-physical characterization, comprising nitrogen physi-sorption, specific heat and thermo-gravimetric equilibrium curves of ethanol adsorption over the sorbents, has been carried out. The equilibrium data have been fitted by means of the Dubinin e Astakhov equation. On the basis of the experimental data, a thermodynamic evaluation of the achievable performance of each adsorbent pair has been estimated by calculating the maximum COP (Coefficient of Performance) under typical working boundary conditions for refrigeration and air conditioning applications. The innovative composite material shows the highest thermodynamic performances of 0.64e0.72 for both tested working conditions. Nevertheless, the best carbonaceous material reaches COP value comparable with the synthesized composite. The results have demonstrated the potential of the chosen adsorbents for utilization in adsorption cooling systems. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Activated carbons Adsorption refrigeration Ethanol Porous composite adsorbent
1. Introduction With the rapid development of economy and society, our modern life and industry consume large amounts of energy for cooling. Furthermore, the depletion of fossil fuels and the threat of global warming over the last decades have challenged air conditioning industry to develop new cooling technologies alternative to the conventional vapor compression systems. Heat driven adsorption cooling cycles have the great energy saving potential due to utilization of the solar energy and low grade waste heat, e.g. engine exhaust, industrial waste heat, etc. [1,2]. The interest in adsorption cooling systems first is started to the oil crisis in the 1970s, and then later, in the 1990s, because of ecological problems related to the use of CFCs and HCFCs as refrigerants. Accordingly, systems that can recover waste heat at low
* Corresponding author. Tel.: þ39 090 624 243; fax: þ39 090 624 247. E-mail address: [email protected] (V. Brancato). http://dx.doi.org/10.1016/j.energy.2015.02.077 0360-5442/© 2015 Elsevier Ltd. All rights reserved.
temperature levels e such as adsorption systems e can be an interesting alternative to common vapor compression ones, for a wiser energy management [3]. Sorption refrigeration technologies are thermally driven systems, in which the conventional mechanical compressor of the common vapor compression cycle is replaced by a ‘thermal compressor’ [4]. This technology could be suitable, for instance, for fishing boat using waste heat discharged from the internal combustion engine. A sorption refrigeration system is noise-free, easy controllable, environmental friendly and virtually maintenance-free [3]. However the main drawback of these systems, which hinder their practical dissemination, is their reduced performance in terms of specific cooling capacity and COP, as well as cooling power [5,6]. The adsorption characteristics of the adsorbent/refrigerant pair are one of the most essential parameters that affect the system performance [7e9]. An adsorbent must have the ability to adsorb large quantities of an adsorbate in a narrow range of temperatures and to desorb it easily when temperature rises. Its properties
178
V. Brancato et al. / Energy 84 (2015) 177e185
should not change with age and use (i.e. high hydrothermal stability). Water, ammonia, methanol and ethanol are typical refrigerants of the adsorption refrigerator. While silica gel, zeolites and activated carbons are the most common adsorbents used in these systems. Among the common refrigerants, ethanol is one of the most interesting due to its low freezing point (114 C) and to its ecological compatibility. Respect to methanol, ethanol has a similar saturation pressure, but its latent heat is about 30% lower than methanol. Nevertheless, methanol is characterized by high toxicity and inflammability compared to ethanol and for this reason special care must be used during the phases of storage, handling and use. Methanol also shows the problem of dissociation above 120 C, especially in the presence of copper. The typical ethanol adsorbents used in adsorption applications are activated carbons, in fact these latter have a large surface area and exhibit good affinity towards ethanol. A good number of papers related to ethanol adsorption refrigeration have already been published. The activated carbons are common materials proposed as ethanol adsorbents for adsorption cooling [10]. Accordingly, other authors further investigated the features of this activated carbon as discussed below. El-Sharkawy et al. [11] measured the adsorption isotherms Maxsorb III/ethanol pair. They found that this pair is attractive for solar cooling application. Indeed, a cooling cycle realized with this adsorption pair can achieve a specific cooling effect as high as 420 kJ/kg at evaporator temperature of 7 C in combination with heat source and heat sink of temperatures 80 and 30 C, respectively. Maxsorb III/ethanol as adsorption pair was investigated also by Uddin et al. [12]. They studied the characteristics of ethanol adsorption onto “parent” Maxsorb III and surface treated Maxsorb III with controlled oxygen content. Experimental results show that adsorption capacity improves in H2 treated Maxsorb III, the KOHeH2 treated Maxsorb III highlights faster adsorption kinetics but the isosteric heat of adsorption remains higher in the parent Maxsorb III/ethanol pair. Loh et al. [13] have presented the performance analysis of both ideal single-stage and single-effect double-lift adsorption cooling cycles working at partially evacuated and pressurized conditions of six specimens of adsorbents and refrigerant pairs. They found that the activated carbon fibers (AC/20)/ethanol pair provides the highest value of specific cooling effect among the tested pairs. El-Sharkawy et al. [14] have also investigated the activated carbon fibers/ethanol pairs for adsorption cooling system. They have studied the adsorption capacity and adsorption rate of the ACF A-15 and A-20. They found that the equilibrium adsorption capacity of A-20/ethanol pair is considerably larger than that of A-15/ ethanol pair and that the ACF A-20 seems to be a promising adsorbent as it has large ethanol adsorption capacity related per gram of the adsorbent. However the low packing density of ACF reduces their capacity related per the unit volume. The same authors [15] have also analyzed the adsorption characteristics of ethanol onto two promising adsorbents based on spherical phenol resin treated with different mass ratios of KOH named as KOH4-PR and KOH6-PR. They found that the adsorption uptake of ethanol onto the sorbents are 1.43 and 2 kg/kg, respectively. This result shows that the newly developed adsorbents have promising adsorption characteristics with ethanol that may lead to the development of next generation of adsorption chillers, in particular for air conditioning cycles. Under conditions of ice making cycle, means at evaporation temperature 3 C the ethanol uptake on KOH4-PR and KOH6-PR is much lower 0.3e0.4 g/g. Recently, Rezk et al. and [16] Saha et al. [17] have been presented the experimental and theoretical investigations of adsorption characteristics of
ethanol onto metal organic framework namely MIL-101Cr. The adsorption capacity of MIL-101Cr reaches as high as 1.1 g/g at ethanol relative pressure of 1, while under conditions of the typical air conditioning cycle (evaporation and adsorption temperatures of 10 and 30 C, respectively) it is 0.9 g/g. At ice making cycle (evaporator temperature of 3 C) the uptake falls down but remains quite high e 0.45 g/g. The results demonstrate that the MIL-101Cr e ethanol is promising working pair, however more studies on the durability, thermal properties (namely the heat capacity, thermal conductivity) etc. are required for evaluation of its practical potential for the adsorption cooling applications. Recently, a new adsorbent, namely composites sorbent salt/silica gel, have been proposed for ethanol sorption cooling. Gordeeva et al. [18] found that the composite LiBr/SiO2 is the promising composite ethanol adsorbent due to the large variation in the ethanol uptake for both air conditioning and ice making cycles. Thus, the data available in the literature shows that various carbonaceous materials, MIL-101Cr and composite LiBr/SiO2 are the most promising ethanol adsorbents in the adsorption cooling. However, the major part of the studies are related to the conditions of air conditioning cycle. The materials having the greatest potential in terms of ethanol uptake, namely KOH4-PR, KOH6-PR, and MIL-101Cr are not commercially available, and consequently can hardly be considered for practical application in the adsorption chillers. As regards the traditional ethanol adsorbents, activated carbons, it was shown by Li et al. [19] that an activated carbon e ethanol working pair cannot allow ice production. Therefore, an empty room still exists for the search of ethanol adsorbents suitable for practical application in the adsorption chillers. The goal of the present paper is the investigation of the adsorption characteristics of ethanol on some on-purposely selected commercially available activated carbons, produced from different bases (coal, coconut shell) and having different morphology (grains, pellets, fibers) with the aim to find the most suitable material for a sorption cooling system, in particular, for an ice making cycle. Additionally, the previously mentioned novel composite LiBr/SiO2 was investigated and compared with the selected activated carbons. The ethanol sorption isobars of the selected samples were valued by a thermo-gravimetric system, specifically realized for alcohols adsorption measurement under real adsorption cooling conditions. The measured adsorption data were correlated with the nitrogen physi-sorption results and they were analyzed according to the DubinineAstakhov theory in order to evaluate the equilibrium adsorption coefficients and sorption heat. Calorimetric measurements were carried out for all the samples to evaluate the specific heat. The evaluated experimental data were used to carry out thermodynamic analysis with the aim to estimate the performance of a refrigeration and air conditioning cycle in terms of cooling COP. 2. Experimental 2.1. Selected materials Five different commercial activated carbons, supplied by different companies, have been characterized and evaluated for the use in adsorption refrigeration systems. The samples are chosen with different bases, ranging from coconut shell to coal based, and with different shapes, from grains to fibers. Table 1 reports the list of the tested samples. In addition, the innovative composite sorbent LiBr/silica gel (SG/ LiBr) has been studied. The silica gel KSK (“Salavatnefteorgsyntes”, Russia) with the specific surface area Ssp ¼ 280 m2/g, the pore volume Vp ¼ 1.0 cm3/g, the average pore size dav ¼ 15 nm was used as a porous matrix. The composite sorbent SG/LiBr was prepared
V. Brancato et al. / Energy 84 (2015) 177e185 Table 1 Tested samples.
179
Table 2 Specifications of load cell and of pressure gauges.
Sample
Manufacture
Product form
Based
Particle size (mm)
SRD 1352/3 FR20 AP4 e 60 ATO COC eL1200 SG/LiBr
Chemviron Kuraray Chemviron Eurocarb Carbon Activated e
Grains Fiber Pellets Grains Grains Grains
Coconut shell e Coal Coconut shell Coconut shell Silica gel/LiBr
0.5e2 F 0.01 >4 0.25e0.6 0.42e1 0.7e0.85
following the preparation reported in Ref. [18]. The LiBr content (CLiBr) in the anhydrous composite was 19 wt.%. Anhydrous ethanol, with a purity of 99.9%, has been used as refrigerant. 2.2. Experiments The evaluation of the different tested materials is based on detailed data of thermo-physical properties including surface and pore characteristics, adsorption equilibrium data and specific heat. The texture properties of the activated carbons and of the porous composite were characterized by a Micromeritics ASAP 2020, using N2 as the adsorbate at 77 K. Prior to the measurements, the samples have been dried at 160 C under continuous evacuation for 3 h. Adsorption and desorption isotherms were performed in order to detect possible hysteresis effects. The BET (BrunauereEmmetteTeller) theory was applied to determine the specific surface area of mesoporous sample (SG/LiBr) while Langmuir theory was selected to evaluate the specific surface area of activated carbons. HorvatheKawazoe [20] method was applied to evaluate the cumulative pore volume and the median pore width in the microporous samples, while cumulative volume of pores of the composite material was evaluated from BJH (BarretteJoynereHalenda) desorption theory. In this last case the pore width was calculated as 4V/ABET [21]. Ethanol adsorption measurements were performed by a thermo-gravimetric system specifically designed to this scope. Fig. 1 reports both the system layout and the realized apparatus. The core component of the experimental apparatus is the load cell (see Table 2 for specifications), that measures the variations of the sample mass during the adsorption process. In accordance to Fig. 1a, the testing chamber (2) is submersed into a thermo cryostat (1). An evaporator made by glass (4), connected to a thermo cryostat (5), two pressure gauges (P1 and P2 e see Table 2 for specifications), three valves (V1, V2, V3) and a vacuum pump (3) are the other components of the plant. All physical parameters (pressure, temperature) were acquired every 60s through a display system
Load cell mod. PW4C3-300g supplied by HBM
Maximum capacity (Emax) Minimum LC verification interval (vmin) Sensitivity (Cn) Zero signal Pressure gauge mod. VSC43MA4 Measuring Range supplied by Thyracont Accuracy
g g
300 0.05
mV/V 1.0 ± 0.1 0 ± 0.1 mbar 1-1400 e 0.3% f.s.
and data acquisition by means of a software interface specifically developed in LabView. For each test 7 e 8 g of adsorbent material were used. Prior to all the measurements the samples were dried in an oven at 150 C for at least 12 h to determine their dry mass (reference mass). After drying, the sample was put in a nickel crucible in the testing chamber and then evacuated by a vacuum pump for at least 6 h at 120 C. Then the valve between the test section and the evaporator was opened and adsorption started. The measurements were conducted under isobaric conditions at four different ethanol pressures: 12.6 mbar (3 C), 18.0 mbar (þ2 C), 25.5 mbar (þ7 C) and 35.4 mbar (þ12 C). Seven different equilibrium points were measured between 30 C and 120 C. The specific heat of the samples has been measured by a differential scanning calorimeter, DSC mod. 27 HP (Mettler Toledo). Approximatively, 10e15 mg of sample were dried for 12 h at 150 C to remove moisture before the experiment and then inserted into a sealed crucible, to prevent adsorption of moisture from the air. Subsequently, the measurements were carried out from room temperature up to 160 C, with a heating rate of 2 C/min in air. 3. Results and discussion 3.1. Physi-sorption nitrogen measurements Fig. 2 shows the N2 adsorption/desorption isotherms for activated carbons (Fig. 2a) and for the porous composite material SG/ LiBr (Fig. 2b). Only one isotherm is represented for activated carbons AP-60 and COC L-1200, because these two sorbents showed overlapped curves due to the same texture characteristics. All the tested activated carbons exhibit adsorption isotherm of type I according to IUPAC [22] typical of microporous solids. The very steep region at low p/ps is due to the volume filling of very narrow pores and limiting uptake is dependent on the accessible micropore volume rather than on the internal surface area [23]. It is observed that isotherm curves of carbon SRD1352/3, ATO and COC L-1200 show a little hysteresis at p/ps > 0.45, which may be caused
Fig. 1. a) schema of thermo-gravimetric plant; b) photo of thermo-gravimetric plant.
180
V. Brancato et al. / Energy 84 (2015) 177e185
Fig. 2. Adsorption/desorption isotherm of nitrogen on tested samples: a) activated carbons; b) composite (SG/LiBr).
by capillary condensation, due to the presence of some mesopores in these carbons. On the contrary, the isotherm of the activated carbon fibers (FR 20) is completely reversible in the whole range of p/ps indicating its microporous structure. As illustrated in Fig. 2b, porous composite SG/LiBr exhibits adsorption isotherm of type IV with a strong hysteresis at p/ps ¼ 0.8e0.9 that is typical of mesoporous solids. The calculated surface areas of the activate carbon are in the range from 1412 to 2613 m2 g1 (see Table 3) and it does not depend on the precursor of the carbon. Instead, the evaluated area of the porous composite SG/LiBr is 181 m2 g1. As before demonstrated by Gordeeva [18] no pore blocking of the silica gel by LiBr occurs during the composite preparation. In the activated carbons a large surface area is preferable for ensuring large adsorption capacity. However, large surface area in a limited volume inevitably gives rise to large number of small size pores between adsorption surfaces. The size of pore determines the accessibility of adsorbate molecules to the adsorption surface, so the pore size width is one of the most important properties for characterizing adsorptive of this adsorbent [23]. To value the pore size a series of method based on Kelvin equation have been proposed, but this theory becomes progressively less accurate as the pore size decreases and breaks down for micropores. In the present work, to value the cumulative pore volume and the median pore width in activated carbons the HorvatheKawazoe [20] method was applied, which is a semi-empirical, analytic model of adsorption in micropores that is commonly used for determining the pore size distribution of microporous materials. Instead, the cumulative pore volume and the pore width of the composite material was evaluated from BJH [21] desorption theory because this material is mesoporous. Table 3 shows a comparison of the found values for each tested sample. Among the activated carbon, the fibers FR 20 exhibit the largest pore volume (0.75 cm3/g), while the granular AP4-60 the smallest (0.47 cm3/g) with a pore width of 0.59 nm and 0.64 nm, respectively. A value of 0.73 cm3/g and pore width of 17.2 nm were found for the mesoporous composite sample.
Table 3 Thermophysical properties of tested samples determined by nitrogen adsorption. Sample
Surface area (m2/g)
Pore size (nm)
Cumulative pore volume (cm3/g)
SRD 1352/3 FR20 AP4 e 60 ATO COC e L1200 SG/LiBr
2613 2180 1428 1745 1412 181
0.56 0.59 0.64 0.59 0.59 17.2
0.65 0.75 0.47 0.64 0.49 0.73
Thus, all the carbonaceous materials show high surface area and an essentially microporous structure that promotes a high ethanol adsorption. 3.2. Ethanol adsorption measurements Ethanol adsorption measurements have been performed under isobaric conditions. In particular, four different isobars were measured at different pressure from 12.6 mbar (3 C) to 35.4 mbar (þ12 C). Some measures have been done in adsorption/desorption mode to verify the presence of hysteresis effects. Fig. 3 shows the adsorption e desorption isobars for all the tested samples. All the materials exhibit the presence of hysteresis effect between adsorption and desorption branches even if in different magnitude. The trend of the uptake depends on the nature of the adsorbent. Activated carbon shows a monotonic increase of the uptake with the decrease of temperature. Differently, the porous material SG/LiBr exhibits a critical temperature (e.g. 60 C at 12.6 mbar) under which the ethanol adsorption becomes significant. Gordeeva et al. [24] found that the sorption rise is caused by the reaction between the salt and the ethanol and the sorption ability of the composite depends mainly on the nature of the confined salt, while the matrix acts as a media that disperses the salt and provides efficient heat and mass transfer. 3.3. Dubinin e Astakhov transformation The measured data were transformed into a characteristic curve following the Dubinin formalism that describes the volume filling of micropores [25]. A number of empirical and semi-empirical isotherms like traditional Toth, Frendlich, Sips equations, as well as a new one for multi-types adsorption, which includes the loading, the adsorbent/adsorbate interaction and the surface structural heterogeneity factors has been developed [26,27]. Nevertheless, the DubinineRadusgkevich and Dubinin e Astakhov approaches are still the most employed model, especially to analyze the adsorption equilibrium of microporous carbonaceous materials for which the adsorption mechanism is the volume filling of micropores [28,29]. Therefore, the ethanol adsorption data were fitted to the Dubinin e Astakhov equation [30].
n A A n ¼ W0 exp W ¼ W0 exp E bE0
(1)
in which: W (cm3/g) defines the adsorbed volume for the given working pair; W0 (cm3/g) is the total accessible pore volume, which corresponds to a maximum loading; E (J/mol) is the characteristic energy for the working pair and depends on the sorbent as well as
V. Brancato et al. / Energy 84 (2015) 177e185
181
Fig. 3. Adsorption and desorption isobars: a) porous composite (SG/LiBr); b) activated carbon SRD 1352/3; c) fiber of activated carbon FR20; d) activated carbon AP4 e 60; e) activated carbon ATO; f) activated carbon COC-L1200.
on the adsorptive; b is the coefficient of affinity and depends only on the adsorbent; E0 (J/mol) indicates the characteristic adsorption energy depending only on the adsorbent. The exponent n is empirically determined. Its value depends on the pore structures. High values indicate a material with a homogenous pore structure, while low values indicate a heterogeneous structure. A (J/mol) is the differential adsorption potential, defined as follows:
A ¼ RTln
ps p
(2)
The adsorption potential A represents the difference of the chemical potentials of the adsorbent and the free liquid. Moreover, it is the differential change of free energy in a reversible, isothermal transition from liquid phase to the adsorbate phase. The reference pressure ps is the saturated vapor pressure at T.
The data transformation from gravimetric units (gEtOH/gAds), as reported in Fig. 3, into volumetric unit (cm3EtOH/gAds) has been performed considering ethanol density as a function of temperature, by means of the equation suggested by Ref. [31]. The differential heat of adsorption can be calculated from the adsorption equilibrium. Since the adsorption equilibrium was derived from the characteristic curve, the differential heat of adsorption results from the following equation [32], where the non-ideality of gaseous phase effects is not considered:
DH ¼ L þ A TaW
vDA vW
(3) T
where L (J/mol) is the heat of evaporation (evaluated considering the medium temperature of the working cycle), a is the coefficient
182
V. Brancato et al. / Energy 84 (2015) 177e185
of thermal expansion of the adsorbent and W is the adsorbed volume. The isobaric data were then transformed according to Dubinin e Astakhov approach as illustrated in Fig. 4. As can be seen from the diagram, the transformed data generate a very smooth curve. Results of the characterization including all parameters for DA equation are provided in Table 4. The goodness of the fit of the activated carbons has been evaluated considering the plot of ln(W) vs (A)n at the optimal value of n. Table 4 also reports the values of R2 estimated from the plot. All values are close to 1, so this confirms that DA transformation properly describes the adsorption properties of the investigated working pairs. In addition, the physi-sorption measurements have also confirmed the goodness of the fit because the calculated values of W0 from nitrogen adsorption are in close agreement with those of ethanol sorption.
Table 4 DA parameters for the tested samples. Sample
SG/LiBr SRD 1352/3 FR 20 AP4 e 60 ATO COC e L1200
Ethanol adsorption
Nitrogen adsorption
W0 (cm3/g) E (kJ/mol) n
R2
W0 (cm3/g) E (kJ/mol) n
0.68 0.82 0.75 0.45 0.61 0.44
e 0.988 0.990 0.988 0.978 0.994
e 0.78 0.74 0.44 0.63 0.52
6.9 8.78 13.5 10.6 11.2 13.3
1.8 1.5 2 2 1.7 2
e 18.9 20.3 18.9 20.9 23.9
e 1.7 1.6 1.5 1.4 1.4
Composites “salt inside porous matrix” are mesoporous solids and adsorption of water and alcohols includes different physicalchemical process, namely the adsorption on the active surface centers, the formation of the crystalline solvates “salt e sorbate”
Fig. 4. DA transformation of the equilibrium values for: a) composite (SG/LiBr); b) activated carbon SRD 1352/3; c) fiber of activated carbon FR 20; d) activated carbon AP4 e 60; e) activated carbon ATO; f) activated carbon COC - L1200.
V. Brancato et al. / Energy 84 (2015) 177e185
and, finally, the formation of the solution “ salt e sorbates” inside pores [33]. However, it was shown earlier [34] that, despite mesoporous structure of the composites “salt inside porous matrix” and complex sorption mechanism, the water sorption on these materials can be formally described by DA equation as well. Therefore, for complete discussion of the results, DA equation has been applied also to ethanol adsorption on the composite material SG/ LiBr. As expected, the DA parameter W0 ¼ 0.68 cm3/g calculated from ethanol adsorption for SG/LiBr presents a good correspondence with pore volume (0.73 cm3/g) calculated from nitrogen adsorption. The DA equation is not applicable for nitrogen adsorption on the composite because in this case the adsorption occurs only on the surface and the capillary condensation in mesopores is also present. For calculation of DA parameters of all activated carbons, the value of affinity coefficient b was 0.61 as reported by Lopez-Ramon [35] and by Wood [36]. The maximum loading capacity W0 follows the increasing surface area, starting from the carbon COC L-1200 to the carbon SRD 1352/3. In fact, COC L-1200 exhibits a surface area of 1412 m2/g and a W0 of 0.44 cm3/g, while the SRD 1352/3 shows the highest surface area (2613 m2/g) and the highest adsorbed volume (0.82 cm3/g). However, all samples reveal a poor adsorption capacity for an adsorption potential higher than 450 J/g. The porous composite SG/LiBr shows a smaller sorption at A > 250 J/g and a higher slope in the range 0e250 J/g compared to the curves of all activated carbons. This behavior is attributed to the different mechanism of the ethanol sorption by the composite. 3.4. Specific heat measurements The evaluation of specific heat of the dry adsorbents has been carried out by a differential scanning calorimeter DSC. The measured specific heats for all the samples are within the same range from approximately 0.80 J/g K at 40 C to 1.10 J/g K at 100 C. Table 5 shows the measured values for each tested sample. The activated carbon with coal as precursor (AP4-60) exhibits the lowest specific heat, while the carbon SRD 1352/3, produced from coconut shell, the highest. In addition, the found value for the porous composite SG/LiBr is in close agreement with the value estimated by Gordeeva et al. [24]. 3.5. Estimation of the cooling performance The obtained data allow an estimation of the coefficient of performance COP for both refrigeration and air conditioning cycles for each working pairs, considering the following characteristic temperature: 90 C as maximum desorption temperature (Tdes) Table 5 Experimental specific heat for tested samples. Sample
T ( C)
Cp (J/g∙K)
Sample
T ( C)
Cp (J/g∙K)
ATO
40 60 80 100 40 60 80 100 40 60 80 100
0.94 0.97 1.04 1.10 1.05 1.07 1.14 1.21 0.73 0.77 0.82 0.86
COC e L1200
40 60 80 100 40 60 80 100 40 60 80 100
0.85 0.88 0.96 1.03 0.80 0.82 0.90 0.95 0.81 0.82 0.84 0.90
SRD 1352/3
AP4 e 60
FR 20
SG/LiBr
183
30 C as minimum adsorption and condensation temperature (Tad-cond); 2 C as evaporation temperature (Tev) for refrigeration cycle and þ7 C as evaporation temperature (Tev) for air conditioning cycle. COP has been calculated as:
COP ¼
Q ev ðQ sens þ Q lat Þ
¼
L$Dw cp eq ðwmax Þ$ðT2 Tads Þ þ cp eq ðwav Þ$ðTdes T2 Þ þ DH$w (4)
where Qev (J/g), represents the useful cooling effect due to the ethanol evaporation, Qsens (J/g), the sensible energy delivered to heat up the adsorbent and Qlat (J/g), the latent heat needed to desorb ethanol from the adsorbent material. In the calculation, the heating up of the ethanol inside the evaporator, due to the flowing of the liquid from the condenser, has been omitted, since its effect can be considered negligible. In particular, Dw (gEtOH/gAds) is the uptake variation between adsorption and desorption; DH (J/g) is the desorption heat; L (J/g) is the latent heat of ethanol, evaluated at the evaporation temperature; cpeq(wmax) (J/g$K), is the equivalent specific heat evaluated during the isosteric heating phase; cpeq(wav) (J/g$K) is the equivalent specific heat evaluated during the desorption phase and finally T2 is the temperature of the end of isosteric heating phase. According to the following equation (5), the equivalent specific heat of the adsorbents saturated with ethanol up to uptake w has been evaluated as linear combination of the specific heat of the dry adsorbent and that of the adsorbed ethanol in the liquid state, as suggested by Ref. [37].
cp eq ¼
cp adsorbent þ cp EtOH $w 1þw
(5)
The equation (5) implies that the specific heat of adsorbed ethanol is considered equal to that for liquid ethanol. The estimated values of variation in ethanol uptake and COP are reported in Table 6. The ethanol uptake variation has been evaluated as difference between the maximum and the minimum uptake calculated starting from the equilibrium values of Dubinin's fit for the considered cycle. Starting from the metrological data of the employed sensor (see Table 2), according to the theory of the uncertainty of measurements, the typical experimental error was calculated by the method of the extended uncertainty where the composed error contribute of each measurement device as well as the effect of the desired confidence level are considered, as reported in Ref. [38]. An uncertainty around 4% has been calculated for COP evaluation. For the refrigeration cycle, the lowest COP value, 0.39, was found for activated carbon COC L-1200. This carbon presents also the
Table 6 COP for tested samples. Sample
SG/LiBr SRD 1352/3 FR 20 AP4 e 60 ATO COC e L1200
Refrigeration cycle
Air conditioning cycle
Dw (gEtOH/gAds)
COP
Dw (gEtOH/gAds)
COP
0.162 0.151 0.103 0.086 0.103 0.062
0.64 0.55 0.47 0.49 0.48 0.39
0.260 0.235 0.143 0.122 0.152 0.086
0.72 0.63 0.53 0.58 0.55 0.45
184
V. Brancato et al. / Energy 84 (2015) 177e185
lowest variation in ethanol uptake, 0.062 g/g. Instead the carbon SRD 1352/3 reveals the highest variation in ethanol uptake (0.151 g/ g) and the highest COP (0.55) among the activated carbons. These results are in close agreement with the others above discussed, indeed the activated carbon SRD 1352/3 has the highest specific surface area and the highest maximum loading capacity evaluated by DA transformation, vice versa the COC L-1200 shows the worst features. The other carbons show an intermediate behavior between COC L-1200 and SRD 1352/3. The activated carbon based on coal (AP4-60) exhibits a low variation in ethanol uptake (0.086 g/g), but it has also the lowest specific heat so the resulting COP is satisfying (0.49). The same considerations can be applied for the obtained results for the air conditioning cycle. All found values are not far from those found by El-Sharkawy [11,39] and by Saha [40]. In both cases, the highest COP value has been obtained for the composite SG/LiBr. Moreover, COP and ethanol uptake variation have been also calculated for a refrigeration cycle with a Tad-cond equal to 25 C, because this is a realistic operative condition for some applications, such as in the fishing boat refrigeration, abovementioned, that employs sea water for condensation. Interestingly, the decreasing of only 5 C of Tad-cond improves the COP value and the Dw that become comparable or even higher than the values found for the air conditioning cycle. The SG/LiBr was again the best adsorbent showing a COP of 0.69 and a Dw of 0.243 g/g. Among activated carbons the SRD 1352/3 exhibited again the highest performances (COP ¼ 0.62 and Dw ¼ 0.231 g/g) while the COC L-1200 the worst behavior (COP ¼ 0.46 and Dw ¼ 0.094 g/g). Fig. 5 shows the COP vs the regeneration temperature of the samples SG/LiBr and SRD 1352/3, reported as instance of the studied materials, for a refrigeration cycle considering the characteristic temperatures above mentioned (Tad-cond ¼ 30 C). The diagram highlights that high regeneration temperatures are preferred for activated carbon SRD 1352/3 while the regeneration temperature has a lower influence on the performance of the composite material SG/LiBr. Indeed, this last shows a slower decrease of the COP with the decrease of the regeneration temperature. On the contrary, regeneration temperatures lower 80 C negatively affect the achievable performance of both samples (activated carbon and composite material). The maximum theoretical COP ¼ 0.3e0.5, reported in Ref. [41] for “activated carbons e methanol” and “ACFs e methanol” pairs at regeneration temperature above 100 C is significantly lower than those estimated for SRD 1352/3 e ethanol and SG/LiBr e ethanol pairs. Some adsorption machines based on zeolites e
methanol, activated carbons e methanol, activated carbons e ammonia, zeolites e water working pairs were realized in actual practice and operated with real COP varying from 0.12 to 0.33 [42,10]. The modeling of a commercially available two-bed silica gel/water adsorption chiller with “MIL-101Cr e ethanol” working pair demonstrated that COP of 0.18 can be achieved in the refrigeration cycle [16]. Thus, the values of the maximum COP estimated here for SRD 1352/3 and SG/LiBr demonstrate their high potential for practical application in adsorption chillers, in particular for refrigeration cycles. 4. Conclusions Nitrogen physi-sorption, thermo-gravimetric and calorimetric measurements were carried out over selected commercially available activated carbons and over the novel composite sorbent SG/ LiBr to estimate the thermo-physical properties, the ethanol adsorption isobars and the specific heat. Isobaric data of ethanol adsorption were fitted by Dubinin e Astakhov equation. Finally, an estimation of the coefficient of performance (COP) of refrigeration and air conditioning cycle was performed for all the sorbents. The results obtained reveal that: 1. all the carbonaceous materials show high surface area and an essentially microporous structure that guarantees a high adsorption, vice versa SG/LiBr is mesoporous; 2. SRD 1352/3 exhibits the best features in terms of superficial area (2613 m2 g1) and adsorption ability (Dw ¼ 0.151 gEtOH/gAds) for a refrigeration cycle; 3. all the materials exhibit the presence of some hysteresis effect between ethanol ad/desorption branches, which however is not pronounced for the most promising sorbents SG/LiBr and SRD 1352/3; 4. for activated carbons the DA parameters evaluated from nitrogen adsorption are in close agreement with those of ethanol sorption; 5. measured specific heats are all within the same range from approximately 0.80 J/g K at 40 C to 1.10 J/g K at 100 C; 6. SG/LiBr composite allows the highest COP of 0.64 and 0.72 for a refrigeration cycle and for an air conditioning cycle, respectively; The characterization performed within this work allows the identification of SRD 1352/3 activated carbon and SG/LiBr composite as the most promising, from thermodynamic point of view, adsorbent materials suitable for the practical application in adsorption refrigeration systems based on ethanol as refrigerant. In order to identify the best working pair, also dynamic performance has to be taken into account. Such an activity will be soon performed by means of a proper experimental kinetic characterization, working under real operating conditions of a refrigerator. Acknowledgment The financial support by “Progetto Bandiera RITMARE SP2-WP3” e La Ricerca Italiana per il Mare e Coordinato dal CNR e finanziato dal MIUR nell'ambito del Programma Nazionale della Ricerca 2011e2013 is gratefully acknowledged. Nomenclature
Fig. 5. COP vs the regeneration temperature of the samples SG/LiBr and SRD 1352/3 for a refrigeration cycle.
A COP Cp DSC
differential adsorption potential, J g1 coefficient of performance specific Heat, J g1 K1 differential scanning calorimetry
V. Brancato et al. / Energy 84 (2015) 177e185
E E0 EER L p Q R T W W0 w
a b DH
characteristic energy of the working pair, J g1 characteristic energy of the adsorbent, J g1 energy efficiency ratio latent heat, J g1 pressure, Pa energy, J g1 universal gas constant, J mol1 K1 temperature, K adsorbed volume, cm3 g1 maximum adsorbed volume, cm3 g1 refrigerant uptake, g g1 coefficient of thermal expansion, K1 coefficient of affinity adsorption enthalpy, J g1
Subscripts 2 end of isosteric heating phase ads adsorption av average des desorption eq equivalent ev evaporation des desorption ish isosteric heating max maximum s saturation References [1] Zheng X, Ge TS, Wang RZ. Recent progress on desiccant materials for solid desiccant cooling system. Energy 2014;74:280e94. [2] Choudhury B, Saha BB, Chatterjee PK, Sarkar JP. An overview of developments in adsorption refrigeration system towards a sustainable way of cooling. Appl Energy 2013;12:554e67. [3] Wang RZ, Oliveira RG. Adsorption refrigeration e an efficient way to make good use of waste heat and solar energy. Prog Energy Combust Sci 2006;32: 424e58. [4] Wang SG, Wang RZ. Recent developments of refrigeration technology in fishing vessels. Renew Energy 2005;30:589e600. [5] Wang LW, Wang RZ, Oliveira RG. A review on adsorption working pairs for refrigeration. Renew Sustain Energy Rev 2009;12:518e34. [6] Fan Y, Luo L, Souyri B. Review of solar sorption refrigeration technologies: development and application. Renew Sustain Energy Rev 2011;11:1758e75. [7] Wang DC, Li YH, Li D, Xia YZ, Zhang JP. A review on adsorption refrigeration technology and adsorption deterioration in physical adsorption systems. Renew Sustain Energy Rev 2010;14:334e53. [8] Wang LW, Wang RZ, Wu JY, Wang K, Wang SG. Adsorption ice makers for fishing boats driven by the exhaust heat from diesel engine: choice of adsorption pair. Energy Convers Manag 2004;45:2043e57. [9] Aristov Y. Challenging offers of material science for adsorption heat transformation: a review. Appl Therm Eng 2013;50:1610e8. [10] Luo L, Feidt M. Comportement transitoire d'une machine frigorifique a adsorption. Etude experimentale du systeme alcool/charbon actif. Rev ne rale Therm 1997;36:159e69. Ge [11] El- Sharkawy II, Saha BB, Koyama S, He J, Ng KC, Yap C. Experimental investigation on activated carbon e ethanol pair for solar powered adsorption cooling applications. Int J Refrig 2008;31:1407e12. [12] Uddin, K., El-Sharkawy, I. I., Miyazaki, T., Saha, B.B.; Koyama, S., Kil, H.-S., et al., Adsorption characteristics of ethanol onto functional activated carbons with controlled oxygen content. Appl Therm Eng, 1e8, Available online May 2014. [13] Loh WS, El-Sharkawy II, Ng KC, Saha BB. Adsorption cooling cycles for alternative adsorbent/adsorbate pairs working at partial vacuum and pressurized conditions. Appl Therm Eng 2009;29(4):793e8. [14] El- Sharkawy II, Kuwahara K, Saha BB, Koyama S, Ng KC. Experimental investigation of activated carbon fibers/ethanol pairs for adsorption cooling system application. Appl Therm Eng 2006;26:859e65.
185
[15] El - Sharkawy II, Uddin K, Miyazaki T, Saha BB, Koyama S, Kil HS, et al. Adsorption of ethanol onto phenol resin based adsorbents for developing next generation cooling systems. Int J Heat Mass Transf 2015;81:171e8. [16] Rezk A, AL-Dadah R, Mahmoud S, Elsayed A. Investigation of ethanol/metal organic frameworks for low temperature adsorption cooling application. Appl Energy 2013;112:1025e31. [17] Saha BB, El-Sharkawy II, Miyazaki T, Koyama S, Henninger SK, Herbst A, et al. Ethanol adsorption onto metal organic framework: theory and experiments. Energy 2015;79:363e70. [18] Gordeeva L, Frazzica A, Sapienza A, Aristov Y, Freni A. Adsorption cooling utilizing the “LiBr/silica - ethanol” working pair: dynamic optimization of the adsorber/heat exchanger unit. Energy 2014;75:390e9. [19] Li M, Huang HB, Wang RZ, Wang LL, Cai WD, Yang WM. Experimental study on adsorbent of activated carbon with refrigerant of methanol and ethanol for solar ice maker. Renew Energy 2004;29:2235e44. [20] Horvath G, Kawazoe M. Method for calculation of effective pore size distribution in molecular sieve carbon. J Chem Eng Jpn 1983;16(5):470e5. [21] Barrett EP, Joyner LG, Halenda PP. The determination of pore volume and area distributions in porous substances. I. Computations from nitrogen isotherms. J Am Chem Soc 1951;73(1):373e80. [22] Ryu Z, Zheng J, Wang M, Zhang B. Characterization of pore size distributions on carbonaceous adsorbents by DFT. Carbon 1999;37:1257e64. [23] Hamamoto Y, Alam KCA, Saha BB, Koyama S, Akisawa A, Kashiwagi T. Study on adsorption refrigeration cycle utilizing activated carbon fibers. Part 1. Adsorption characteristics. Int J Refrig 2006;29:305e14. [24] Gordeeva L, Aristov Y. Novel sorbents of ethanol “salt confined to porous matrix” for adsorptive cooling. Energy 2010;35:2703e8. [25] Dubinin MM. Chemical reviews. Refrigeration 1960;60:235e66. [26] Chakraborty A, Sun B. An adsorption isotherm equation for multi-types adsorption with thermodynamic correctness. Appl Therm Eng 2014;72(2): 190e9. [27] Sun B, Chakraborty A. Thermodynamic formalism of water uptakes on solid porous adsorbents for adsorption cooling applications. Appl Phys Lett 2014;104(20):201901. [28] Aristov Y. Adsorptive transformation of heat: principles of construction of adsorbents database. Appl Therm Eng 2012;42:18e24. [29] Henninger SK, Schicktanza M, Hugenell PPC, Sievers H, Henning H-M. Evaluation of methanol adsorption on activated carbons for thermally driven chillers part I: thermophysical characterisation. Refrigeration 2012;35: 543e53. [30] Dubinin MM, Astakhov VF. Development of the concepts of volume filling of micropores in the adsorption of gases and vapors by microporous adsorbents. Bull Acad Sci USSR Div Chem Sci 1971;20(1):3e7. [31] Perry RH, Green DW. Perry's chemical engineers' handbook. McGraw Hill Companies; 1999 (s. l). [32] Bering BP, Dubinin MM, Serpinsky VV. On thermodynamics of adsorption in micropores. J Colloid Interface Sci 1972;38(1):185e94. [33] Gordeeva LG, Aristov Yu I. Composites “salt inside porous matrix” for adsorption heat transformation: a current state of the art and new trends. Int J Low-Carbon Technol 2012;7(4):288e302. [34] Prokopiev SI, Aristov YI. Selective water sorbents for multiple applications, 9. Temperature independent curves of sorption from theory of volume filling of micropores. React Kinet Catal Lett 1999;67(2):345e52. [35] Lopez-Ramon MV, Stoeckli F, Moreno-Castilla C, Carrasco-Marin F. Specific and non specific interaction between methanol and ethanol and active carbons. Langmuir 2000;16:5967e72. [36] Wood GO. Affinity coefficients of Polanyu/Dubinin adsorption isotherm equations. A review with compilations and correlations. Carbon 2001;39: 343e56. [37] Aristov Y, Tokarev MM, Cacciola G, Restuccia G. Heat capacity and heat conductivity of aqueous solutions of calcium chloride in silica gel pores. Russ J Phys Chem A 1997;71(3):327e30. [38] Doebelin Ernest O. Measurements systems: application and design. 4th ed. Singapore: McGraw e Hill International Editions; 1990. [39] El-Skarkawy II, Hassan M, Saha BB, Koyama S, Nasr MM. Study on adsorption of methanol onto carbon based adsorbents. Int J Refrig 2009;32:1579e86. [40] Saha BB, El-Sharkawy II, Chakraborty A, Koyama S. Study on an activated carbon fiber-ethanol adsorption chiller: part II e performance evaluation. Int J Refrig 2007;30:96e102. [41] Wang RZ, Jia JP, Zhu YH, Teng Y, Wu JY, Cheng J, Wang QB. Study on a new solid adsorption refrigeration pair: activated carbon fiber-methanol. J Sol Energy Eng 1997;119(3):214e8. [42] Dieng AO, Wang RZ. Literature review on solar adsorption technologies for ice-making and air conditioning purposes and recent developments in solar technology. Renew Sustain Energy Rev 2001;5:313e42.