Environ. Sci. Technol. 2003, 37, 1008-1012
Recycled Wool-Based Nonwoven Material as an Oil Sorbent M A J A M . R A D E T I CÄ , * , † D R A G A N M . J O C I CÄ , † PETAR M. JOVANC ˇ I CÄ , † Z O R A N L J . P E T R O V I CÄ , ‡ A N D HELGA F. THOMAS§ Textile Engineering Department, Faculty of Technology and Metallurgy, University of Belgrade, Karnegijeva 4, 11120 Belgrade, Yugoslavia, Institute of Physics, P.O. Box 68, 11080 Zemun,Yugoslavia, and Deutsches Wollforschungsinstitut an der RWTH Aachen, Veltmanplatz 8, 52062 Aachen, Deutschland
The aim of this study was to highlight the possibility of using recycled wool-based nonwoven material as a sorbent in an oil spill cleanup. This material sorbed higher amounts of base oil SN 150 than diesel or crude oil from the surface of a demineralized or artificial seawater bath. Superficial modification of material with the biopolymer chitosan and low-temperature air plasma led to a slight decrease of sorption capacity. Loose fibers of the same origin as nonwoven material have significantly higher sorption capacities than investigated nonwoven material. White light scanning interferometry analysis of the fibers suggested that roughness of the wool fiber surface has an important role in oil sorption. The laboratory experiments demonstrated that this material is reusable. Recycled wool-based nonwoven material showed good sorption properties and adequate reusability, indicating that a material based on natural fibers could be a viable alternative to commercially available synthetic materials that have poor biodegradability.
Introduction It has been known for decades that some natural fibers have higher sorption capacities for oil than commercially available synthetic fibers (1). Synthetic fibers play a dominant role in oil spill cleanup because of their hydrophobic (i.e., oleophilic) properties that are important preconditions for an efficient oil sorbent. Increasing environmental concern, especially after several hazardous incidents in the past decades when large quantities of oil were spilled into the sea, renewed the interest for natural fibers (2). Choi and co-workers demonstrated that cotton, milkweed, and kenaf have 1.5-3 times better sorption properties than polypropylene fibers (3, 4). Excellent oil sorption properties and high biodegradability of natural fibers make them particularly attractive as a possible alternative to synthetic fibers. The introduction of strict ecological legislation, the lack of available space for waste deposition, and the need for higher responsibility of current and future generations to achieve further economic expansion but with limited re* Corresponding author e-mail:
[email protected]; phone: +381-11-3370-460; fax: +381-11-3370-387. † University of Belgrade. ‡ Institute of Physics. § Deutsches Wollforschungsinstitut an der RWTH Aachen. 1008
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 5, 2003
sources of raw material led to a consideration of broadening applications of waste and recycled materials (5). Recent studies in Germany demonstrated that about 1.8 million ton of textile material is thrown away every year (6). From that quantity, 23% goes to the second-hand section, 73% is deposited (i.e., burned), and only 4% is recycled. To address all these environmental requirements, we have developed recycled wool-based nonwoven material. The aim of this study was to examine the possibility of using this material as a sorbent that can be applied for the removal of oils from the water surfaces. The effect of chitosan (CHT) and low-temperature air plasma (LTP) treatment on oil sorption properties has also been investigated.
Experimental Section Materials. The experiments were performed using recycled wool-based nonwoven material (78% wool/22% polyester). Second-hand knitted pullovers of the same quality and characterstics were torn off, washed, decolorized with reducing agent, dried, and garneted in industrial conditions. To avoid the effect of chemical binders on oil adsorption, the needlepunch process was chosen to produce the nonwoven material (7). The material was produced from recycled fibers on the Dilo (Germany) needle loom. The physical and mechanical properties of the obtained material are given in Table 1 (8). The characteristics of investigated oils are given in Table 2. To obtain constant experimental conditions (i.e., to exclude the effect of light hydrocarbons that evaporate easily), oil samples were stored in an appropriate hood for 48 h (3, 4). Chitosan (Vanson, USA) with viscosity 16 cps and deacetylation degree 88.6% was used without further purification. Artificial seawater was produced in accordance with DIN 50905-4 (9). The procedure for treatment of material with chitosan is previously described (8). Low-temperature plasma treatment was carried out in a capacitively coupled, radiofrequency plasma operating at 13.56 MHz in air (10). Treatment time was 5 min and pressure was 0.2 mbar with the power supply maintained at the constant level of 100 W. Methods. The desired amount of oil was placed in 500 mL of artificial seawater or demineralized water in a 1-L glass beaker. One gram of dry material was then put in the beaker and shaken in a laboratory shaker (3, 4). After 1 min of drainage in a sustainer, the wet material was weighed. The amount of oil sorbed (oil capacity of the material) was determined by the following equation:
q)
mf - (mo + mw) mo
(1)
where q is the amount of oil sorbed (oil sorption) (g/g), mf is the weight of the wet material after 1 min of drainage (g), mo is the initial weight of the material (g), and mw is the water content (g). Water content was determined by distillation with a mixture of toluene and xylene (20/80, v/v) as a solvent in accordance with ASTM D95-83 (11). To analyze the amount of oil sorbed on the material in oil medium without any water, 1 g of material was placed in 100 mL of oil in a glass beaker of 200 mL (3, 4). As in previous procedure, after shaking for 10 min, material was drained in a sustainer for 1 min and weighed. The amount of oil sorbed was determined as a difference between the weight of wet material and the initial weight of material. 10.1021/es0201303 CCC: $25.00
2003 American Chemical Society Published on Web 01/31/2003
TABLE 1. Physical and Mechanical Properties of Untreated Nonwoven Material weight (g/m2)
breaking strengtha (N)
bursting strength (N)
thickness (mm)
235
19.23
21.97
1.56
a
Machine direction.
TABLE 2. Characteristics of Oil Samples wt loss (%) sample
viscosity KINcst (40 °C)
sp gr (15.6 °C)
after 24 h
after 48 h
base oil SN 150a diesel crude oil
30.30 2.97 11.19
0.871 0.844 0.879
0.22 1.50
0.37 5.59
a
Paraffinic solvent neutral (S.U.S. 150).
FIGURE 2. Sorption of SN 150, diesel, and crude oil in demineralized water.
FIGURE 1. Sorption kinetics of untreated nonwoven material in demineralized water containing 40 g of oil. Reusability of material was studied only for the case of sorption in oil medium by the above-explained procedure. Five cycles of sorption process were performed for each sample. Between each cycle, material was squeezed between rollers and weighed again. For investigation of fiber surface roughness, white light scanning interferometry analysis (WLSI) was applied using Wyko profilometer NT 2000 (Veeco Metrology Group, USA). The profilometer is used for noncontact measurements of 3D surface profiles of materials with high resolution (12). This is accomplished by illuminating the sample with white light and analyzing interferometrically the reflected beam and an internal standard flat. 3D images that demonstrate physical surface characteristics are constructed from these data. The system measures roughness, form, and waviness of surface down to the nanometer scale.
Results and Discussion Oil Sorption from Aqueous Medium. Sorption kinetics of untreated nonwoven material in demineralized water containing 40 g of oil (SN 150, diesel, and crude oil) is shown in Figure 1. The prolongation of sorption process over 10 min has no significant influence on sorption of SN 150 and diesel oil. In case of crude oil, sorption properties were worsened for longer sorption times. Therefore, time of 10 min was chosen for all further investigations on wool sorption of oil. The influence of the initial content of oil in demineralized water and artificial seawater bath on sorption capacity of untreated material for SN 150, diesel, and crude oil is demonstrated in Figures 2 and 3, respectively. No significant difference between the sorption in demineralized water and that in artificial seawater bath was observed. The results indicate that all investigated oil samples showed similar sorption trends since they have similar specific gravities.
FIGURE 3. Sorption of SN 150, diesel, and crude oil in artificial seawater. Sorption capacity increased with the content of oil in the bath until the material was saturated with oil. Sorption capacity of nonwoven material increased in the following order:
SN 150 > diesel > crude oil Sorption capacity of the material for SN 150 is about 11% higher than for diesel and 17% for crude oil. This could be due to the higher viscosity of SN 150 as compared to other oil samples. High viscosity of oil can cause two opposite effects: decrease of sorption as the penetration through the interior of the fiber is inhibited and improved sorption since the oil is better adhered to the material surface (4). This was particularly noticeable in case of sorption of diesel oil that has almost 10 times lower viscosity than SN 150. Initially, a great amount of diesel oil was sorbed to the material, but subsequently it was rapidly drained during the 1-min drainage in sustainer. Sorption behavior of the dry sample and the sample presoaked by water for SN 150, diesel, and crude oil is shown in Figure 4. Dry samples showed a slightly better sorption properties than wet samples. It is well-established that chitosan has the ability to bind heavy metal ions, fats, and proteins effectively (13-15). However, an attempt to treat the investigated nonwoven wool VOL. 37, NO. 5, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1009
FIGURE 4. Oil sorption of dry and wet untreated material in demineralized water bath with 40 g of SN 150, diesel, and crude oil.
FIGURE 5. Effect of treatment on sorption in artificial seawater bath containing 40 g of oil. material with chitosan in order to improve its oil sorption properties was not successful. CHT treatment of the material leads to a partial hydrophilization of the wool fiber and accordingly to reduction of wool sorption capacity shown in Figure 5. CHT treatment caused the reduction of sorption capacity by 9.89% in case of SN 150, by 15.34% for diesel, and by 4.52% for crude oil as compared to untreated material. Knowing that fibers with hydrophobic (i.e., oleophilic) nature have predisposition for excellent oil sorption properties, it was expected that LTP treatment would impair even worse oil sorption than CHT treatment of the material. Such prediction was based on the fact that such LTP treatment results in extremely high hydrophilization of the wool fiber because of the modification of a covalently bound layer of fatty acids on the fiber surface and the formation of new polar functional groups (16). During the plasma treatment, the fiber surface undergoes a severe bombardment of different plasma particles (neutrals, ions, free radicals, metastables and UV photons) that have enough energy to modify the surface and cause etching (17). Consequently, 0.9 nm thick fiber surface fatty acid layer (F-layer) that is mostly responsible for the hydrophobic nature of wool is modified and partially removed (18). Unexpectedly, the LTP-treated samples demonstrated better sorption properties as compared to CHT-treated samples and even better than the untreated sample in case of crude oil as shown in Figure 5. These results pointed out that hydrophobicity and oleophilicity are important but are not the only factors that influence the oil sorption on wool. It is also obvious that oil sorption on wool is a very complex process that is governed by several different mechanisms. It is well-known that physical properties of the fiber such as crimp, twist, surface roughness, and porosity have significant influence on the oil sorption properties (19). Choi and Moreau suggested that adsorption should be the most prominent mechanism of oil sorption on wool because of (i) the existence of waxes and grease giving a hydrophobic nature to the fiber, (ii) the scale like structure with large pores, and (iii) the fiber crimp providing the space for the deposition of oil and formation of capillary bridges of oil between the fibers (19). However, these authors also stressed that medulla, 1010
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 5, 2003
FIGURE 6. WLSI images of (a) untreated, (b) LTP-treated, and (c) CHT-treated sample. the central part of some wool fiber types, can significantly contribute to the oil absorption mechanism of oil sorbability (19). Previous studies pointed out that good wool sorption performance for oil was mainly attributed to the presence of grease (lanolin) and waxes (1, 19). However, properties of recycled wool fibers are supposed to be different from raw wool fibers since they have undergone different mechanical and chemical changes during the primary material production, the use and maintenance, and the production of nonwoven material. Lanolin and waxes are removed in great quantities during the washing process (20, 21). Accordingly, it can be presumed that recycled wool fibers do not possess waxes or possess waxes in a small amount. Therefore, adsorption by van der Waals forces and hydrophobic interactions between fiber surface and oil is not expected to be the dominant mechanism of oil sorption in the case of recycled wool fibers. WLSI images of untreated, LTP-treated, and CHT-treated samples are demonstrated in Figure 6. In Figure 6a, a characteristic scaly surface of the wool fiber can be observed. WLSI analysis revealed that plasma treatment increases surface microroughness as compared to the untreated sample. The surface of the LTP-treated sample is more rough (i.e., more peaks and valleys are introduced). However, CHT-treated samples are more flat as compared to both untreated and LTP-treated samples. It appears that
FIGURE 7. Oil sorption of nonwoven material and loose fibers in the seawater bath containing 40 g of oil. the polymer has masked the wool scales and flattened the surface. The absence of pitts and peaks can be noticed. The Rv parameter presents the height difference between the mean line and the lowest point over the evaluation length. It demonstrates the maximum depth of valley in the analyzed area and provides information about possible retention of oil in that area. Average Rv values obtained for untreated, LTP-treated, and CHT-treated samples are -118.0, -163.2, and 108.5 nm, respectively. The valleys in LTP-treated samples are deeper as compared to those of untreated samples. The positive value of Rv in the case of CHT-treated sample indicates that valleys are likely to be filled with biopolymer and that the fiber surface has become more flat. The results pointed out that oil sorption on wool is most likely governed by an adsorption process because of a scaly structure of the fiber that promotes formation of capillary bridges of oil between fibers and surface roughness, which has significant influence on the increase of the surface area (19). Thus, the acceptable explanation for good sorption behavior of LTP-treated samples despite their high hydrophilicity is that LTP treatment induces an enlargement of the fiber specific surface area (i.e., significant morphological changes) (22, 23) and an increase of surface microroughness, promoting the adsorption by physical trapping. The comparison of the results presented here with corresponding literature data related to raw wool loose fibers pointed out that the latter have more than 100% better sorption properties for the oil of similar characteristics (1, 19). Accepting the fact that type and construction of material have considerable influence on sorption process and that once the fibers are incorporated in material the sorption properties are permanently changed (3), the analysis under the identical experimental conditions with loose fibers has been performed. The loose fibers originated from the same primary material as investigated nonwoven material and have been collected after garnetting. Oil sorption of nonwoven material and loose fibers in the artificial seawater bath containing 40 g of oil (SN 150, diesel, and crude oil) is shown in Figure 7. Obviously, loose fibers sorb oil remarkably better than the nonwoven material. It was noticeable that irregular position of fibers as well as more available space between them intensified the sorption of oil. The values of sorption capacities of investigated recycled wool fibers are close to the literature data relating to raw wool fibers (1). Sorption capacity of loose fibers increased 103% for SN 150, 74% for diesel, and 111% for crude oil as compared to nonwoven material. Loose fibers sorbed from 52.7 to 72.5% of the maximum amount of 40 g of oil depending on investigated oil. On the contrary, after oil sorption on nonwoven material, approximately 70% of oil was retained in the bath. Previous work of Choi et al. showed that the variation of cotton nonwoven material parameters such as the number of passes can dramatically change the sorption processes (3). The capillary interaction of oil between neighboring fibers is inhibited since nee-
FIGURE 8. Oil sorption properties of untreated, CHT-treated, and LTP-treated in SN 150, diesel, and crude oil bath.
FIGURE 9. Reusability of untreated, CHT-treated, and LTP-treated nonwoven material for SN 150. dlepunch process induces the reduction of space between the fibers. Accordingly, oil sorption capacity is decreased. Our results pointed out that further attempts in optimization of the oil sorption process should be focused on the variation of the nonwoven material parameters. Oil Sorption from Oil Medium. To investigate the maximum oil sorption capacity of the nonwoven material, its sorption properties were analyzed in oil medium without any water. Oil sorption properties of untreated, CHT-treated, and LTP-treated in SN 150, diesel, and crude oil bath are demonstrated in Figure 8. The sorption capacities of the material are slightly higher in oil than in aqueous medium. Here again, sorption properties increased in the following order:
untreated > LTP-treated > CHT-treated Similar to the sorption in aqueous medium, SN 150 is the best-sorbed oil, followed by crude and diesel oil. Reusability of the Nonwoven Material. Oil sorption capacities of the nonwoven material after five cycles of the sorption process for SN 150 are shown in Figure 9. The results demonstrated that approximately 88% of oil was removed from the material after squeezing it between rollers. It is obvious that sorption decreases with the increase of sorption cycles. However, sorption capacities of the material after five cycles of sorption were good enough, particularly for SN 150, indicating that it could be applied as an efficient oil sorbent for several uses. Similar behavior has occurred in case of diesel and crude oil where the decrease of the sorption capacities was more pronounced on CHT- and LTP-treated samples than on the untreated sample (8). Obtained results are comparable with literature data related to nonwoven material made of polypropylene fibers as the most commonly used fibers for sorbent material (4). VOL. 37, NO. 5, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1011
Comparison of literature data for the sorption capacities of polypropylene fibers with recycled wool loose fibers we have used showed that wool has a much higher sorption capacity (19). However, previous researches indicated that several factors (fibers fineness, form of material, crimp, oil characteristics, etc.) determine the sorption behavior of the material and that all of them should be taken into account while comparing the different sorbents (1, 3, 4). Recycled wool-based nonwoven material is obviously an efficient sorbent either in water or in oil medium, and it can be a viable alternative to commercially available synthetic materials because of the relatively high oil sorption capacity, reusability, and biodegradability.
Acknowledgments We express our gratitude to Mr. M. Suboticˇki (BVK, Yugoslavia) and Mr. Z. Blagojevic´ (INTEKS, Yugoslavia) for their help in the preparation of nonwoven material as well as to Mrs. M. Kolb (NIS, Yugoslavia) for supplying us with oil samples.
Literature Cited (1) Johnson, R. F.; Manjrekar, T. G.; Halligan, J. E. Environ. Sci. Technol. 1973, 7, 439. (2) Crighton, J. S. Proceedings, 9th International Wool Textile Research Conference, Biella, Italy; Citta degli Studi: Biella, Italy 1995; Vol. I, p 345. (3) Choi, H. M.; Kwon, H. J.; Moreau, J. P. Text. Res. J. 1993, 63, 211. (4) Choi, H. M.; Cloud, R. M. Environ. Sci. Technol. 1992, 26, 772. (5) Jovanovic´, R. Natural and Chemical Protein Fibers; IRO Gradjevinska knjiga: Beograd, 1989; p 301 (in Serbian). (6) Elsasser, N.; Maetschke, O.; Wulfhorst, B. Melliand Textilber. 1998, 79, 768.
1012
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 5, 2003
(7) Tattersall, R. Proceedings, World Textile Congress, Industrial Technical and High Performance Textiles, Huddersfield, U.K.; University of Huddersfield: Huddersfield, U.K., 1998; p 1. (8) Radetic´, M.; Jocic´, D.; Jovancˇic´, P.; Rajakovic´, Lj.; Petrovic´, Z. Lj.; Thomas, H. DWI Rep. 2002, No. 125, 463. (9) DIN 50905-4, Corrosion of Metals, 1987. (10) Radetic´, M.; Jocic´, D.; Jovancˇic´, P.; Trajkovic´, R.; Petrovic´, Z. Lj. Text. Chem. Color. Am. Dyest. Rep. 2000, 32 (4), 55. (11) ASTM D-95-83, Standard Test Method for Water in Petroleum Products and Bituminous Materials by Distillation, 1969. (12) Kaist-PEM Precision Engineering & Metrology. http:// pemkaist.ac.kr/psi&wsi.html (accessed Oct 2002). (13) Maslova, G.; Krasavtsev, V. Proceedings, 7th International Conference on Chitin, Chitosan and Euchis ‘97, Lyon, France; European Chitin Society: Villeurbanne, France 1997; p 554. (14) Kulak, Z.; Niekraszewicz, A.; Struszczyk, H. Fibres Text. East. Eur. 1999, October/December, 60. (15) Yoshida, H.; Kishimoto, N.; Kataoka, T. Ind. Eng. Chem. Res. 1995, 34, 347. (16) Dai, X. J.; Hamberger, S. M.; Bean, R. A. Aust. J. Phys. 1995, 48, 939. (17) Lee, K. S.; Pavlath, A. E. Proceedings, 5th International Wool Textile Research Conference, Aachen, Germany; Deutsches Wollforschungsinstitut an der RWTH: Aachen, Germany, 1975; Vol. III, p 275. (18) Klausen, T.; Thomas, H.; Ho¨cker, H. Proceedings, 9th International Wool Textile Research Conference, Biella, Italy; Citta degli Studi: Biella, Italy, 1995; Vol. II, p 241. (19) Choi, H. M.; Moreau, J. P. Microsc. Res. Tech. 1993, 25, 447. (20) Hartley, F. R. Wool Sci. Rev. 1969, 37, 23. (21) Poole, A. J.; Cord-Ruwisch, R.; Jones, F. W. Water Res. 1999, 33, 1981. (22) Hesse, A.; Ho¨cker, H.; Umbach, K. M.; Mecheels, J. Presented at the IWTO Meeting, Harrogate, U.K., June 1995; Report 12. (23) Denda, B. Ph.D. Dissertation, RWTH, Aachen, 1999.
Received for review July 4, 2002. Revised manuscript received November 23, 2002. Accepted December 23, 2002. ES0201303
