Influence of Functionalization Degree on the Rheological Properties of Isocyanate-Functionalized Chitin-and Chitosan-Based Chemical Oleogels for Lubricant Applications

This work deals with the influence of functionalization degree on the thermogravimetric and rheological behaviour of NCO-functionalized chitosanand chitin-based oleogels. Chitosan and chitin were functionalized using different proportions of 1,6-hexamethylene diisocyanate (HMDI) and subsequently dispersed in castor oil to promote the chemical reaction between the –NCO group of the modified biopolymer and the –OH group located in the ricinoleic fatty acid chain of castor oil, thus resulting in different oleogels with specific thermogravimetric and rheological characteristics. Biopolymers and oleogels were characterized through Fourier transform infrared spectroscopy (FTIR) and thermogravimetric analysis (TGA). Small-amplitude oscillatory shear (SAOS) measurements were performed on the oleogels. Oleogels presented suitable thermal resistance, despite the fact that the inclusion of HMDI moieties in the polymer structure led to a reduction in the onset temperature of thermal degradation. The insertion of low amounts of HMDI in both chitin and chitosan produces a drastic reduction in the values of oleogel viscoelastic functions but, above a critical threshold, they increase with OPEN ACCESS Polymers 2014, 6 1930 the functionalization degree so that isocyanate functionalization results in a chemical tool to modulate oleogel rheological response. Several NCO-functionalized chitosanand chitin-based oleogel formulations present suitable thermal resistance and rheological characteristics to be proposed as bio-based alternatives to traditional lubricating greases.

In the frame of the general tendency to promote the replacement of non-renewable raw materials by renewable resources in order to avoid the impact that process technologies and products of different industrial sectors cause on the environment, the lubricant industry is especially involved [33,34].In this sense, different alternatives based on modifications of vegetable oils have been proposed to replace mineral and synthetic lubricant base oils [35][36][37][38][39][40].Although vegetable oils are being increasingly used as bio-based lubes, the substitution of traditional thickening agents, such as metallic soaps and polyureas, in lubricating greases by others derived from renewable resources is a much more complicated task.For instance, some limitations regarding physical and mechanical stabilities and thermo-rheological behaviour of previously mentioned cellulosic derivatives-based gel-like dispersions [27][28][29] were pointed out as a consequence of a certain degree of chemical incompatibility between oil and such polar biopolymers.Most of these problems can be overcome by inducing the formation of chemical gels, for instance by dispersing reactive isocyanate-functionalized methylcellulose in the vegetable oil medium [41,42].Following the same procedure and based on some preliminary studies [43], the possibility of using NCO-functionalized chitin and chitosan to chemically gel castor oil is explored in this work.As well known, chitin is one of the most abundant organic materials, being second only to cellulose in the amount produced annually by biosynthesis.It is an important constituent of the exoskeleton in animals especially in crustaceans, molluscs and insects and the principal fibrillar polymer in the cell wall of certain fungi [44,45].Chitosan is typically obtained by deacetylation of chitin under alkaline conditions in a variety of useful forms, characterized by displaying interesting properties such as biocompatibility or biodegradability [46][47][48], which also produces non-toxic, non-immunogenic and non-carcinogenic degradation products [49,50].Nowadays, there is a growing interest in the chemical modification of chitosan in order to improve its solubility or the cross-linking ability in different media and therefore widen its applications [51][52][53][54].The reaction between chitosan and diisocyanate compounds has been previously promoted to produce bio-polyurethanes for the development of molecular carriers in the biomedical and environmental fields [55], beads for cholesterol absorption [56] or materials with controlled thermal or elastomeric properties [57,58].As a continuation of the above mentioned research on the development of oleogels for lubricant purposes and taking into account the interest of these biopolymers, the effect of NCO-functionalization degree of chitin and chitosan on the rheological properties and thermal degradation of resulting chemical oleogels has been studied in this work.

General Methods
Functionalization reactions were performed in flasks that were previously flame-dried under a positive flow of argon to eliminate surface moisture.HMDI was stored at 4 °C and handled under argon atmosphere.Solvents were purified according to standard literature techniques and stored under argon.Toluene was freshly distilled immediately prior to use from sodium/benzophenone and strictly deoxygenated for 30 min under argon.Reagents were purchased at the higher commercial quality and used without further purification, unless otherwise stated.Characterization tests were performed at least in duplicate.

Functionalization Reaction of Chitosan and Chitin with HMDI
The functionalization reactions of chitosan and chitin with HMDI were carried out according to the protocol described in preliminary studies [43].Chitosan or chitin was added to a bottom round flask with toluene while stirring at room temperature until it became a suspension.Then, triethylamine (Et 3 N) and HMDI were also added to the system, the last one drop-wisely.The solution was vigorously stirred at room temperature for 24 h.Afterwards, the mixture was vacuum filtered and the resulting product was immediately used to prepare the oleogels.The amount of HMDI, Et 3 N and toluene was modified with the aim of achieving different functionalization degrees of -OH and -NH 2 groups in the chitosan and chitin derivatives synthesized.The different molar ratios of HMDI, Et 3 N and toluene with respect to free -OH and -NH 2 groups in the non-functionalized chitin and chitosan monomers are listed in Table 1.

Preparation of Oleogels
Oleogel samples were prepared in an open vessel (batches of 60 g), using a controlled-rotational speed mixing device (70 rpm) RW 20 (Ika, Staufen, Germany), equipped with an anchor impeller to disperse the different biopolymers in the oil.Functionalized biopolymers were slowly added to the castor oil at different concentrations (see Table 2).The mixing process was maintained at 70 rpm for 24 h at room temperature.Finally, the resulting dispersion was homogenized with an Ultra-Turrax T50 (Ika) rotor-stator turbine, at 8800 rpm for 1 min.Table 2 also shows the codes applied for each oleogel sample.

Thermogravimetric Analysis (TGA)
Measurements of mass loss versus temperature were carried out by using a Thermogravimetric analyser, model Q-50 (TA Instruments, New Castle, DE, USA) under N 2 purge.Typically, 5-10 mg of sample were placed on a platinum pan and heated from 30 °C to 600 °C, at 10 °C•min −1 .

Fourier Transform Infrared (FTIR) Spectroscopy
FTIR spectra were obtained with a Digilab FTS3500ARX (Varian, Palo Alto, CA, USA) apparatus.Chitin and chitosan samples were prepared as KBr pellets, whereas a small drop of oleogels was placed between two KBr disks (32 mm × 3 mm).In both cases, the set was placed into an appropriate sample holder.The spectra were obtained in a wavenumber range of 400-4000 cm −1 , at 4 cm −1 resolution, in the transmission mode.The degree of deacetylation provided in Section 2.1 for original

Rheological Characterization
Rheological characterization of oleogels was carried out using controlled-stress rheometers (Gemini, Bohlin, Malvern Instruments Ltd., Worcestershire, United Kingdom, and Rheoscope, ThermoHaake, Karlsruhe, Germany).Small amplitude oscillatory shear (SAOS) tests were performed inside the linear viscoelastic region, using roughened plate-plate geometries (20 mm, 25 mm and 35 mm, and 1 mm gap), in a frequency range of 0.03-100 rad•s −1 , at 25 °C.Measurements were done one day, one week, one month and four months after oleogel preparation.Average molecular weight of original chitin and chitosan samples were determined by means of viscosity measurements according to the methods and Mark-Houwink-Sakurada equation constants provided by Poirier and Charlet [60] and Takahashi et al. [61], respectively.

Spectroscopic Characterization
Chemical modification of chitin and chitosan after functionalization with different proportions of HMDI and the occurrence of further reaction with -OH groups of castor oil were evaluated by FTIR. Figure 1 shows the FTIR spectra for non-functionalized chitin (Figure 1a) and chitosan (Figure 1d), selected NCO-functionalized samples, CTIN-1 (Figure 1b) and CSAN-1 (Figure 1e), and the corresponding oleogels prepared with these biopolymers and castor oil, CTIN-1-25 (Figure 1c) and Polymers 2014, 6 CSAN-1-25 (Figure 1f).FTIR spectra of functionalized biopolymers were in concordance with the proposed chemical structure previously reported [43].Table 3 lists the most relevant peaks detected in NCO-functionalized chitosan and chitin spectra.The reaction of chitin and chitosan with the -NCO groups of HMDI to form urethane linkages produces a significant reduction in the intensity of the band around 3200-3550 cm −1 , corresponding to the -OH group in the molecules.The higher HMDI content, the more significant the reduction in this band is.The urethane bands (Figure 1b,e) can be detected at approximately 3400 cm −1 and 1740 cm −1 for N-H and C=O stretching vibrations in the urethane linkage, respectively, and 1580 cm −1 for the N-H bending vibration.Moreover, an intense peak at around 2270 cm −1 is noticed, confirming the presence of unreacted -NCO groups, as intended.The functionalization degree was controlled by means of carrying out the reaction at room temperature, the amount of HMDI and Et 3 N used and the reaction time (24 h in this case).Under these conditions, both a rapid attack of free -OH to -NCO groups of HMDI and an excessive cross-linking of -NCO groups present either in the unreacted HMDI or in the HMDI segments attached to the chitin or chitosan chains are significantly dampened.Thereby, the reaction can be controlled and allows obtaining selectively polymers with different degrees of functionalization.Finally, the reaction between the -NCO group of functionalized biopolymers and the -OH group of castor oil was also monitored by analyzing the FTIR spectra of oleogels (Figure 1c,f).As can be observed, the absorption band assigned to free -NCO groups at around 2271 cm −1 is negligible in the FTIR spectra of the oleogels, corroborating the occurrence of the reaction promoted.

Thermogravimetric Characterization
Thermal degradation of chitin, chitosan and derived NCO-functionalized compounds was studied by means of thermogravimetric analysis (TGA). Figure 2 shows TGA curves, in the form of weight loss derivative function versus temperature, for chitosan (Figure 2a) and chitin (Figure 2b) functionalized in different degrees, compared to those obtained with the corresponding starting materials, i.e., non-functionalized biopolymers and HMDI.The onset temperature for thermally-induced weight loss (T onset ), the maximum weight-loss rate temperature (T max ), the weight loss at the end of each decomposition step (ΔW) and the percentage of non-degraded residue were calculated from the thermogram of each sample studied (Table 4).Disregarding the initial weight loss between 40 °C and 80 °C, which is inferior to 8%, as a consequence of water evaporation, thermal degradation of non-functionalized chitosan and chitin takes place in one main stage.This main thermal event occurs in the range of 343-405 °C for chitin (Figure 2b) and is due to the degradation of the saccharide structure of the molecule, including the dehydration of saccharide rings and decomposition of the acetylated and deacetylated units [62].In the case of chitosan (Figure 2a), thermal degradation starts at lower temperature, indicating poorer thermal stability, but in a wider range (281-472 °C).This thermal range corresponds to the degradation peak of the part of the molecule that is deacetylated, showing a T max at 301 °C, and an overlapped small signal at around 390 °C, due to the degradation of the acetylated part of the molecule.
On the other hand, for NCO-functionalized biopolymers, the inclusion of HMDI segments in the polymer structure reduces their thermal stabilities, with marked impact in the highest functionalized polymer, as expected (Figure 2, Table 4).Thermal degradation curves of NCO-functionalized chitosans exhibit two main events.The first one at around 83-179 °C, with T max at 116-153 °C, corresponds to the loss of -NCO segments, which is corroborated with HMDI thermal degradation occurring at 137 °C.As can be observed in Table 4, the T onset and, logically, weight loss for this degradation stage increases with -NCO content.The second main degradation stage, as previously discussed for the non-functionalized sample, is related to the loss of the deacetylated part of the molecule of chitosan and takes place between 271 °C and 340 °C, with a T max at around 300 °C.The weight loss assigned to this thermal degradation process generally becomes more important as functionalization degree decreases (Table 4).A characteristic temperature corresponding to the acetylated part, expected at around 400 °C in biopolymers with lower quantity of HMDI segments in their structures (CSAN-3 and CSAN-4), also appears in the thermograms.Moreover, a final decomposition stage with T max at around 450 °C was detected, which can be assigned to excessively cross-linked chitosan fractions originating more rigid polyurethanes [63].Thermal degradation of NCO-functionalized chitin polymers (Figure 2b) is rather similar to that discussed for chitosan samples.The loss of HMDI segments was found from 83 °C to 154 °C, with maximum weight-loss rate temperatures between 119 °C and 135 °C, whereas the other main thermal degradation event takes place at around 390 °C, which corresponds to the loss of the acetylated units of the biopolymer.Finally, a third degradation stage occurs at 439-512 °C, similarly to that found in chitosan samples.Thermal decomposition of NCO-functionalized chitosan-and chitin-based oleogels is shown in Figure 3.The thermal degradation of non-functionalized chitosan oleogel takes place from 278 °C to 482 °C and three different thermal events with T max at 302 °C, 389 °C and 449 °C can be detected (Figure 3a, Table 5).The first one corresponds to chitosan degradation, the second one is mainly due to castor oil evaporation [30], the main oleogel component, and the last one refers to secondary products.For NCO-functionalized chitosan-based formulations (Figure 3a, Table 5), the thermal degradation occurs between 266 °C and 483 °C, also in three different stages.The first one, at 302-307 °C, is mainly noticed in those oleogels containing lower amounts of HMDI in their formulation (CSAN-3-30 and CSAN-4-30) and is attributed to the degradation of deacetylated chitin units.The second stage is again a consequence of castor oil evaporation.The temperature range for this event is expanded when the ratio of HMDI in the formulation increases (CSAN-1-30 and CSAN-2-30), which may be attributed to the urethane segments achieved during the reaction between functionalized chitosan and ricinoleic fatty chains.In the case of chitin-based oleogels (Figure 3b, Table 5), thermal decomposition always takes place in two clearly different stages.The main one can be observed between 324 °C and 410 °C and is a consequence of both thickener and castor oil degradations, which appear overlapped in one peak.The second minor event is observed at around 450 °C and refers to secondary products.Finally, it can be concluded that the disappearance of the first isocyanate-functionalized polymer degradation peak, due to HMDI decomposition, in the resulting oleogel confirms the chemical reaction promoted between modified polymers and the castor oil.

Rheology of Chitosan-and Chitin-Based Oleogels
Figure 4 shows the SAOS response, inside the linear viscoelastic range, of NCO-functionalized and non-functionalized chitosan and chitin-based oleogels, at 30% w/w concentration, as a function of the functionalization degree.In most cases, NCO-functionalized chitosan-based oleogels exhibit values of the storage modulus, G′, significantly higher than those found for the loss modulus, G″, in the whole frequency range studied (Figure 4a).This evolution, the so-called "plateau region", where G′ slightly increases with frequency and G″ displays a minimum, is typically found in highly entangled polymeric systems [64] and standard lubricating greases [65].Only when using the lowest amount of HMDI (CSAN-4-30), the plateau region appears to be not well developed, indicating the consecution of a weaker gel.The plateau region is also exhibited by some of the NCO-functionalized chitin-based oleogels such as CTIN-3-30 and CTIN-4-30 (Figure 4b).However, interestingly, a crossover between both viscoelastic functions was found at high frequency for the highly-functionalized chitin-based oleogels (CTIN-1-30 and CTIN-2-30).In these cases, this behaviour must be attributed to a higher degree of cross linking achieved during oleogel formation, resulting in a rheological response closer to the transition region of the mechanical spectrum.In general, as can be deduced from Figure 4, -NCO content in the functionalized biopolymers largely influences the values of the linear viscoelastic functions of resulting oleogels.In the case of chitosan, a high degree of functionalization (CSAN-1-30 and CSAN-2-30) increases the values of the SAOS functions with respect to those shown by the non-functionalized chitosan oleogel (CSAN-30), as a consequence of an extensive crosslinking.However, the values of the SAOS functions decrease when reducing the amount of HMDI incorporated to the biopolymer, yielding similar (CSAN-3-30) or lower (CSAN-4-30) G′ and G″ values than those exhibited by the non-functionalized chitosan-based oleogel (CSAN-30).On the contrary, for chitin-based oleogels, the most noticeable effect of HMDI segments is that they always produce lower values of the viscoelastic moduli than those shown by the non-functionalized chitin-based oleogel (CTIN-30).Apart from this, oleogels with high amounts of HMDI in their formulation (CTIN-1-30 and CTIN-2-30) exhibited the highest values of the linear viscoelastic  (rad/s) functions but, contrary to the expected, CTIN-3-30 showed lower values than CTIN-4-30.All these effects that functionalization degree exerts on the rheological response of chitin-and chitosan-based oleogels can be more noticeably summarized in Figure 5, where the values of G′ and the loss tangent (tanδ = G″/G′), at 1 rad•s −1 , are plotted versus the HMDI proportion applied in the functionalization of biopolymers.As can be observed, the insertion of low amounts of HMDI in both chitin and chitosan produces a drastic reduction in the values of SAOS functions, i.e., G′ 1 (G′ at 1 rad•s −1 ) especially for chitosan-based oleogels.However, above a critical threshold, SAOS functions increase with the functionalization degree.Interestingly, this increase in both G′ and G″ is associated to a lower relative elasticity in the case of chitin-based oleogels and higher in the case of chitosan-based ones, as can be seen in the tanδ vs. HMDI proportion plot (Figure 5).As mentioned above, only in the case of chitosan-based oleogels a high degree of functionalization makes possible to reach values of the viscoelastic moduli higher than those displayed by the non-functionalized biopolymer-based oleogel.However, as previously reported [27,42], it must be remarked that these physically stabilized gels present some limitations regarding their physical and mechanical stability, which can be overcome by mean of functionalization.As extensively investigated [66][67][68], typical G′ values in conventional lubricating greases of NLGI (National Lubricating Grease Institute) grade 1-2 range from 10 4 -10 5 Pa, around one order of magnitude higher than G″ values, depending on compositions and processing conditions.This particular rheological response has been obtained for several of the NCO-functionalized chitosan-and chitin-based oleogels studied, emphasizing that the degree of NCO-functionalization can be used as a chemical tool to modulate the rheological behaviour and consistency of these oleogels as desired.Rheological data collected in Figure 4a,b correspond to oleogels 1 and 4 months aged, respectively.As previously reported for NCO-functionalized methyl cellulose-based oleogels [42], these chemical gels require a certain ageing time to obtain stable rheological properties.Figure 6 shows the evolution of the linear viscoelastic functions with frequency, as a function of ageing time, for selected chitosan-(CSAN-3-25) and chitin-(CTIN-1-30) based oleogels.In most of the cases studied, the values of SAOS functions significantly increase during the first seven days of ageing and then slightly increase or remain almost constant after several months of ageing.These results can be explained on the basis of the slow reactivity of residual -NCO functional groups in a high viscous medium [69].However, interestingly, the loss tangent was almost unaffected (data not shown).Finally, as typically performed in conventional lubricating greases, the rheological response can be modified by changing the thickener concentration, in this case the functionalized biopolymers used as gelling agents.Figure 7 shows the mechanical spectra of functionalized chitosan-and chitin-based oleogels, containing different concentrations of the same biopolymer in their respective formulations.As can be observed, the values of the viscoelastic moduli increase with the concentration of both functionalized biopolymers.In the case of chitosan-based oleogels (Figure 7a), G′ exhibits values significantly higher than those found for G″, i.e., the plateau region, for concentrations ranging from 25%-35% w/w (CSAN-3-25, CSAN-3-30 and CSAN-3-35).Nevertheless, a liquid-like weak gel was obtained with a 15% biopolymer concentration (CSAN-3-15).For chitin-based oleogels (Figure 7b), almost three decades of both SAOS functions can be covered by modifying the thickener concentration between 15% and 30% w/w.In all cases, G′ is higher than G″ at low frequencies.However, a crossover between both linear viscoelastic functions was always observed, which is shifted to lower frequencies as biopolymer concentration decreases.

Conclusions
A systematic and controlled functionalization protocol for chitosan and chitin with HMDI resulted in several biopolymers functionalized in different degrees.The dispersion of these biopolymers in castor oil promoted the reaction between free -NCO segments of the biopolymer and -OH groups of the ricinoleic fatty acid chain thus providing different chemical oleogels with specific characteristics.
The resulting oleogels presented suitable thermal resistance despite the fact that the inclusion of HMDI segments in the polymer structure reduces their thermal stabilities.Oleogels thermal degradation basically takes place in one main stage which comprises the thermal decomposition of both the castor oil and the thickening agent.
Typical rheological behaviour found in conventional lubricating greases was achieved in different formulations studied, providing a cross-linked viscoelastic gel with a well-developed plateau region in the evolution of SAOS functions with frequency.The insertion of low amounts of HMDI in both chitin and chitosan produces a drastic reduction in the values of SAOS functions but, above a critical threshold, viscoelastic functions increase with the functionalization degree.Therefore, isocyanate functionalization appears as a chemical tool to modulate the rheological response.
The rheological properties of functionalized chitosan-and chitin-based oleogels evolve during several months of ageing, but mainly during the first week, due to the progress of the reaction between -NCO functional groups and castor oil -OH groups.Oleogel linear viscoelastic response is also highly dependent on the concentration of NCO-functionalized biopolymers.
Several NCO-functionalized chitosan-and chitin-based oleogel formulations present suitable thermal resistance and rheological characteristics to be proposed as bio-based alternatives to traditional lubricating greases.

Figure 2 .
Figure 2. Thermal degradation curves, under inert atmosphere, for NCO-functionalized chitosan (a) and chitin (b) samples and starting materials.

Figure 7 .
Figure 7. Frequency dependence of the storage, G′, and loss, G″, moduli for NCO-functionalized chitosan-(a) and chitin-(b) based oleogels, as a function of biopolymer concentration.

Table 4 .
TGA characteristic parameters for NCO-functionalized chitosan and chitin samples studied (main degradation stages in bold).

Table 5 .
TGA characteristic parameters for non-functionalized and NCO-functionalized chitosan and chitin-based oleogels studied (main degradation stages in bold).