This study describes a method to expand chitin into a foam by chemical techniques that require no specialized equipment.
Chitin is an underexploited, naturally abundant, mechanically robust, and chemically resistant biopolymer. These qualities are desirable in an adsorbent, but chitin lacks the necessary specific surface area, and its modification involves specialized techniques and equipment. Herein is described a novel chemical procedure for expanding chitin flakes, derived from shrimp shell waste, into foams with higher surface area. The process relies on the evolution of H2 gas from the reaction of water with NaH trapped in a chitin gel. The preparation method requires no specialized equipment. Powder X-ray diffraction and N2-physisorption indicate that the crystallite size decreases from 6.6 nm to 4.4 nm and the specific surface area increases from 12.6 ± 2.1 m2/g to 73.9 ± 0.2 m2/g. However, infrared spectroscopy and thermogravimetric analysis indicate that the process does not change the chemical identity of the chitin. The specific Cu adsorption capacity of the expanded chitin increases in proportion to specific surface area from 13.8 ± 2.9 mg/g to 73.1 ± 2.0 mg/g. However, the Cu adsorption capacity as a surface density remains relatively constant at an average of 10.1 ± 0.8 atom/nm2, which again suggests no change in the chemical identity of the chitin. This method offers the means to transform chitin into a higher surface area material without sacrificing its desirable properties. Although the chitin foam is described here as an adsorbent, it can be envisioned as a catalyst support, thermal insulator, and structural material.
Chitin is a mechanically robust and chemically inert biopolymer, second only to cellulose in natural abundance1. It is the major component in the exoskeleton of arthropods and in the cell walls of fungi and yeast2. Chitin is similar to cellulose, but with one hydroxyl group of each monomer replaced with an acetyl amine group (Figure 1A,B). This difference increases the strength of hydrogen bonding between adjacent polymer chains and gives chitin its characteristic structural resilience and chemical inertness2,3. Due to its properties and abundance, chitin has attracted significant industrial and academic interest. It has been studied as a scaffold for tissue growth4,5,6, as a component in composite materials7,8,9,10,11, and as a support for adsorbents and catalysts11,12,13,14. Its chemical stability, in particular, makes chitin attractive for adsorption applications that involve conditions inhospitable to common adsorbents14. In addition, the abundance of amine groups make chitin an effective adsorbent for metal ions15. However, the protonation of the amine groups under acidic conditions reduces the metal adsorption capacity of chitin16. A successful strategy is to introduce adsorption sites more resistant to protonation17,18. Instead, herein is described a simple method to increase the specific surface area and, therefore, the number of adsorption sites in chitin.
Figure 1. Chemical structure. (A) cellulose, (B) chitin, (C) chitosan. Please click here to view a larger version of this figure.
In spite of its many potential uses, chitin is underutilized. Chitin processing is challenging due to its low solubility in most solvents. A key limitation to its use in catalysis and adsorption is its low specific surface area. While typical carbon and metal oxide supports have specific surface areas in the order 102-103 m2/g, commercial chitin flakes have surface areas in the order of 10 m2/g19,20,21. Methods to expand chitin into foams exist, but they invariably rely on high temperature and pressure, strong acids and bases, or specialized equipment that represent a significant entry barrier5,21,22,23,24,25. In addition, these methods tend to deacetylate chitin to form chitosan (Figure 1C)-a more soluble and reactive biopolymer5,25,26.
Herein, a method is described to expand chitin into solid foams, increase its specific surface area and adsorption capacity, and maintain its chemical integrity. The method relies on the rapid evolution of gas from within a chitin gel and requires no specialized equipment. The increased adsorption capacity of the expanded chitin is demonstrated with aqueous Cu2+-a common contaminant in the local groundwater26.
Unit | Neat Flake | Baked Foam | Lyophilised Foam | |
Crystallinity | % | 88 | 74 | 58 |
Crystal size | nm | 6.5 | 4.4 | 4.4 |
Surface Area | m2/g | 12.6 ± 2.1 | 43.1 ± 0.2 | 73.9 ± 0.2 |
Cu Uptake | mg/g | 13.8 ± 2.9 | 48.6 ± 1.9 | 73.1 ± 2.0 |
Cu Uptake | atom/nm2 | 10.5 ± 2.8 | 10.7 ± 0.4 | 9.4 ± 0.3 |
Table 1. Summary of material properties. Chitin foams have lower crystallinity and crystal size relative to neat chitin flakes. However, the specific surface area and Cu uptake of the chitin foams are proportionally higher than that of the neat chitin flakes.
1. Preparation of expanded chitin
Figure 2. Preparation of expanded chitin foam. (A) The initial chitin in LiCl/DMAc solution. (B) The addition of the NaH/DMAc slurry. (C) The chitin foam after addition of water. (D) The chitin foam as extracted from the reaction flask. (E) The chitin foam during washing with water. Please click here to view a larger version of this figure.
2. Development of the adsorption isotherms
Expanded chitin shows the same morphology regardless of the drying method. Figure 3 shows images of neat chitin flakes (Figure 3A1), oven-dried expanded chitin (Figure 3B1), and lyophilized expanded chitin (Figure 3C3). While the neat flakes have the appearance of coarse sand, the expanded chitin foam has the appearance of a kernel of popped corn. Scanning electron micrographs show a similar change at smaller scales. While the neat chitin flakes (Figure 3A2, 3A3) have a compact, dense structure, the oven dried (Figure 3B2, 3B3) and lyophilized (Figure 3C2, 3C3) expanded chitin resemble crinkled paper or wrinkled sheets. The samples were sputter coated with gold before imaging with a secondary electron detector, with a 15 kV accelerating voltage, and at a working distance in the range of 29-31 mm.
Figure 3. Photographs and micrographs of neat flake and expanded chitin. The photographs correspond to chitin (A1) in its neat flake form and in its expanded foam form dried by (B1) baking at 80 °C and (C1) lyophilizing. The scanning electron micrographs correspond to two magnifications of chitin (A2, A3) in its neat flake form and in its expanded foam form dried by (B2, B3) baking at 80 °C and (C2,C3) lyophilizing. Note the more compact form of the neat flakes relative to the expanded foam. Please click here to view a larger version of this figure.
These visual and microscopic observations agree with the powder X-ray Diffraction (XRD) and N2-physisorption analyses of the samples. Diffractograms show a widening of crystalline reflections and an increase in the intensity of the amorphous peak in the expanded foams relative to the neat flakes (Figure 4). This observation can be illustrated by comparing the semi-quantitative crystallinity index and the crystallite size estimates of the neat and expanded chitin. The crystallinity index is the normalized difference of crystalline to amorphous diffraction intensities29. It is given by the equation:
For chitin, the crystalline diffraction intensity typically used is that of crystal plane (110) at 19.3° and the amorphous diffraction intensity is that at 16.0°29. The crystallinity index drops from 88% in the neat flakes, to 74% in the oven-dried expanded foam, and to 58% in the lyophilized expanded foam (Table 1). The crystallite size can be estimated by the Scherrer equation30:
We assume a shape factor of 1 and the instrument used Cu Kα radiation (wavelength = 15.4 nm). Using the diffraction of the (110) plane at 19.3°, the crystallite size drops from 6.6 nm in the neat chitin to 4.4 nm in the expanded chitin (Table 1).
Figure 4. X-ray diffractograms of neat and expanded chitin. The figure shows the diffractograms of chitin in its neat flake form and in its expanded foam form dried by two different methods-baking at 800 °C and lyophilizing. All three diffractograms are normalized to the maximum intensity of reflection at 19.3 °, which corresponds to plane (110). Note the general widening of peaks in the expanded foams relative to the neat flakes. Please click here to view a larger version of this figure.
Measurements of specific surface area, obtained from N2-physisorption isotherms at 77 K using the Brunauer-Emmett-Teller (BET) equation31, lead to similar observations. For all the materials, the N2 adsorption isotherms show the uptake volume to increase linearly with partial pressure in the range p/po = 0.05-0.25 (Figure 5A), as is expected of N2 multilayer condensation32. However, the uptake volume is greatest for the expanded foams. The BET plot (Figure 5B,5C), show a positive linear correlation with partial pressure and positive intercept, indicating that the data is within the valid range of the BET equation33. As such, the specific surface area of the materials is proportional to the inverse of the sum of the slope and intercept of those lines31. While the specific surface area of the neat flakes is 12.6 ± 2.1 m2/g, that of oven dried foam is 43.1 ± 0.2 m2/g, and that of the lyophilized foam is 73.9 ± 0.2 m2/g. The changes in crystallinity index, crystallite size, and specific surface area indicate that the material either (1) forms a more open and porous structure, or (2) is degraded into smaller particles. The micrographs in Figure 3 suggest the former, but the latter cannot be ruled out without a thorough pore-size distribution analysis.
Figure 5. N2 adsorption isotherms and BET plots. (A) N2 adsorption isotherms of chitin in its neat flake form and in its expanded foam form dried by two different methods-baking at 80 °C and lyophilizing-for partial pressures in the BET range. (B, C) BET plot for the same materials and the range of partial pressures. The specific surface areas are proportional to the inverse of the sum of intercept and slope of the lines in the BET plots. Please click here to view a larger version of this figure.
In spite of the morphological changes described above, the expansion process does not appear to affect the chemical structure of chitin. The IR spectrum, obtained as attenuated total reflectance (ATR), of all chitin samples remain virtually unchanged regardless of processing (Figure 6). Note the similarity of the peaks at 1650 cm-1 and 1550 cm-1 which correspond to the amide functional group23.
Figure 6. ATR IR spectrograms of neat and expanded chitin. The figure shows the IR spectra of chitin in its neat flake form and in its expanded foam form dried by two different methods-baking at 80 °C and lyophilizing. The differences in the spectra are minimal and suggest no significant chemical changes between neat flakes and expanded foam chitin. Please click here to view a larger version of this figure.
The thermal decomposition behavior also indicates minimal chemical changes between the three samples (Figure 7). The shape of the thermogravimetric profile is identical for the expanded chitin regardless of drying method, but both differ from that of the neat flakes (Figure 7A). This is ascribed to mass and thermal diffusion limitations associated with the more compact flakes. The onset of thermal decomposition of all three samples occurs at 260 °C (Figure 7B), but the maximum decomposition rate for chitin flakes occurs at higher temperatures due to its more compact morphology.
Figure 7. Thermogravimetric profiles of neat and expanded chitin. The figure shows the integral (above) and differential (below) thermogravimetric profiles of chitin in its neat flake form and in its expanded foam form dried by two different methods-baking at 80 °C and lyophilizing. The onset of thermal decomposition of all three materials is at 260 °C, but flakes decompose over a longer temperature range relative to the foams. Please click here to view a larger version of this figure.
The increase in specific surface area is accompanied by an expected increase in the maximum uptake Cu by chitin. While the neat flakes uptake 13.8 ± 2.9 mg/g, the oven-dried foam uptakes 43.1 ± 1.9 mg/g and the lyophilized foam uptakes 73.1 ± 2.0 mg/g (Table 1). The increase in Cu uptake is more clearly illustrated by comparing the standard (Figure 8A) and linearized (Figure 8B) Langmuir adsorption isotherms. The maximum uptake is represented by the asymptotic limit in the standard isotherm and the inverse of the slope in the linearized isotherm. However, these differences in the uptake disappear when the Cu uptake is normalized by the surface area (Table 1). While the neat flakes uptake 10.5 ± 2.8 atoms/nm2, the oven-dried foam uptakes 10.7 ± 0.4 atoms/nm2, and the lyophilized foam uptakes 9.4 ± 0.3 atoms/nm2 (Table 1). This suggests that the surface of the expanded chitin is chemically similar to that of the initial chitin flakes, which agrees with spectroscopy and thermogravimetric observations.
Figure 8. (A) Standard and linearized (B, C) Cu adsorption isotherm. The figure shows Cu adsorption isotherms of chitin in its neat flake form and in its expanded foam form dried by baking at 80 °C and lyophilizing. Each data point is the average of three measurements and the error bars represent two standard deviations. Error bars for the expanded foams in the linearized isotherm are small and can only be seen in (C). The solid lines show the best fit Langmuir adsorption isotherms. The maximum uptake is the asymptotic value in the standard adsorption isotherm and the inverse slope in the linearized ones. Expanded chitin shows a higher Cu uptake than that of chitin flakes by at least a factor of 4. Please click here to view a larger version of this figure.
The proposed method for chitin foam fabrication allows for the production of such foams without the need for specialized equipment or techniques. Production of the chitin foam relies on the suspension of sodium hydride within a chitin sol-gel. Contact with water from the atmosphere induces gelling of the chitin matrix and evolution of hydrogen gas by decomposition of the sodium hydride. Therefore, the critical steps of the preparation are (1) formation of the sol-gel, (2) introduction of the sodium hydride in anhydrous conditions, and (3) the reaction of atmospheric water with the chitin sol-gel and sodium hydride suspension.
Two important limitations arise from the third step. First, the process scales up poorly. The chitin sol-gel is a highly hygroscopic and readily absorbs moisture, but as the reaction volume increases, water diffusion limitations may prevent gelling. In fact, we observed that doubling the reaction volume increased the gelling time from days to weeks. Second, the process relies on atmospheric moisture. Local climate and seasonal weather will cause variations in the gelling time. A possible modification to the procedure is to use Schlenk techniques to maintain the reaction atmosphere air and moisture free, and then gradually add water to the chitin sol-gel and sodium hydride suspension. However, such a change requires resources and skills that would limit applicability.
Both the crystallinity index and crystal size reported above are only semi-quantitative estimates. The crystallinity index was calculated as described by Focher, et al.29, and is therefore not a true crystallinity fraction. It was not obtained by comparing peak areas to those of standards of known purity. Similarly, use of the Scherrer equation to obtain crystallite size from the line broadening only provides estimates. Other phenomena, such as non-uniform strain, can also contribute to line broadening34. For this reason, it is more appropriate to focus on trends rather than the absolute values of crystallinity index and crystallite size. As recommended elsewhere, those values are reported without associated errors or variances34.
Calculating specific surface areas by applying the BET equation to N2 physisorption isotherms requires thorough drying and degassing of samples prior to the analysis. The presence of moisture and adsorbates on the sample will alter specific area measurements in two ways: (1) by blocking and lowering the effective number of vacant adsorption sites, and (2) by desorbing volatiles, increasing the measured pressure above the sample, and lowering its apparent adsorption. To prevent these errors, carbon and oxide samples are typically degassed at temperatures near 300 °C under flowing N2 or vacuum for at least 1 h. Although structurally robust, chitin will thermally decompose under such conditions (Figure 6). Instead, specific surface area measurements of expanded chitin foams were most reliable for samples degassed at 50 °C under flowing N2 for 1 week immediately after oven drying or lyophilizing.
Conducting adsorption isothermal experiments is routine, but the specific protocols vary greatly based on the adsorbent, solution, mixing method, available instruments, and convenience. For that reason, this study includes a detailed protocol based on a procedure for waste water analysis28. The adsorption of Cu on chitin is low relative to other adsorbents, such as carbons. Chitin requires high Cu concentrations in the range of 100-500 mg/L in order to reach saturation35. However, the colorimetric bicinchoninate method has a Cu detection ceiling of only 5 mg/L27. This means that aliquots had to be diluted 100 times for their Cu concentration to be measurable by the instrument. Dilutions can introduce significant experimental error into measurements, so the dilution and measurements were repeated three times per sample. Using a graduated cylinder to perform the dilutions, the observed variance in the measured concentrations were low-less than 3.7 % for low Cu concentrations and less than 0.35% for high Cu concentrations. The variance could be decreased by using volumetric flasks to perform the dilution. In addition, it is important to minimize the headspace during the adsorption experiments. Any adsorbent that adheres to the container walls above the liquid line will not equilibrate with the solution and will induce error in the experiment. This can be prevented by placing the containers at a 15° angle relative to the orbital plane of the shaker, and routinely shaking the containers by hand to dislodge any adsorbent adhered to the inner walls.
The Langmuir model for isothermal, non-dissociative adsorption assumes that (1) the analyte adsorbs in a single layer, (2) adsorption sites are energetically equivalent and can contain a single analyte molecule or ion, and (3) adsorbed molecules or ions do not interact with one another. The collected Cu adsorption data fits the Langmuir model and validates these assumptions. However, we used refined chitin harvested from a single species as the starting material. Using lower purity chitin, or chemically modifying the surface17,36, can result in greater morphological and energetic variation between adsorption sites, which would call for a different adsorption model.
The authors have nothing to disclose.
The research was sponsored by the Combat Capabilities Development Command Army Research Laboratory (Cooperative Agreement Number W911NF-15-2-0020). Any opinions, findings and conclusions, or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the Army Research Lab.
We thank the Center for Advanced Materials Processing (CAMP) at Montana Technological University for the use of some of the specialized equipment required in this study. We also thank Gary Wyss, Nancy Oyer, Rick LaDouceur, John Kirtley, and Katherine Zodrow for the technical assistance and helpful discussions.
Ammonium bicarbonate | Sigma-Aldrich | 9830 | NH4HCO3, ≥99.5 % |
Chitin | Sigma-Aldrich | C7170 | Pandalus borealis, practical grade |
Colorimeter | Hanna Instruments | HI83399-01 | Photometer for wastewater analysis |
Copper High Range Checker | Hanna Instruments | HI702 | Bicinchoninate colorimetric titration |
Copper nitrate hydrate | Sigma-Aldrich | 223395 | Cu(NO3)2 · 2.5 H2O, 98 % |
Dimethylacetamide (DMAc) | Sigma-Aldrich | 271012 | Anhydrous, 99.8 % |
IR Spectrophotometer | Thermo Nicolet | Nexus 670 | Fitted with an ATR cell |
Lithium chloride | Sigma-Aldrich | 310468 | LiCl, ≥99 % |
N2 Physisorption Apparatus | Micromeritics | Tristar II | |
Nitric acid | BDH | BDH7208-1 | HNO3, 0.1 N |
Scanning electron microscope | Zeiss LEO | 1430 VP | 15 kV, secondary electron detector, 29-31 mm working distance |
Sodium hydride | Sigma-Aldrich | 223441 | NaH, packed in mineral oil, 90 % |
Thermogravimetric analyzer | TA Instruments | Q500 | 100 ml/min N2, 10 °C/min to 800 °C |
Water Purification System | Millipore | Milli-Q | Type A water (18 MΩ) |
X-Ray Diffractometer | Rigaku | Ultima IV | Cu K-α radiation, 8.04 keV |