Available online at www.sciencedirect.

com

ScienceDirect
Physics Procedia 67 (2015) 1199 – 1205

25th International Cryogenic Engineering Conference and the International Cryogenic Materials
Conference in 2014, ICEC 25–ICMC 2014

Activated carbon-hydrogen based continuous sorption cooling in

single adsorbent bed with LN2 heat sink
Susmita Koley*, Indranil Ghosh
CryogenicEngineering Center, IIT Kharagpur, Kharagpur-721302, India

Abstract

Quick and periodic inflow-outflow of adsorbate in an adsorbent column creates a differential temperature between the two ends
of it, allowing for the generation of continuous sorption cooling in a single adsorbent tube. The concept has been proven
experimentally and theoretically for near room temperature applications using activated carbon-nitrogen. The feasibility of
generating continuous solid sorption cooling in a single adsorbent tube in the cryogenic domain has been studied theoretically
with a different adsorbent-adsorbate pair, namely, activated carbon-hydrogen. Precooling of gaseous hydrogen (before it enters
the adsorbent column) and removal of the heat of adsorption has been achieved using liquid nitrogen. Theoretical estimation
shows nearly 20 K temperature difference between the two ends under no load condition. Finally, parametric variations have
been performed.
©
© 2015
2014TheTheAuthors. Published
Authors. by by
Published Elsevier B.V.B.V.
Elsevier This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of ICEC 25-ICMC 2014.
Peer-review under responsibility of the organizing committee of ICEC 25-ICMC 2014
Keywords:continuous cooling; adsorben/adsorbate pair; activated carbon bed

1. Introduction

The use of vibration-free solid sorption cooling in space cryogenics has been proved successful for ages (Jones
and Golben, 1985; Prakash, et al., 2000; Bhandari, et al., 2004; Borders, et al., 2004). Periodically adsorbed (or
desorbed) gas to (or from) solid adsorbent particles generate a pressure swing required in Linde-Hampson cycle
based liquefaction processes. The role of adsorbent beds in the heat driven process is that of a compressor as

* Corresponding author. Tel.: 08145362186.

E-mail address:[email protected]

1875-3892 © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of ICEC 25-ICMC 2014
doi:10.1016/j.phpro.2015.06.189
1200 Susmita Koley and Indranil Ghosh / Physics Procedia 67 (2015) 1199 – 1205

mentioned in Sun, et al. (1996) and Olivera (2011). Alternatively, Leppard and Leslie (1980) have shown that
adsorbent beds can also be used as condenser or evaporator, while a mechanical compressor generates the required
pressure swing. The cooling power in a ‘thermal compressor’ driven system is due to evaporation of the refrigerant.

Nomenclature

A1 Internal cross-sectional area of the container (m2)

A2 Area of the annular section (m2)
Cp Heat capacity of adsorptive (J/kg-K)
Cps Heat capacity of activated carbon (J/kg-K)
Cpw Heat capacity of wall (J/kg-K)
Dax Axial dispersion coefficient (m2/s)
heff Effective heat transfer coefficient between bed and surrounding (watt/m2-K)
ǻH Heat of adsorption/desorption (J/mol)
kb Bed thermal conductivity (W/m-K)
kwall Wall thermal conductivity (W/m-K)
Mg Molecular weight of adsorptive (kg/kmol)
q (x,t) Gas adsorbed at any instant of time (kmol/kg)
R Universal gas constant (J/kg-K)
T Temperature (K)
t Time (s)
u Interstitial velocity (m/s)
Greek symbols
‫ܭ‬ Void Fraction
ȡ Density of gas (kg/m3)
ȡs Density of adsorbent (kg/m3)
ȡw Density of wall material (kg/m3)

The cooling power in the ‘mechanical compressor’ driven system is because of desorption of the gases from the
adsorbent (Sward and LeVan, 1999). Application of the second type of cooling process in cryogenics is rare. Use of
multiple beds and alternatively connecting them to heat sink and heat source enhances operational hazards and
reduces thermal performance. Recently, a new concept has been proposed for creating a temperature differential
between the two ends of a single adsorbent column with the quick and sequential inflow and outflow of adsorbate
through the bed (Koley and Ghosh, 2013a). This sorption cycle is fundamentally different from pulse tube operation
in terms of their operating frequencies and the origin of cooling. The cooling in ‘single bed’ is due to the heat of
desorption, whereas, the expansion of gaseous helium produces cryogenic refrigeration in a pulse tube.
The proposed sorption cooling cycle is a combination of two half cycles, namely, adsorption and desorption
(Koley and Ghosh, 2013a). Firstly, pressurization (adsorption) of the bed is accomplished by allowing the
compressed adsorbate from the compressor to enter into the bed through an inlet solenoid valve. During this process
the other solenoid valve (outlet valve) remains closed. In the second step, closure of the inlet valve and opening of
the outlet valve results in depressurization (desorption) of the bed. These two processes occurring in quick
succession create a temperature gradient in the bed. Consequently, the end where gas enters or leaves the bed
remains cooler than the other end. Other than the two solenoid valves, there is an orifice or small opening at the hot
end. The orifice is kept open throughout the whole process enabling the creation of a pressure gradient between the
two ends of the column. Koley and Ghosh (2014) in another work have shown that incorporating this orifice opening
increases the magnitude of the temperature drop between the hot and cold ends.
While different adsorbent-adsorbate combinations are suitable for different temperature zones, theoretical
investigations involving the activated carbon-hydrogen pair have been made for generating refrigeration around the
normal boiling point of liquid nitrogen. Cooling obtained in this process depends on different operating and
geometrical parameters related to the system. Operating conditions such as charging pressure, degree of orifice
 Susmita Koley and Indranil Ghosh / Physics Procedia 67 (2015) 1199 – 1205 1201

opening, heat-sink temperature have significant effect on the cold end temperature. Use of different activated carbon
types as adsorbent has an effect on the cold-end temperature. Theoretical studies involving activated carbon and
hydrogen as adsorbent-adsorbate pair are described in the following section assuming liquid nitrogen as the coolant
for removing the heat of adsorption isothermally.

2. Methodology

In order to investigate the feasibility of generating cooling around the normal boiling point of liquid nitrogen
(~77 K), a continuous solid sorption cooling cycle has been simulated theoretically. The choice of adsorbate for
generating cooling near 77 K is limited. Hydrogen has been chosen as adsorbate so as to investigate its use in
combination with activated carbon adsorbent. The effect of different operating conditions and some of the system
parameters have been examined theoretically.
The proposed sorption cooling process (Koley and Ghosh, 2013a) has been modified slightly as shown in Fig. 1.
Liquid nitrogen precooling of the gaseous hydrogen has been introduced before entering the sorption bed. Liquid
nitrogen is taken from the same storage for removing the heat of adsorption at the hot end.

LN2 7

4 5 6
3

1 QC QH

1 Compressor 5 Adsorbent Tube

2 Buffer 6 Hot End Heat Exchanger
3 Solenoid Valve 7 Orifice
4 Cold End Heat Exchanger
2

Fig. 1. schematic of the physical model of the activated carbon bed.

Experimental and theoretical details of the cooling process in activated carbon-nitrogen system have been
described in an earlier work (Koley and Ghosh, 2014). These studies were limited to the non-cryogenic temperature
domain (around room temperature). In order to study the suitability of this process in the realm of cryogenic
temperature, a theoretical investigation has been envisaged with the hydrogen-activated carbon (adsorbate-
adsorbent) combination, assuming that liquid nitrogen is available as coolant for removing heat of adsorption. The
simulation has been restricted within the adsorbent bed, presuming continuous supply of compressed hydrogen at
77 K to the column. In the following paragraphs, a brief summary of the theoretical model used in this study is
described.
Transient mass and energy balance equations, Eqs. (1) and (2) have been used for finding out the density and
temperature profiles in the bed:

డఘ డ డఘ డ௤ሺ௫ǡ௧ሻ
ߝ డ௧ ൅ ߝ డ௫ ቂ‫ ߩݑ‬െ ‫ܦ‬௔௫ డ௫ ቃ ൅ ሺͳ െ ߝሻߩ௦ ‫ܯ‬௚ ൌ Ͳ, (1)
డ௧

డ் డ௤ሺ௫ǡ௧ሻ డ డ்
ൣߝ‫ܣ‬ଵ ߩ‫ܥ‬௣ ൅ ሺͳ െ ߝሻ‫ܣ‬ଵ ߩ௦ ‫ܥ‬௣௦ ൅ ߝ‫ܣ‬ଶ ߩ௪ ‫ܥ‬௣௪ െ ߝ‫ܣ‬ଵ ܴߩ൧ ቀ ቁ െ ሺͳ െ ߝሻ‫ܣ‬ଵ ο‫ߩܪ‬௦ ൌ ቂ‫ܣ‬ଵ ݇௕ ൅
డ௧ డ௧ డ௫ డ௫
డ்
‫ܣ‬ଶ ݇௪௔௟௟ െ ߝ‫ܣ‬ଵ ߩ‫ܥ‬௣ ‫ܶݑ‬ቃ െ  ݄௘௙௙ ሺߨ‫ܦ‬଴ ሻሺܶ െ ܶ଴ ሻ. (2)
డ௫
1202 Susmita Koley and Indranil Ghosh / Physics Procedia 67 (2015) 1199 – 1205

The terms in the Eq. (1) from left side respectively represent the amount of accumulated gas in the intra-particle
space, the convective mass flux including axial dispersion of gas and the amount of adsorbed gas. In the energy
balance equation, the coefficient of the partial time derivative of temperature includes four terms which respectively
denote the heat accumulation in the gas phase, in the adsorbent and in the wall and heat generation due to
compression of gas. The second term in the left hand side of Eq. (2) signifies the heat liberated due to exothermic
adsorption process. The spatial derivative in the right hand side contains conduction heat flux in the bed as well as in
the wall and the convective heat flux. Thermal interaction between adsorbent column and surroundings has been
taken into account through the last term in Eq. (2). The initial and boundary conditions and other relevant
expressions (like rate equations etc.) are omitted in the present manuscript. Details can be obtained from the earlier
work (Koley and Ghosh, 2014). The governing transient mass and energy balance equations Eqs. (1) and (2) have
been discretized using the Euler implicit method and central difference scheme. The resulting equations have been
solved by writing a C++ program. Validation of the theoretical predictions and experimental results has been
performed (Koley and Ghosh, 2014). Finally, the adsorption and desorption models have been coupled together for
simulating the sorption cycle.

3. Results and discussion

Desirable features primarily looked for in adsorbents to be used are the high equilibrium adsorption capacity and
the heat of adsorption/desorption (Koley and Ghosh, 2013b). In the present analysis, three different activated carbon
samples, characterised and reported in the literature, have been selected (Chahine, et al., 2001; Zheng. et al., 2014
and Delahaye, et al., 2002). At the onset, the equilibrium adsorption capacity (at 77 K) multiplied by the
corresponding heat of adsorption have been plotted as a function of charging pressure (Fig. 2). The heat of
adsorption for the activated carbons of Chahine, et al. (2001), Zheng, et al. (2014) and Delahaye, et al. (2002) have
been taken as 6.4 kJ/mol, 6.4 kJ/mol and 3.9 kJ/mol respectively. Fig. 2 indicates that the carbon samples would
result in different degrees of cooling in the different operating pressures zones. Beyond 2.3 MPa operating pressure,
it is expected to obtain more cooling with the activated carbon of Chahine, et al. (2001) than the same generated with
the activated carbon of Zheng, et al. (2014) or Delahaye, et al. (2002). Fig. 3 indicates the correctness of this
hypothesis. Lowest cold end temperature produced with different activated carbons, when plotted as a function of
the operating pressure (Fig. 3) shows that the activated carbon of Chahine, et al. (2001) generates more cooling at
pressure higher than 2 MPa as compared to the other carbons. The smaller value of the product term for activated
carbon of Delahaye, et al. (2002) justifies the insignificant cooling obtained with it. In the following studies
involving parametric variations, the activated carbon of Chahine, et al. (2001) has been considered.

Fig. 2. Product of equilibrium adsorption capacity and heat of Fig. 3. Effect of using different activated carbons on cold
adsorption for three different activated carbons. end temperature.
 Susmita Koley and Indranil Ghosh / Physics Procedia 67 (2015) 1199 – 1205 1203

Fig. 4. Temperature and pressure profile of activated carbon bed at 3 MPa pressure.

Lowering the temperature at the cold end of the adsorbent tube depends on multiple variables including operating
conditions and the geometrical parameters. The effects of some of the important ones are described in the following
section. Simulation of typical time-temperature-pressure variations at the two ends of the adsorbent column for
5 s adsorption and 5 s desorption cycle is shown in Fig. 4. Results have been simulated up to 200 s. The bed has been
pressurized up to 3 MPa during adsorption. In the other half of the cycle, the bed is allowed to desorb down to
0.1 MPa. Nevertheless, a small difference in pressure between the hot and cold ends indicates that the orifice valve is
kept open slightly allowing some of the compressed gas to flow through the bed. Starting from an initial temperature
of 77 K, the cold end temperature inside the bed has been found to reach a minimum of 57 K.
With the increase in the operating pressure from 2 to 4MPa, decrease in the minimum cold end temperature has
been observed (Fig. 5). However, the minimum cold end temperature cannot be lowered indefinitely with the
increase in pressure when the rate of adsorption attains saturation.
It has been observed (Koley and Ghosh, 2013a; Koley and Ghosh, 2014) that orifice opening has a significant
effect on the cold end temperature. The change in orifice opening in the present study has been simulated using
different valve flow coefficient (Cv). Fig. 6 depicts the pressure variation of the cold end temperature for three
different valve flow coefficients keeping the entry and exit solenoid valve flow coefficients constant. Larger Cv
values implying larger opening of the orifice help reducing the cold end temperature up to a certain value of the Cv,
beyond which, the decrease in cold end temperature becomes insignificant. The effects of the entry and exit solenoid
valve flow coefficients become important. It may also be noted that at higher pressure, lowering of the cold end
temperature is independent of the valve flow coefficient.
A notable change in the temperature profile can be seen if the cooling provided by liquid nitrogen is removed. In
absence of liquid nitrogen, the adsorbate (gaseous hydrogen) enters the bed at ambient temperature (assumed to be at
300 K) and the heat rejection at the hot end also takes place around 300 K. Under these circumstances, the resulting
temperature profile between the two ends (shown in Fig. 7) reveals hardly any temperature drop as compared to the
situation when nitrogen precooling and a liquid nitrogen heat sink is used. The different degree of cooling obtained
at the different temperature zones can be explained by the second term on the left hand side of the energy balance
equation (Eq. 2). This term contains the adsorption rate and the heat of adsorption for the adsorbent/adsorbate pair.
Adsorption capacity is a strong function of temperature and pressure. When the isotherms at 77 K and 300 K,
multiplied by the corresponding heat of adsorption, are plotted as a function of pressure, one can observe a large
difference between the two (Fig. 8). It is due to this wide difference in the resulting product (of sorption rate and heat
of adsorption) that activated carbon-hydrogen is not a suitable pair for near room temperature application (around
300 K), while the same combination can be effective at the liquid nitrogen temperature region (~77 K).
1204 Susmita Koley and Indranil Ghosh / Physics Procedia 67 (2015) 1199 – 1205

Fig. 5. Effect of operating pressure on cold end temperature. Fig. 6. Effect of orifice opening on cold end temperature.

Fig. 7. Temperature profile in activated carbon bed at 77 K and 300 K. Fig. 8. Product of equilibrium adsorption capacity and heat of
adsorption for 77 K and 300 K.

4. Conclusion

Theoretical study reveals that a mechanical compressor driven solid sorption cooling can be generated
continuously in activated a carbon-hydrogen based system with the removal of heat of adsorption at liquid nitrogen
temperature.
While this adsorbent-adsorbate pair is not suitable for cold production near room temperature applications, this
pair has been found effective at low temperature. The term defined by the product of equilibrium adsorption capacity
and heat of adsorption determines the production of temperature drop at different conditions.
Higher operating pressure creates larger temperature drop due to high adsorption rate of adsorptive at elevated
pressure. A larger opening of the orifice increases the cold generation, this is limited by the rate of entry of gas
through the valve at entrance.
 Susmita Koley and Indranil Ghosh / Physics Procedia 67 (2015) 1199 – 1205 1205

References

Agarwal A., Biegler L. T., Zitney S. E., 2009. Simulation and optimization of pressure swing adsorption systems using reduced order modeling.
Industrial and Engineering Chemistry Research; 48, 2327-2343.
Alpay E., Scott D. M., 1992. The linear driving force model for fast-cycle adsorption and desorption in a spherical particle, Chemical Engineering
Science 47, 499-502.
Benard P., Chahine R., 2001. Determination of adsorption isotherm of hydrogen on activated carbons above the critical temperature of the
adsorbate over wide temperatures and pressure ranges. Langmuir 17, 1950-1955.
Bhandari P., Prina M., Bowman Jr R. C., Paine C., Pearson D., Nash A., 2004. Sorption coolers using continuous cycles to produce 20K for the
Planck flight mission. Cryogenics 44, 395-401.
Borders J., Morgante G., Prina M., Pearson D., Bhandari P., 2004. Optimized autonomous operations of a 20K space hydrogen cryocooler.
Cryogenics 44, 565-573.
Delahaye A., Aoufi A., Gicquel A. Pentchev I., 2002. Improvement of hydrogen storage by adsorption by using 2-D modeling of heat effects.
Energy and environmental engineering 48, 2061-2073.
Glueckauf E., 1955. Theory of Chromatography, Part 10 - Formula for diffusion into spheres and their application to chromatography,
Transaction of the Faraday Society 54, 1540-1551.
Jones J.A., Golben P.M., 1985. Design, life-tasting and future-designs of cryogenic hydride refrigeration systems. Cryogenics 25, 212-219.
Koley S. Ghosh I., 2013. New technique for generating continuous sorption cooling in a single adsorbent column. Applied Thermal Engineering;
55, 33-42.
Koley S., Ghosh I., 2014. Generating continuous solid sorption cooling in a single adsorbent tube – Experiment and generalised transient
analysis. International Journal of Heat and Mass Transfer; 72, 470-478.
Koley S., Ghosh I., Activated carbon based solid sorption cooling using different adsorptives in different temperature zones. Proceedings of the
22nd National and 11th International ISHMT-ASME Heat and Mass Transfer Conference. HMTC1300161.
Lamari M., Aoufi A., Malbrunot P., 2000. Thermal effects in dynamic storage of hydrogen by adsorption. Environmental and Energy
Engineering; AIChE Journal 46, 632-646.
Leppard C. J., Leslie S. 1980. Adsorption heat pump, US Patent No. 4,183,734.
Leva M., 1949. Pressure drop correlation for a single incompressible fluid. Chemical Engineering 56, 115-117.
Nakao S. I, Suzuki M., 1983. Mass transfer coefficient in cyclic adsorption and desorption, Journal of Chemical Engineering of Japan 16, 114-
119.
Olivera R. G., 2011. Solar powered sorption refrigeration and air conditioning. Refrigeration: Theory, technology and applications; ISBN: 978-1-
61668-930-8. 205-238.
Prakash M. J., Prasad M., Rastogi S. C., Akkimaradi B. S., Gupta P. P., Narayanamurthy H., Srinivasan K., 2000 Development of a laboratory
model of activated charcoal-nitrogen adsorption cryocooler. Cryogenics 40, 481-488.
Rodrigues A. E, Dias M. M., 1989. Linear driving force approximation in cyclic adsorption processes: simple results from system dynamics
based on frequency response analysis, Chemical Engineering and Process 37, 489–502.
Ruthven D. M., 1984. Principles of Adsorption and Adsorption Processes, John-Wiley & Sons, New York.
Sircar S., Hufton J. R., 2000. Why does the linear driving force model for adsorption kinetics work? Adsorption 6, 137-147.
Sun L.M., Feng Y., Pons M., 1996. Numerical investigation of adsorptive heat pump systems with thermal wave heat regeneration under uniform
pressure condition. International Journal of Heat and Mass Transfer 40, 281-293.
Sward B. K., LeVan M. D. 1999. Examination of the performance of a compression-driven adsorption cooling cycle. Applied Thermal
Engineering 19, 1-20.
Zheng Q., Wang X., Gao S., 2014. Adsorption equilibrium of hydrogen on graphene sheets and activated carbon. Cryogenics 61, 143-148.