What Is the Real Value of Chitosan's Surface Energy

Jan 26, 2008 - Because of conflicting reports and unrealistic literature values, a systematic study of the surface energy of chitin, chitosan, and the...
1 downloads 16 Views 109KB Size
610

Biomacromolecules 2008, 9, 610–614

What Is the Real Value of Chitosan’s Surface Energy? Ana G. Cunha,* Susana C. M. Fernandes, Carmen S. R. Freire, Armando J. D. Silvestre, Carlos Pascoal Neto, and Alessandro Gandini CICECO and Department of Chemistry, University of Aveiro, Campus de Santiago, 3810-193 Aveiro, Portugal Received October 31, 2007; Revised Manuscript Received December 6, 2007

Because of conflicting reports and unrealistic literature values, a systematic study of the surface energy of chitin, chitosan, and their respective monomeric counterparts was carried out using contact angle measurements on films and pellets, before and after different purification procedures. All the commercial samples of these polymers were shown to contain nonpolar impurities that gave rise to enormous errors in the determination of the polar component of their surface energy. After their thorough removal, the value of the total surface energy (γs), and particularly of its polar component, increased considerably and reached the classical polysaccharide figures of γsd ∼ 30 and γsp ∼ 30 mJ/m2. The characterization of the impurities by gas chromatography/mass spectrometry analysis indicated the presence of significant amounts of higher alkanes, fatty acids, and alcohols and sterols.

Introduction The rapidly growing interest in the exploitation of renewable resources for the synthesis of novel macromolecular materials as viable alternatives to petroleum-based counterparts includes of course the use of natural polymers like polysaccharides and lignin. Whether they are investigated as blend components or as precursors to chemical modifications, knowledge of their surface properties and, more specifically, their surface energy is obviously essential. Cellulose and starch have similar polymer structures, dominated by OH functions, and both the dispersive and the polar components to their surface energy are high, viz., 30–40 and 20–30 mJ/m2, respectively,1–3 for different purified materials. These values reflect convincingly the facts that, on the one hand, they refer to macromolecules, hence their high dispersive component, and, on the other hand, they are associated with a predominance of OH groups at their surface, hence a high polar contribution. Factors like the percentage of crystallinity, the surface morphology, the possible presence of oxidized moieties like COOH groups, and the amylose/amylopectin composition in starch are responsible for their variations. It follows that polysaccharides, with their prevalent polarity, display values of the total surface energy which are among the highest measured for macromolecular materials. Only those synthetic polymers incorporating similarly dominant polar moieties in their monomer units, like poly(acrylamide) and epoxy resins,4 have total surface energies (γs ∼ 50 mJ/m2) approaching those of cellulose and starch, whereas most of the monomer units of other common macromolecular materials are made up either exclusively of nonpolar moieties, like polyolefins (γs ) 30–35 mJ/m2),4 or of a variable combination of nonpolar and polar moieties, like polyesters and poly(vinyl chloride) (γs ) 35–40 mJ/m2).4 We recently started working with chitosan (Figure 1), both as a physical component of blends and coatings and as a precursor to novel materials, through its surface or bulk chemical modification. This polymer and its derivatives have been attracting growing attention, because of their remarkable properties and aplications.5 A bibliographic search related to its surface * Corresponding author: e-mail, [email protected]; tel, +351 234 401 405; fax, +351 234 370 084.

energy6–11 revealed some puzzling data, in the sense that in all the publications in which both the polar and the dispersive components were determined, the former contribution was systematically very low, varying from 1 to 8 mJ/m2. The values of the latter contribution, most around 30 mJ/m2, were more in tune with a polysaccharide structure and in reasonable agreement among themselves,7–10 with only one6 much lower figure of 17 mJ/m2. In another study,11 only the total surface energy was reported with, again, an exceedingly low value of 18 mJ/m2. All these data were based on contact angle measurements. Chitosan, cellulose, and starch are all polysaccharides, the only difference residing in the replacement of an OH group in each saccharide unit of cellulose by an NH2 counterpart in chitosan and by the presence of branched structures in starch. It is therefore surprising to encounter repeatedly very low values of the polar contribution to the surface energy, spanning some 15 years of publications, considering moreover their lack of reproducibility from one study to the next. No cross reference was provided in any of these publications nor any discussion related to these seemingly abnormal results. The use of inverse gas chromatography, which ensures a clearcut approach to the dispersive component of the surface energy of solids, yielded values of about 50 mJ/m2 for chitosans with different degrees of deacetylation,3 which are higher than the corresponding values determined by contact angle measurements,7–10 a frequently observed difference between the two techniques.12 Given this confused state of affairs, we decided to undertake a brief investigation aimed at unravelling the issue by calling upon different samples of chitosan and chitin, as well as model compounds and applying a series of purification procedures to assess their effect on the free energy of the ensuing surfaces.

Experimental Section Sample Characteristics. Chitosan samples were kindly provided by Mahtani Chitosan Pvt. Ltd. (India) and Norwegian Chitosan AS (Norway), and their molecular weight and deacetylation degree (DDA) are listed in Table 1. The chitin sample (Figure 1) used in this study, also a generous gift from Mahtani Chitosan Pvt. Ltd. (India), had a DDA of 30%. Their respective model compounds, D-(+)-glucosamine hydrochloride 99% (GlcN) and N-acetyl-D-glucosamine 99% (GlcNAc)

10.1021/bm701199g CCC: $40.75  2008 American Chemical Society Published on Web 01/26/2008

Chitosan’s Surface Energy

Biomacromolecules, Vol. 9, No. 2, 2008 611

Figure 1. Chemical structure of chitosan (1 - x . x) or chitin (x . 1 - x). Table 1. Characteristics of the Chitosan Samples Used in This Study sample

source

DDA (%)

mol wt

Ch98 Ch95 Ch79 Ch67

Mahtani Chitosan Pvt. Ltd. Mahtani Chitosan Pvt. Ltd. Norwegian Chitosan AS Norwegian Chitosan AS

98 95 79 67

350000 543000 58000 58000

(Figure 2), were purchased from Sigma and used as received. The DDAs of the chitosan samples were determined by 1H NMR spectroscopy in D2O containing 1% of CD3COOD using a DRX-300 Brüker instrument. Their molecular weights were determined by viscosimetry.13 All the samples were dried in a vacuum oven before being used. Chitosan Purification. All the chitosan samples were purified by reprecipitation using a 10% NaOH aqueous solution from a 1% aqueous AcOH solution, previously filtered successively through Porafil membranes (3, 1.2, and 0.8 µm). The ensuing precipitate was filtered and thoroughly washed with methanol/water mixtures, whose composition was progressively varied from 70/30 up to 100/0 (v/v).14 Films and pellets were prepared from all these samples. The commercial and the reprecipitated chitosan powder samples, as well as the commercial chitin sample, were sequentially Soxhlet extracted with dichloromethane, n-hexane, and acetone. After each extraction, the solution was evaporated to dryness and the residues were analyzed by GC-MS after derivatization. The films were only extracted with acetone. GC-MS Analysis. Before GC-MS analysis, samples of the extracted impurities were silylated in pyridine at 70 °C for 30 min with trimethylchlorosilane in the presence of bis(trimethylsilyl)trifluoracetamide, according to a described procedure.15 The GC-MS analyses of the derivatized extracts were performed using a Trace Gas Chromatograph 2000 series, equipped with a DB-1 J&W capillary column (30 cm × 0.32 mm, 0.25 µm film thickness), using helium as carrier gas (35 cm/s), which was coupled with a Finnigan Trace MS mass spectrometer. The chromatographic conditions were as follows: initial temperature, 80 °C for 5 min; heating rate, 5 °C/min; final temperature, 285 °C for 15 min; injector temperature, 200 °C; transfer-line temperature, 280 °C; split ratio 1:35. Preparation of the Pellets. The pellets of both commercial and variously purified chitosan and chitin, as well as of the respective model compounds, were prepared using a Graseby Specac (6 ton during 1 min) laboratory press. Preparation of the Films. Films were obtained by the casting method using 1% w/v solution of the purified chitosan samples (except Ch95, Table 1, which was used in its commercial form) in 1% v/v aqueous acetic acid. The initial dispersions were stirred for 12 h to achieve total dissolution, before spreading them onto an acrylic plate and drying them at 30 °C in a ventilated oven for 16 h. Contact Angles and Surface Energy. Contact angles (CA) with water, formamide, and diiodomethane were measured with a “Surface Energy Evaluation System” commercialized by Brno University (Czech Republic). Each θ value (average of three to five measurements, with an associated standard deviation of (2°) was the first captured by the

Figure 2. Structures of (a) D-(+)-glucosamine and (b) N-acetyl-Dglucosamine. Table 2. Total surface energy, together with its polar and dispersive components, relative to all the pellets prepared from the samples’ powders analyzed here

GlcNAc GlcN Chitin Ch98 Ch95 Ch79 Ch67

γsP (mJ/m2)

γsd (mJ/m2)

γs (mJ/m2)

29 29 11 0.1 0.4 3 ∼0.0

33 33 41 41 38 40 31

62 62 52 41.1 38.4 43 31

instrument following the drop deposition on the sample surface, which had previously equilibrated with the vapor of the liquid to be tested. All measurements were carried out at the laboratory temperature, which varied between 23 and 25 °C. The values of the dispersive and polar components to the surface tension of the liquids used for the CA measurements are available in the literature.16 The CA values were then used to calculate the dispersive and polar contributions to the surface energy of the samples, using Owens-Wendt’s approach.17

Results And Discussion The unreasonably modest values for the surface energy of chitosan, particularly in relation to the polar component, published by several authors in the last 15 years, strongly suggested to us that nonpolar impurities were responsible for this anomalous feature. Hence, a series of purification procedures were applied to the chitosan and chitin samples in order to verify the validity of this assumption. Contact Angles and Surface Energies. As shown in Table 2, the values for the polar component of the surface free energy of the commercial chitosan and chitin samples used in this work were particularly low, thus confirming the general trend related to chitosan previously reported, albeit without comments, by other authors.6–11 These results are in complete divergence with the corresponding values obtained for the model compounds, namely, GlcN and GlcNAc (Table 2), for which both polar and dispersive components of the surface energy are high and in excellent tune with those of starch and cellulose, viz., γsp ≈

612 Biomacromolecules, Vol. 9, No. 2, 2008

Cunha et al.

Figure 3. Variation of the surface energy of the chitosan pellets before and after purification treatments (Ext, extracted; RepExt, reprecipitated and extracted).

Figure 4. Variation of the surface energy of the chitin pellets before and after Soxhlet extraction (Ext, extracted).

γsd ≈ 30 mJ/m2. Although GlcN hydrochloride is a watersoluble substance, the deposition of a water droplet on the surface of its pellets gave enough time to register the corresponding contact angles (see Experimental Section) before any substrate dissolution by diffusion. It seemed therefore most unlikely that, when joined in a macromolecular chain, these structures should behave in such a way, as to lose most of their polarity when exposed to the atmosphere! This dramatic difference in behavior constituted the first indication corroborating the idea of nonpolar impurities present in the commercial polymers, but absent in their monomeric counterparts. The origin of these impurities is clearly associated with the natural crustacean morphologies from which chitin is extracted and then converted into chitosan, that are rich in lipids,

dyes, calcium carbonate, and proteins.18 We therefore decided to apply different purification procedures to the commercial chitosan and chitin samples in order to detect any increase in surface energy and to identify the ensuing impurities. First of all, sequential Soxhlet extractions of the chitosan and chitin samples were carried out with dichloromethane, n-hexane, and acetone. After each Soxhlet extraction, the contributions to the surface energy were assessed on pellets of the residual material, which showed that both the dispersive and the polar components had increased, more so the latter. Figures 3 and 4 show the values obtained after the three extractions, which emphasize the drastic increase in γsp with all the samples, albeit in different quantitative proportions. With the “purest” commercial sample according to the manufacturer, Ch98, the

Chitosan’s Surface Energy

Biomacromolecules, Vol. 9, No. 2, 2008 613

Figure 5. Variation of the surface energy of some chitosan films before and after Soxhlet extraction (Ext, extracted).

respective contributions had already reached values close to those of GlcN and other polysaccharides. Moreover, the values obtained for the extracted chitin, namely, γsp ) 23.2 and γsd ) 39.1 mJ/m2 were very similar to those published by Nair et al.,19 viz., γsp ) 20 and γsd ) 32.6 mJ/m2, using the same approach, and an almost identical value for the dispersive component was reported by Belgacem et al. (γsd ) 38.3 mJ/ m2),3 using inverse gas chromatography. The four chitosan samples were also purified by reprecipitation followed by the same sequential Soxhlet extrations. Once again, Figure 3 shows that after this double treatment, the polar component was enhanced to higher levels than with the extraction sequence alone, suggesting that a higher proportion of impurities had been removed. Furthermore, Figure 3 indicates that, whereas the initial quality of the samples played an important role in the extent of purification level achievable (with Ch98 attaining surface energies entirely comparable with those of its monomeric structure and of other polysaccharides), the DDA did not appear to be a crucial factor affecting the surface energy, since the Ch95 and the Ch79 gave similar values for the polar component after the purification treatments. This important aspect is corroborated even more strongly by the fact that the two monomer models gave identical values of γsp and γsd, despite the fact that their structure differs by the presence of relatively less polar acetylamide moiety. In other words, the polar contribution to the surface energy of these substrates, in both a monomeric (NH3+,Cl-, or amide) and a polymeric form (acetate for chitosan films, and also for chitosan precipitated without neutralization, or NH2 for precipitated and neutralized chitosan), is not significantly affected by the specific nature of this nitrogen-bearing moiety in the presence of the very strong accompanying contribution of the two OH groups. Obviously, these conclusions have nothing to do with the actual chemical reactiVity of the different N-containing groups and only relate to their role in determining the surface energy of the corresponding substrates. Figure 5 shows that the film casting of both pristine and purified chitosan samples did not provide the same increase in γsp as with pelleted powders. This can be rationalized by considering that, during the slow process of film formation, even

minute amounts of residual nonpolar impurities were adsorbed efficiently at the liquid surface, just like surfactants, and thereafter remained imprisoned as solid monolayers, a phenomenon which is obviously much less pronounced when chitosan powders are solvent extracted and/or reprecipitated. The validity of this interpretation was unambiguously proved by scraping the surface of the films of the purified chitosans, an operation which resulted in a drastic decrease in the water contact angle, typically going from 95–110 to 40–60°, the latter values being the same as those measured for GlcN and the purified Ch98. This simple experiment provided strong evidence that the nonpolar impurities had indeed migrated (almost) entirely to the film surfaces. Interestingly, scraping the surface of the pellets produced a much more modest effect and indeed none at all for GlcN and purified Ch98. The possible role of the surface roughness on the CA values was assessed by preparing pellets of different surface morphology, by varying the particle size of the sample and the pressure applied in the fabrication of the pellets. No significant trend was encountered, outside the standard CA deviation, which suggested that in the present context the roughness parameter did not influence appreciably the CA measurements. As for the scraping experiments, the same doubt arose concerning the inevitable change in surface roughness associated with this operation. In order to check for a possible effect of scraping as such, i.e., in the absence of surface impurities, we applied it to a pure cellophane film. Several tests revealed that the CA values, compared with those taken on the unscraped surface, tended to vary randomly with (10°, thus ruling out a univocal role of scraping, which could have cast a doubt on our above interpretation related to the removal of low-energy impurities from the surface of chitosan films. GC-MS Analysis. After each Soxhlet extraction, the extracted impurities from both chitosan and chitin, were silylated and analyzed by GC-MS. The most abundant compounds, identified by this technique and reported in Table 3, had predominantly nonpolar structures like higher alkanes, fatty acids, and alcohols. These results are entirely in tune with the fact that the chitin and chitosan samples employed in this investigation were extracted from the exoskeleton of crustaceans.

614 Biomacromolecules, Vol. 9, No. 2, 2008

Cunha et al.

Table 3. Identification of the Main Compounds Extracted from the Chitin and Chitosan Samples family alkanes alcohols

fatty acids

sterols a

compound

%a

heptacosane nonacosane triacontane glycerol tetradecanol hexadecanol (9Z)-octadecenol octadecanol octacosanol tetradecanoic acid hexadecanoic acid oleic acid octadecanoic acid docosanoic acid cholesterol

5.6 8.1 5.8 2.5 0.6 14.0 20.0 11.5 1.0 9.7 10.6 4.0 4.4 0.9 1.5

Percentage of each impurity relative to the total identified amount.

This external anatomical feature is constituted by several layers, namely, epicuticle, exocuticle, and endocuticle. The latter two contain the chitin macromolecules and are linked to the former, which contains waxes and paraffines, fatty acids, esters, and alcohols.18 The presence of these impurities in commercial chitin and chitosan constitutes, as clearly shown above, an enormous source of error in the determination of the surface free energy of these biopolymers.

Conclusions The origin of the widely different and anomalous results reported for the surface energy of chitosan has been thoroughly explained by this study, which showed it to be associated with nonpolar impurities in even the best-quality commercial samples of this renewable resource. Given the rapidly growing interest in the development and applications of materials based on chitosan, the clarification of such a negatively relevant ambiguity represents an important contribution to this realm.

Acknowledgment. The authors thank Mahtani Chitosan Pvt. Ltd. (India) and Norwegian Chitosan AS. (Norway) for the generous gifts of chitosans and chitin. A. G. Cunha and S. C. Fernandes thank the Fundação para a Ciência e a Tecnologia (Portugal) for a Ph.D. grant (SFRH/BD/31134/2006) and a scientific research grant (SFRH/BI/33050/2007), respectively.

References and Notes (1) Gandini, A.; Belgacem, M. N. Cellulose Fibre Reinforced Polymer Composites; Old City Publishing: Philadelphia, PA, 2007; Chapter 3. (2) Angellier, H.; Molina-Boisseau, S.; Belgacem, M. N.; Dufresne, A. Langmuir 2005, 21, 2425–2433. (3) Belgacem, M. N.; Blayo, A.; Gandini, A. J. Colloid Interface Sci. 1996, 182, 431–436. (4) (a) Bortolotti, M.; Brugnara, M.; Della Volpe, C.; Maniglio, D.; Siboni, S. J. Colloid Interface Sci. 2006, 296, 292–308. (b) Lewin, M.; MeyMarom, A.; Frank, R. Polym. AdV. Technol. 2005, 16, 429–441. (5) Rinaudo, M. Prog. Polym. Sci. 2006, 31, 603–632. (6) Wong, D. W. S.; Gastineau, F. A.; Gregorski, K. S.; Tillin, S. J.; Pavlath, A. E. J. Agric. Food Chem. 1992, 40, 540–544. (7) Rillosi, M.; Buckton, G. Pharm. Res. 1995, 12 (5), 669–675. (8) Yamatoto, H.; Nishida, A.; Ohkawa, K. Colloids Surf., A 1999, 149, 553–559. (9) Amaral, I. F.; Granja, P. L.; Melo, L. V.; Saramago, B.; Barbosa, M. A. J. Appl. Polym. Sci. 2006, 102 (1), 276–284. (10) Sionkowska, A.; Kaczmarek, H.; Wisniewski, M.; Skopinska, J.; Lazare, S.; Tokarev, V. Surf. Sci. 2006, 600, 3775–3779. (11) Ma, Y.; Jia, Y.-L.; Shang, Y.-L.; Liao, F.; Li, J.-R.; Zhang, S.; Zhang, O. J. Appl. Polym. Sci. 2007, 105, 2427–2432. (12) Belgacem, M. N.; Gandini, A. In Interfacial Phenomena in Chromatography; Pefferkorn, E., Ed.; Marcel Dekker: New York, 1999; Chapter 2. (13) Rinaudo, M.; Milas, M.; Le Dung, P. Int. J. Biol. Macromol. 1993, 15, 281–285. (14) Rinaudo, M.; Pavlov, G.; Desbrières, J. Polymer 1999, 40, 7029–7032. (15) Freire, C. S. R.; Silvestre, A. J. D.; Pascoal Neto, C. Holzforschung 2002, 56, 143–149. (16) Della Volpe, C.; Siboni, S. J. Colloid Interface Sci. 1997, 195, 121– 136. (17) Owens, D. K.; Wendt, R. C. J. Appl. Polym. Sci. 1969, 13 (8), 1741– 1747. (18) Neville, A. C. Biology of the Arthropod Cuticle; Springer-Verlag: Berlin, 1975; pp 7–60. (19) Nair, K. G.; Dufresne, A.; Gadini, A.; Belgacem, M. N. Biomacromolecules 2003, 4, 1835–1842.

BM701199G