Yifan Yang,
Dipak Rana* and
Christopher Q. Lan*
University of Ottawa, Department of Chemical and Biological Engineering, 161 Louis Pasteur Private, Ottawa, Ontario K1N 6N5, Canada. E-mail: rana@uottawa.ca; clan@uottawa.ca
First published on 24th June 2015
A solid super desiccant was developed based on the concept of impregnating a super liquid hygroscopic agent in a superabsorbent polymer to take advantage of the vast vapour absorption capacity of the hygroscopic agent and the liquid withholding capacity of the superabsorbent polymer. The maximum hygroscopic salt loading was determined based on different salt concentrations in sodium polyacrylate. The newly developed desiccant showed a vapor absorption capacity of 276 ± 20%, which is superior in comparison to that of conventional solid desiccants such as silica gel, activated alumina and molecular sieves. Thus they have great potential in applications requiring solid desiccants such as desiccant cooling, air conditioning, storage of moisture sensitive materials, and dehydration of natural gas.
Nevertheless, liquid desiccants, mostly lithium chloride solutions, are not applicable in many scenarios. For instance, desiccant cooling is designed for personal heat mitigation. A safer desiccant is required, specifically toxic-free. Therefore, solid desiccants are preferred due to their dryness, and simplicity of their structure.5 Solid desiccants are favoured in many other scenarios over liquid ones for their ease of operation, simplicity of device, and chemical inertness.6 In contrast, liquid desiccants are usually corrosive, difficult to handle, and require complicated systems.
Most commonly encountered conventional solid desiccants are silica gel, activated alumina, molecular sieves, and activated carbon. Further, certain synthetic polymers have also been recognized as great desiccant options. For instance, various polymers and proteins were studied in terms of water vapor adsorption at a temperature of 25 °C, and the carboxyl groups presented a strong affinity with the adsorbed water.7,8
Unlike liquid desiccants, which absorb water molecules into the entire volume of the solution, solid desiccant relies on the adsorption of vapor molecules onto their surface to retain moisture. As a result, solid desiccants are inherently of low vapor adsorption capacity in comparison to liquid desiccants.
In the 1960's, researchers developed cross-linked polyacrylamide which can absorb liquid water up to 15–75 times its own weight and retained the water under pressure at about 2.5 psi. In the 1970's, the Agricultural Research Administration of US (Northern Regional Res. Lab. (NRRL), Bureau of Agricultural and Industrial Chemistry, Peoria, IL) developed a new material that can absorb more than 103 times its weight in water and was called a super absorbent polymer (SAP). Since then, this type of polymer, which has great water retention ability, has been widely used in disposable diapers and in agriculture industries.
We hypothesized that the superb liquid holding capacity of SAP, could be exploited in combination with superior vapour absorption capacity of liquid desiccants to create a super solid desiccant. Since sodium polyacrylate has been proven to be a great water retention agent, and also a decent moisture absorption material, further optimization in order to develop a super solid desiccant seems feasible theoretically. With the motivation of developing a super solid desiccant that is highly efficient, non-corrosive and easy to regenerate, we have proposed the use of a superabsorbent hydrogel composite (SHC), a material that is based on sodium polyacrylate, with the impregnation of a hydroscopic agent (lithium chloride), as a novel solid desiccant material. SHC takes advantages of both, the great absorption capacity of lithium chloride and the large liquid-holding property of SAP. Moreover, commercial sodium polyacrylate is available at around $10 per kilogram, which is comparable in price to the commercial silica gel. This newly developed desiccant has shown big potential in applications in oilfields, HVAC industries, etc.
![]() | ||
Fig. 3 SAP swelling capacities and LiCl retained amount with respect to the molarity of LiCl solution. |
Fig. 3 also shows that a 2.5 M LiCl solution led to the largest LiCl retention in SAP, which can be potentially beneficial because of the larger moisture retention for the resulting SHC desiccant. The existence of a LiCl concentration at which a maximum LiCl retention was obtained was due to the fact that the amount of LiCl retained in the gel, which equals the amount of solution absorbed at equilibrium multiplies the solution concentration. When low concentrations of LiCl solution was used for swelling, little LiCl was contained in the relatively large volume of solution absorbed. On the other hand, when the LiCl concentration was high (i.e., 3 M or higher in this study), the reduced volume of solution absorbed by the SAP led to less LiCl retention despite of the relatively high LiCl concentration.
The swelling behaviour of ionic hydrogel such as SAP is a result of a balance between the elastic energy of the SAP network and the osmotic pressure difference between the liquid phase inside the gel matrices and that outside of it. The elastic term is described by the Flory expression derived from an assumption of Gaussian chain distributions, and in an electrolytic solution, the osmotic pressure is associated with the molarity of the solution and temperature.
Liquid absorption by polymer can be described by the Flory theory for an ionic network.10
The osmotic pressure dependency is through i/S, which is the electronic charge on the polymer structure per polymer unit versus ionic strength of solution. When the solution is more concentrated or solution contains ion with more charges, the swelling capacity of the particular polymer decreased due to raised ionic strength of the solution. The affinity of water and polymer in this case is referring to the water affinity of the hydrophilic group on the polymer chain, more specifically carboxylate and carboxylic acid groups in this system (as shown in Fig. 1).
By recognizing the mechanism of hydrogel swelling, based on the same polymer structure, the affinity of the water and the polymer and cross-linking density of the network remain the same. The swelling capacity is directly affected by the osmotic pressure. Thus, decreasing the ionic strength, more specifically, decreasing the concentration of the LiCl solution for this system, can increase the polymer swelling capacity (liquid retention). We can see that there is a trade-off for an higher amount of liquid retention which requires a low solution concentration and more LiCl amount trapped inside of SAP which generally prefers a higher solution concentration. Fig. 3 has shown the optimum solution concentration for the maximum LiCl retention within SHC.
Vapor absorption experiments on the different loading of SHC were tested with a sample preparation based on a 1 time of solution/SAP equilibrium. As shown in Fig. 4, 2.5 M SHC has the largest 24 h moisture absorption capacity compared to the other loading of SHC. The results followed the same trend as the LiCl solution retained in the SAP gel in Fig. 4, meaning that the larger the amount of LiCl absorbed on the polymer, the greater the moisture absorption ability of the final SHC. The retained LiCl solution in the gel became a salt crystal after drying and staying on the SHC surface. The crystal acts like the separate moisture absorbent on the polymer, which further increases the capacity.
![]() | ||
Fig. 4 24 h absorption capacity versus molarity of LiCl solution for SHC preparation (1 h equilibrium duration). |
It is worth noting that more LiCl salt outside of the polymer did not always positively affect the SHC performance. When large amounts of free lithium chloride crystals absorb moisture, a layer of concentrated LiCl solution film is slowly formed around the SHC, and this is hardly absorbed by the SHC due to its high osmotic pressure. This layer of free LiCl solution becomes more and more diluted and corrosive, and is harmful to the equipment. Therefore a rinsing step as the last step of SHC preparation is important to ensure that the resulting desiccant is non-corrosive and therefore ease of operation.
From Fig. 4, it can also be observed that the absorption capacity increased slightly from 5 M SHC to 10 M SHC. This could imply that the ion exchange degree also affected moisture absorption ability. A 10 M LiCl solution, which is an almost saturated LiCl solution, can exchange the largest amount of Li+ onto the polymer chain at equilibrium conditions compared to a lower concentration solution. However, for 10 M SHC, it retained the least amount of LiCl solution within the polymer structure during SHC formation. Overall, the active lithium ions or lithium chloride salts on the 10 M SHC is still less compared to the others (especially 2.5 M SHC). For similar reasons, the 1 and 5 M SHC, either have a lesser degree of ion exchange or they retained less amounts of LiCl crystal on the polymer, resulting in a less moisture absorption performance.
The 2.5 M SHC shows the best performance with Fig. 4 revealing the impact of ion exchange degree on the moisture absorption capacity. This led to the concerns about ion exchange rate limiting factors, with these two factors being studied. One is equilibrium duration and the other is number of equilibrium.
![]() | ||
Fig. 6 Vapor absorption performance with 2.5 M SHC prepared by different numbers of ion-exchange equilibrium. |
SAP vapor absorption data was also plotted as a baseline for comparison. The 24 h absorption for SAP without further treatment is measured as 1.31 g of vapor per gram of SAP. For different numbers of equilibrium preparation, the higher the equilibrium number, the higher the vapor absorption capacity. The slope for each type of SHC increases as well when a higher number of equilibrium is achieved. It is interesting to note that the 3-equilibrium SHC has a similar vapor absorption curve to that for 5-equilibrium SHC, meaning that the ion exchange might be completed by the end of 3-equilibrium. This hypothesis needs to be further proven by ICP-ES results.
The relative ion exchange degree is calculated based on the ICP-ES results. The sodium ion amount from the leachates of different equilibrium runs was summarized to calculate the total ion exchange capacity for sodium polyacrylate.
The relative ion exchange degrees were then calculated by using the sodium ion amount from each batch of equilibrium divided with the total ion exchange capacity. The calculation is summarized in the following equation:
From the grey curve shown in Fig. 7, we can see that the ion exchange degree reaches a plateau after the 3rd equilibrium, reaching 97.2% of ion exchange degree at the 3rd equilibrium. This trend is maintained for the 24 h moisture absorption tests from different equilibrium degrees of SHC samples. The 5th equilibrium showed the same 24 h moisture absorption capacity as that of the 3rd equilibrium SHC, and these are the largest results of all 5 tests. The sodium polyacrylate sample was tested at the same condition as the control. The reason for this is because more numbers of equilibrium pushed the ion exchange to the Li+ favour direction on the polymer, and more Li+ on the polymer chain can help attract more vapor, resulting in a better desiccant performance.
![]() | ||
Fig. 7 24 h vapor absorption capacity and ion exchange degree with respective to the numbers of equilibrium. |
The morphologies of the SHC under different equilibrium conditions were measured by SEM-EDX. Fig. 8 lists the SEM images for 2.5 M 1-equilibrium, 2.5 M 3-equilibrium and 2.5 M 5-equilibrium SHC at a magnification of 60×. These images clearly showed that the amount of salt crystals (white dots) on the polymer surface decreased when the number of equilibrium increased. To better understand this phenomenon, an elemental analysis was performed by EDX to confirm the crystal content. The EDX results are listed in Fig. 9.
![]() | ||
Fig. 8 SEM images for (A) 2.5 M 1-eq., (B) 2.5 M 3-eq. SHC and (C) 2.5 M 5-eq. SHC (magnification 60×). |
The top images of Fig. 9 are the SEM images for the corresponding SHC at a magnification of 5000×. The actual detecting spots are highlighted in the images, and the bottom images are the EDX peaks showing elements detected from the selected spot. It was found that most of the crystal shown in Fig. 9A was made of sodium chloride, which had a perfect cubic crystal shape. The crystals found on Fig. 9B and C are mainly made of lithium chloride. It is noted that lithium is not detectable by EDX due to its small atomic size. However, the peaks show high levels of chlorine and zero cation. Based on the fact that only lithium and sodium ion exist in the system, it is therefore concluded that the crystal is made of lithium chloride.
Another interesting fact from the SEM images shown in Fig. 9, is that the shape of the crystals found on the surface of 3-equilbirium (Fig. 9B) and 5-equilibrium (Fig. 9C) SHC are not the perfect cubic shape as those found in 1-equilibrium (Fig. 9A). These deformed crystals are all detected as lithium chloride while those cubic shaped crystals found in the Fig. 9A are sodium chloride. The deformation of lithium chloride crystals may be due to the hygroscopic nature of lithium chloride salt. It has a higher affinity for moisture compared with sodium chloride. Therefore, under the same sample preparation period before the SEM measurement, the lithium chloride attracted slight amount of moisture and slowly deformed. This also explains another observation from Fig. 9, that it seems to have more crystals on Fig. 9B comparing to Fig. 9C, this doesn't means that 3-equilibrium has more lithium chloride crystals than 5-equlibrium, it is again due to the trace amount of sodium chloride on the surface of Fig. 9B (cubic shaped crystals found on the bottom right of Fig. 9B). The observed less crystals on Fig. 9C surface is due to the fast moisture capture from the lithium chloride crystal during sample preparation, the ICP-ES results have confirmed that 5-equilibrium has more completed sodium removal comparing to 3-equilibrium, and therefore Fig. 9C should contain slightly more lithium chloride crystal comparing to Fig. 9B. This phenomenon also leads to another conclusion that by performing more equilibrium steps, less sodium ion exists within the system. This leaves more space for lithium chloride crystals to be attached on the polymer surface, thus further improving the vapor absorption capacity.
To better evaluate the proposed solid desiccant, the physical properties and performances of a few common solid desiccants were researched in literature. Silica gel, a generic name for a gel manufactured from sulphuric acid and sodium silicate, has a high water capacity in that it can adsorb up to 45% of its own weight in water. It is easier to regenerate and it costs less than a molecular sieve. Activated alumina, a hydrated form of aluminium oxide (Al2O3), is the least expensive adsorbent for gas dehydration, can produce a dew point below −105.56 °C if applied properly and less heat is required to regenerate than molecular sieves. However, molecular sieves give lower outlet water dew points. Molecular sieves are the most versatile adsorbent because they can be manufactured for a specific pore size, depending on the application and are capable of dehydration for less than 0.1 ppm water content. It is the most expensive adsorbent but offers greater dehydration and requires higher temperatures for regeneration. These are alkaline and are subject to attack by acids. Table 1 has listed the common desiccant physical properties.
Silica gel | Activated alumina | Molecular sieves | SAPb | SHCb | |
---|---|---|---|---|---|
a Data reported from Sigma-Aldrich.b Data were obtained from this study. | |||||
Surface areaa (m2 g−1) | 600–800 | 200–300 | 600–700 | 3.5 ± 0.06 | 4.6 ± 0.08 |
Pore diameter13 (Å) | 10–90 | 15 | 3, 4, 5, 10 | Non-porous | Non-porous |
Static adsorption at RH 100% (ref. 14) | 38% | 36–42% | 23% | 131 ± 27% | 276 ± 20% |
Regeneration temperature15 (°C) | 100 | 120–260 | 175–315 | N/A | <80 |
The cost of commercial sodium polyacrylate is around $10 per kilogram, which is comparable to the commercial silica gel price, considering that the lithium chloride chemical is more expensive, a more detailed cost analysis was performed. By performing ICP-ES tests, we measured the amount of Na in the leachate solution, the total amount of Na removed from SAP is 0.85 mmol per gram of SAP, and this amount of lithium ion is expected to be on the polymer chain. The amount of LiCl obtained by the gel swelling shown in Fig. 3 (2.5 M) is 34.6 mmol per gram of SAP that is 1.5 g of LiCl per g of SAP, this amount can be considered as the amount of LiCl crystals trapped inside of SAP. The cost of the analytical grade of lithium chloride purchased from Sigma-Aldrich Inc. is $0.15 per gram ($368 per bottle of 2.5 kg), this chemical cost can be greatly reduced when approaching it to a non-analytical grade source. Therefore by this proposed modification on sodium polyacrylate, the resulted SHC has doubled the vapor absorption capacity with reasonable amount of cost during preparation.
It is noted that adding LiCl can double the vapor absorption of SAP, which is significant considering that the conventional solid desiccant has the capacity only around 40% of its own weight (shown in Table 1). Furthermore, the vapor absorption capacity reported is 24 h absorption instead of equilibrium capacity. As shown in Fig. 6, the new desiccant was still absorbing vapour after 24 h and the gap between the desiccants and the SAP would become much larger if we push the vapour absorption to equilibrium. The much larger rate of vapour absorption by the desiccant than the SAP provides another important advantage.
There is an important concern in terms of potential LiCl leaching, since this new developed solid desiccant has great potential in various health related applications, for example, life support vacuum desiccant cooling device,16 in which three layers structures including water reservoir, spacer and solid desiccant layer are encapsulated inside of an air-tight bag, water evaporation which can be initiated through vacuuming provides cooling. LiCl solution leaching from desiccant can potentially damage the device and eventually irritates the body skin. Other applications like the compact space cooling unit, also has similar safety concern. Thus solid desiccants are required to not only be efficient but also safe to handle (non-corrosive). The rinsing step as the last step for SHC gel preparation is designed to remove all the freest LiCl outside of SHC gel so that the desiccant will stay solid (gel) with the multiple absorption–desorption cycles, and therefore is safe to the environment and equipment.
This journal is © The Royal Society of Chemistry 2015 |