Production of Chitin from Penaeus vannamei By-Products to Pilot Plant Scale Using a Combination of Enzymatic and Chemical Processes and Subsequent Optimization of the Chemical Production of Chitosan by Response Surface Methodology

The waste generated from shrimp processing contains valuable materials such as protein, carotenoids, and chitin. The present study describes a process at pilot plant scale to recover chitin from the cephalothorax of Penaeus vannamei using mild conditions. The application of a sequential enzymatic–acid–alkaline treatment yields 30% chitin of comparable purity to commercial sources. Effluents from the process are rich in protein and astaxanthin, and represent inputs for further by-product recovery. As a last step, chitin is deacetylated to produce chitosan; the optimal conditions are established by applying a response surface methodology (RSM). Under these conditions, deacetylation reaches 92% as determined by Proton Nuclear Magnetic Resonance (1H-NMR), and the molecular weight (Mw) of chitosan is estimated at 82 KDa by gel permeation chromatography (GPC). Chitin and chitosan microstructures are characterized by Scanning Electron Microscopy (SEM).


Introduction
Prawns and shrimp are the second fish product marketed at world level, and constitute 8% of the total value of internationally traded fish products. In aquaculture, Penaeus vannamei is the top cultured species. Because of its great commercial value, the food industry processes a significant amount of these crustaceans, in many cases involving the removal of the exoskeleton of the tail and the cephalothorax. As a result, about 40-45% of the whole animal is considered by-products. This low yield combined with the commercial importance of this species results in high amounts of waste generated by its industrial processing. Therefore, alternatives for the treatment of these by-products are necessary, ideally scaled at pilot plant level to assess their viability.
The by-products generated contain compounds of interest, such as protein, pigments, and chitin. In particular, the exoskeleton represents the main source of chitin and chitosan [1]. The former is the second most abundant natural biopolymer after cellulose, and is one of the major sources of surface pollution in coastal areas. The biosynthesis of chitin involves the linkage of N-acetylglucosamine

Pilot Plan Production of Chitin
The moisture, ash, protein, and lipid contents of shrimp by-products were 76%, 4.76%, 55% and 1.6% respectively ( Table 1). The compositional values are similar to the samples described in previous reports [28,29]. The process applied in the present work for the industrial production of chitin ( Figure 1) is an improved modification of the alkaline-enzymatic methods reported by other authors [28,[30][31][32]. These improvements are based on the reduction of the concentrations of alkaline (1 M NaOH) and acid reagents (0.4 M HCl), time (6 h for the whole process), and processing temperature (less than 65 • C), when compared to the more extreme conditions described in the cited works. The chitin thus obtained presents similar purity levels in terms of the ash content, the degree of acetylation, and the infrared fingerprint, which demonstrates that milder conditions are sufficient to obtain high quality chitin (Table 1). The degree of N-acetylation can be derived from the carbon/nitrogen ratio (C/N), which is a fundamental parameter for discriminating between chitin and chitosan. Theoretically, that ratio is 6.9 for fully acetylated chitin, and 5.1 for fully deacetylated chitosan. The degree of acetylation (DA) was calculated by the following equation [33]: It resulted in a DA of 96%, and C/N = 6.69 (Table 1) for the dry solids obtained at the end of the process line, which confirms its identity as chitin. Figure 2 shows the comparison of the infrared spectra (IR) of the final product obtained from commercial chitin (A), and P. vannamei by-products (B). The identification of the characteristic chitin bands OH group at 3440-3447 cm −1 ; NH group at 3265-3266 cm −1 ; the CH band at 2927-2933 cm −1 ; and the amide I band at 1630 cm −1 [5,34], as well as the similar pattern, confirms chitin presence in the final product, in agreement with the DA results. In addition, the ash percentage was low (1.44%), which reveals an almost complete demineralization of the sample under the conditions used in the process, and is consistent with the IR spectrum, free from interfering bands.
The scanning electron microscope (SEM) image of chitin displays the conventional surface with irregular shapes and heterogeneous size, observing asymmetric particles or microfibrils ( Figure 3A) [35]. The linear sequence of N-acetylglucosamine monomers linked by covalent β-(1→4) glycosidic bonds is considered the primary structure of chitin. These chains are joined together by multiple hydrogen bonds in three-dimensional microfibrils, which constitutes their secondary structure. Moreover, microfibrils bind to proteins, and create the tertiary structure of chitin. The percentage of crystalline material in chitin (crystallinity index, ICr), calculated according to Segal et al. [36], is 88.1 ± 3.7%. The white flakes of chitin produced at the pilot plant are depicted in Figure 3B. The final yield of chitin recovery reaches 30%.  The degree of N-acetylation can be derived from the carbon/nitrogen ratio (C/N), which is a fundamental parameter for discriminating between chitin and chitosan. Theoretically, that ratio is 6.9 for fully acetylated chitin, and 5.1 for fully deacetylated chitosan. The degree of acetylation (DA) was calculated by the following equation [33]: It resulted in a DA of 96%, and C/N = 6.69 (Table 1) for the dry solids obtained at the end of the process line, which confirms its identity as chitin. Figure 2 shows the comparison of the infrared spectra (IR) of the final product obtained from commercial chitin (A), and P. vannamei by-products (B). The identification of the characteristic chitin bands OH group at 3440-3447 cm −1 ; NH group at 3265-3266 cm −1 ; the CH band at 2927-2933 cm −1 ; and the amide I band at 1630 cm −1 [5,34], as well as the similar pattern, confirms chitin presence in the final product, in agreement with the DA results. In addition, the ash percentage was low (1.44%), which reveals an almost complete demineralization of the sample under the conditions used in the process, and is consistent with the IR spectrum, free from interfering bands.
The scanning electron microscope (SEM) image of chitin displays the conventional surface with irregular shapes and heterogeneous size, observing asymmetric particles or microfibrils ( Figure 3A) [35]. The linear sequence of N-acetylglucosamine monomers linked by covalent β-(1→4) glycosidic bonds is considered the primary structure of chitin. These chains are joined together by multiple hydrogen bonds in three-dimensional microfibrils, which constitutes their secondary The results from the elemental analysis-including heavy metals-are also summarized in Table 1.
The low values obtained agree with the low ash content in the final chitin samples. The Pb, Hg, and Cd levels are consistent with data previously reported [37], and comply with the values established by European Union legislation [38].
Leaving aside unclear patent references, this is the first approach to the scaled-up production of chitin from waste generated in the industrial processing of P. vannamei ( Figure 1A). Additionally, the effluents generated in the process can either represent inputs for other valorization processes, or be safely discarded ( Figure 1B). The first effluent and secondary by-product consists of the protein trimmings obtained in the initial water washings, which could be an interesting substrate, after filtration, for the production of fish protein hydrolysates (FPH). Its high protein content (more than 60%) and the presence of essential amino acids make this waste a candidate to be enzymatically processed to be applied as a high quality nutrient in aquaculture feed [39][40][41]. Leaving aside unclear patent references, this is the first approach to the scaled-up production of chitin from waste generated in the industrial processing of P. vannamei ( Figure 1A). Additionally, the effluents generated in the process can either represent inputs for other valorization processes, or be safely discarded ( Figure 1B). The first effluent and secondary by-product consists of the protein trimmings obtained in the initial water washings, which could be an interesting substrate, after filtration, for the production of fish protein hydrolysates (FPH). Its high protein content (more than 60%) and the presence of essential amino acids make this waste a candidate to be enzymatically processed to be applied as a high quality nutrient in aquaculture feed [39][40][41]. Subsequent to the enzyme proteolysis of shells and hydrolysate filtration, two fractions are recovered: (a) solids containing chitin, mineral salts, pigments, and remaining protein, later used as a substrate for chitosan production; (b) an effluent rich in soluble proteins and pigments, which can be processed with vegetal oil for the recovery of pigments-mainly astaxanthin-in the oil fraction [42,43]. The later remaining aqueous fraction contains a remarkable amount of protein (6.95 ± 0.22 g/L), which is possible to separate after centrifugation. Recently, the extraction of pigments from the wastewaters of the industrial processing of P. vannamei shells has been extensively reported [44]. Such approaches are mainly based on vegetal oil extraction at low-medium temperatures (40-70 • C), sometimes combined with previous concentration steps by ultrafiltration membranes and hydrolysis assisted by commercial proteases. Regarding the protein effluent, it may be an excellent organic nitrogen source (marine peptones) for the culture of several bacteria (e.g., marine probiotics, lactic acid bacteria, etc.). This was confirmed in a recent study using effluents obtained from the enzymatic β-chitin production from the pens of Illex argentinus squid [12]. Subsequent to the enzyme proteolysis of shells and hydrolysate filtration, two fractions are recovered: (a) solids containing chitin, mineral salts, pigments, and remaining protein, later used as a substrate for chitosan production; (b) an effluent rich in soluble proteins and pigments, which can be processed with vegetal oil for the recovery of pigments-mainly astaxanthin-in the oil fraction [42,43]. The later remaining aqueous fraction contains a remarkable amount of protein (6.95 ± 0.22 g/L), which is possible to separate after centrifugation. Recently, the extraction of pigments from the wastewaters of the industrial processing of P. vannamei shells has been extensively reported [44]. Such approaches are mainly based on vegetal oil extraction at low-medium temperatures (40-70 °C), sometimes combined with previous concentration steps by ultrafiltration membranes and hydrolysis assisted by commercial proteases. Regarding the protein effluent, it may be an excellent organic nitrogen source (marine peptones) for the culture of several bacteria (e.g., marine probiotics, lactic acid bacteria, etc.). This was confirmed in a recent study using effluents obtained from the enzymatic β-chitin production from the pens of Illex argentinus squid [12].
The effluents from acid demineralization and alkaline hydrolysis are mixed to achieve neutral pH, reducing thereby the difficulty and the cost of managing strong acid and alkaline solutions. Furthermore, this effluent is a good source of calcium (in carbonate and phosphate forms) and sodium (as chloride), which are of increasing interest as food supplements of marine origin [45,46]. Finally, the bleaching solution, with a slight orange coloration, can be discharged (or employed for washing the pilot plant) without previous depuration.

Optimization of Chitosan Production
The habitual procedure for chitin deacetylation is mediated by the alkaline hydrolysis of the acetyl groups at high temperature. Although it is not a particularly environmentally friendly process, enzymatic alternatives using deacetylases are not a realistic option nowadays, since no commercial deacetylases are available. Recent studies of fermentative deacetylation using fungus, such as Mucor rouxii, have shown promising results, but crystalline chitin must be necessarily pretreated prior to enzyme hydrolysis to improve the accessibility of the enzyme to the acetyl groups [47]. The applications of the mentioned deacetylases, directly or by fungus fermentation, are not yet used at an industrial scale.
In the present work, a two-factor rotatable design has been performed to study the effect of two independent variables (time and alkali concentration) in the maximization of chitosan production from chitin of P. vannamei waste ( Table 2). The main response evaluated is the degree of deacetylation (DD) quantified by Proton Nuclear Magnetic Resonance ( 1 H -NMR). In this context, 1 H-NMR spectroscopy is the technique employed to determine the structural composition and purity of the polysaccharide (chitosan), based on the intensities of the 1 H absorption peaks of the The effluents from acid demineralization and alkaline hydrolysis are mixed to achieve neutral pH, reducing thereby the difficulty and the cost of managing strong acid and alkaline solutions. Furthermore, this effluent is a good source of calcium (in carbonate and phosphate forms) and sodium (as chloride), which are of increasing interest as food supplements of marine origin [45,46]. Finally, the bleaching solution, with a slight orange coloration, can be discharged (or employed for washing the pilot plant) without previous depuration.

Optimization of Chitosan Production
The habitual procedure for chitin deacetylation is mediated by the alkaline hydrolysis of the acetyl groups at high temperature. Although it is not a particularly environmentally friendly process, enzymatic alternatives using deacetylases are not a realistic option nowadays, since no commercial deacetylases are available. Recent studies of fermentative deacetylation using fungus, such as Mucor rouxii, have shown promising results, but crystalline chitin must be necessarily pretreated prior to enzyme hydrolysis to improve the accessibility of the enzyme to the acetyl groups [47]. The applications of the mentioned deacetylases, directly or by fungus fermentation, are not yet used at an industrial scale.
In the present work, a two-factor rotatable design has been performed to study the effect of two independent variables (time and alkali concentration) in the maximization of chitosan production from chitin of P. vannamei waste ( Table 2). The main response evaluated is the degree of deacetylation (DD) quantified by Proton Nuclear Magnetic Resonance ( 1 H -NMR). In this context, 1 H-NMR spectroscopy is the technique employed to determine the structural composition and purity of the polysaccharide (chitosan), based on the intensities of the 1 H absorption peaks of the corresponding chitosan chemical shifts (in ppm), with the residual HOD solvent signal as reference. Two examples from the factorial design are depicted in Figure 4. For a highly deacetylated sample (left) obtained after reaction with 50% NaOH for 24 h, the 1 H-NMR spectrum shows the common peaks of N-acetyl at 2.09 ppm and H2 of glucosamine (GlcN) at 3.20 ppm. The unresolved signals for protons H2-H6 of N-acetylglucosamine (GlcNAc) and H3-H6 of GlcN appear in the region 3.5-4.0 ppm. The signal of AcOH, tipically seen at 2.11 ppm is absent, indicating very little hydrolysis in the dissolved samples. Similar profiles were observed for chitosan produced under other experimental conditions, such as 12.5 h/50% NaOH (spectra not shown). In these cases, DD values are determined from the relative integrals of acetyl (N-acetyl and AcOH) and combined H2-H6 protons of, GlcN and GlcNAc [48,49].
for protons H2-H6 of N-acetylglucosamine (GlcNAc) and H3-H6 of GlcN appear in the region 3.5-4.0 ppm. The signal of AcOH, tipically seen at 2.11 ppm is absent, indicating very little hydrolysis in the dissolved samples. Similar profiles were observed for chitosan produced under other experimental conditions, such as 12.5 h/50% NaOH (spectra not shown). In these cases, DD values are determined from the relative integrals of acetyl (N-acetyl and AcOH) and combined H2-H6 protons of, GlcN and GlcNAc [48,49]. In the spectrum on the right, obtained at 20.6 h/35.9% NaOH, the picture is quite different. The typical signals for chitosan described above show very low intensities, corresponding to also low quantities of chitosan in solution. This indicates that only minute amounts of chitin were converted to chitosan, probably because the conditions applied were too mild to render the material soluble, remaining for the most part as chitin. In this context, the weakness of the signals do not allow to calculate values for DD, so in these cases the equation proposed by Ottøy et al. [50] is used instead.
The DD values obtained by both methods are fitted by the polynomial model (2), followed by a statistical analysis using response surface methodology ( Table 2).
The outcomes of the multivariate analysis indicate that the statistical signification of coefficients, evaluated by a Student t-test (α = 0.05), is only relevant for the quadratic term of the NaOH and the joint effect of time and alkali. The value of the coefficient of determination adjusted is remarkable ( 2 adj R = 0.817), revealing a good correlation between experimental and predicted data, and indicating that the variability of the DD is explained satisfactorily by the second order equation (Table 2). The four F-Fisher tests evaluated are significant, confirming the robustness of the polynomial equation (2).  In the spectrum on the right, obtained at 20.6 h/35.9% NaOH, the picture is quite different. The typical signals for chitosan described above show very low intensities, corresponding to also low quantities of chitosan in solution. This indicates that only minute amounts of chitin were converted to chitosan, probably because the conditions applied were too mild to render the material soluble, remaining for the most part as chitin. In this context, the weakness of the signals do not allow to calculate values for DD, so in these cases the equation proposed by Ottøy et al. [50] is used instead.
The DD values obtained by both methods are fitted by the polynomial model (2), followed by a statistical analysis using response surface methodology ( Table 2). Table 2. Results of the rotatable second-order design of the combined effect of time (t) and alkali concentration (NaOH) on the production of chitosan (as degree of deacetylation) according to Equation    The outcomes of the multivariate analysis indicate that the statistical signification of coefficients, evaluated by a Student t-test (α = 0.05), is only relevant for the quadratic term of the NaOH and the joint effect of time and alkali. The value of the coefficient of determination adjusted is remarkable (R 2 adj = 0.817), revealing a good correlation between experimental and predicted data, and indicating that the variability of the DD is explained satisfactorily by the second order equation (Table 2). The four F-Fisher tests evaluated are significant, confirming the robustness of the polynomial equation (2). DD = 87.37 + 12.69 t + 4.54 NaOH + 20.25 t NaOH − 14.64 t 2 − 31.74 NaOH 2 (2) Figure 5 displays the experimental data of the DD, and the predicted surfaces generated by the model. Based on this equation, the optimal conditions calculated by numerical derivation [51,52] are NaOH = 53.8% and t = 17.5 h. The degree of deacetylation calculated in these values is 92%. Higher Mw chitosan (150 kDa) has been obtained working at higher temperatures and lower NaOH concentrations using the shrimp wastes of non-defined species [52]. A structural analysis by SEM was also performed in four samples ( Figure 6). The top images (A and B) also showed irregular patterns, with the presence of microfibrils indicating a perfect crystal structure as a result of a regular packing of the chains corresponding to chitin. In fact, when comparing these images with the obtained crystallinity index (86% and 85%) and the values of the DD (20% and 19%), it is evident that they are chitin. In the bottom images (C and D), less crystallinity was observed, in agreement with the data of the DD (84% and 85%, respectively). The molecular weights (Mw) of various samples from the previous factorial design were also determined, by gel permeation chromatography (GPC). The values of the Mws are different depending on the experimental conditions applied, ranging from 58 to 82 kDa for the more deacetylated samples to 91-93 kDa for the acetylated ones (basically non-deacetylated chitin).
Higher Mw chitosan (150 kDa) has been obtained working at higher temperatures and lower NaOH concentrations using the shrimp wastes of non-defined species [52]. A structural analysis by SEM was also performed in four samples ( Figure 6). The top images (A and B) also showed irregular patterns, with the presence of microfibrils indicating a perfect crystal structure as a result of a regular packing of the chains corresponding to chitin. In fact, when comparing these images with the obtained crystallinity index (86% and 85%) and the values of the DD (20% and 19%), it is evident that they are chitin. In the bottom images (C and D), less crystallinity was observed, in agreement with the data of the DD (84% and 85%, respectively). Figure 5. Experimental data (points) and theoretical response surfaces describing the joint effects of alkali and processing time on chitosan production from P. vannamei chitin.
Higher Mw chitosan (150 kDa) has been obtained working at higher temperatures and lower NaOH concentrations using the shrimp wastes of non-defined species [52]. A structural analysis by SEM was also performed in four samples ( Figure 6). The top images (A and B) also showed irregular patterns, with the presence of microfibrils indicating a perfect crystal structure as a result of a regular packing of the chains corresponding to chitin. In fact, when comparing these images with the obtained crystallinity index (86% and 85%) and the values of the DD (20% and 19%), it is evident that they are chitin. In the bottom images (C and D), less crystallinity was observed, in agreement with the data of the DD (84% and 85%, respectively).

Raw Materials and Extraction of Chitin
The shrimp (Penaeus vannamei) by-products, mainly exoskeleton of cephalothorax, were kindly provided by Pescanova S.A. (Vigo, Spain) and stored at −20 • C until use. The purification of the chitin was performed in a pilot plant built for this purpose by Grupo Josmar S.L. (Pontevedra, Spain). This plant consists of three conical stainless steel reactors (each one 500 L), fully equipped with controllers of pH, temperature, and agitation as well as auxiliary equipment (transfer pumps, membrane filters, spray-dryer, etc.) designed for the main treatment of crustacean by-products to produce chitin (Figure 1, top): (a) enzymatic proteolysis (by Alcalase); (b) acid demineralization (by HCl); and (c) alkaline hydrolysis (using NaOH) and chemical bleaching (by NaClO). The commercial chitin from shrimp shells (reference C7170) was purchased from Sigma-Aldrich SA (St. Louis, MO, USA).
The chitin production line was performed (in duplicate) using batches of 50 kg of shrimp by-products (Figure 1, bottom). Initially, these materials were ground (0.5-1.5 cm) and washed to eliminate protein trimmings from the rest of the crustacean's body. Subsequently, the shrimp exoskeleton was hydrolysed by a commercial protease, Alcalase 2.4L (Novozymes S.A.), at an enzyme concentration of 0.5 (v/w of substrate) in a solid:liquid ratio of 1:5 (w/v) with distilled water at 55 • C and pH = 8. A constant pH was maintained for 2 h by the addition of 1M NaOH as required. At the end of the enzyme proteolysis, the solids were separated from the protein effluent by filtration (500 µm), and passed on to the next step: acid treatment with 0.4 M HCl using solid:liquid ratio of 1:10 (w/v) for 1 h under agitation (200 rpm) at room temperature. This second stainless steel reactor is internally covered with an epoxy material to prevent a strong acid attack. After the acid treatment, the effluent was filtered and the solid was washed until neutral pH. The acid treatment was performed twice. The last stage was an alkaline treatment with NaOH 1M using a ratio 1:5 (w/v) for 2 h at 65 • C, then the solid was washed to neutrality and treated with 1% sodium hypochlorite (ratio 1:10 w/v) for 15 min at room temperature. Finally, the solution was filtered and the resulting solids (chitin) were dried (by drying oven) at 60 • C for 24 h and milled.

Compositional Characterization of P. vannamei By-Products and Produced Chitin
Different methods were applied to characterize both the shrimp cephalotorax and the purified chitin: the water content was determined by drying at 105 • C until constant weight; the total nitrogen was analyzed according to the Kjeldahl procedure [53]; the ash content was assessed according to the AOAC protocol [53]; and all of the lipids were extracted by the Bligh and Dyer method [54].
In chitin, the presence of ions was quantified by inductively coupled plasma mass spectrometry (ICP-MS), and the heavy metal content by atomic absorption spectroscopy (AAS). The IR measurements were performed at room temperature on a Thermo Nicolet 6700 FTIR infrared spectrometer. The samples were prepared in a KBr pellet at a ratio sample/KBr of 1:10, and measured in transmission mode in the 400 to 4000 cm −1 range using a DTGS type detector (deuterated triglycerin sulfate) and a divider of KBr beam. The C/N ratio was determined by combustion at 1200 • C in a LECO CN-2000 analyzer.

Experimental Design for the Production of Chitosan and Statistical Analysis
A second order rotatable design with quintuple replication in the centre of the experimental domain was performed [55], in order to maximize the deacetylation of the chitin to produce chitosan ( Table 1). The combined effect of NaOH (in %) and the time of hydrolysis (t, in h) on chitosan production was studied, using as dependent variables (responses) the degree of deacetylation (DD) and the solubility (S) in acetic acid of the resulting solids. The experimental conditions varied between 1 and 24 h for t and 30-70% for NaOH; the solid:liquid ratio (1:20) and temperature (90 • C) were kept constant. The experimental units were 300 mL Erlenmeyer flasks with 5 g of chitin and 100 mL of NaOH solution, each unit fully sampled at the established time intervals. The samples were centrifuged, the supernatant stored at −20 • C, and the solid dried at 60 • C for 24 h and stored in sealed bags for further characterization.
The experimental results of the factorial designs were modelled by second-order polynomial equations, as: where Y represents the response to be modelled; b 0 is the constant coefficient, b i is the coefficient of linear effect; b ij is the coefficient of interaction effect; b ii the coefficients of squared effect; n is the number of variables; and X i and X j define the independent variables.
The goodness-of-fit was established as the adjusted determination coefficient (R 2 adj ), the statistical significance of the coefficients was verified by means of the Student t-test (α = 0.05), and the model consistency by the Fisher F test (α = 0.05), using the following mean squares ratios:

Chemical and Structural Characterization of Chitin and Chitosan
Scanning Electron Microscopy (SEM) images were obtained on a JEOL JSM-6400 microscope (Westmont, IL, USA), with acceleration voltage of up to 40 kV and resolution of 36 Å. The samples were prepared on a metal slide with a double-sided adhesive sticker, and covered with a thin layer of gold.
Crystallinity (DRX) was assessed with a PHILIPS X'PERT MPD automatic diffractometer (Almelo, The Netherlands), equipped with a goniometer PW3050 (θ-2θ), at a generator power of 45 kV and 40 mA. The measurements were made at room temperature with Cu Kα1 radiation (wavelength 1.54056 Å), with a graphite monochromator and confocalised geometry (Bragg-Brentano). The step size (2θ) of the measurements was set at 0.040 • , and the duration of the step at 1 s. The angular range studied was 5 • to 40 • 2θ. The percentage of crystalline material in chitin is expressed by a crystallinity index (ICr), determined according to the method proposed by Segal [36].
The deacetylation degree (DD) was determined by Proton Nuclear Magnetic Resonance ( 1 H-NMR) performed at room temperature on a Bruker Avance II spectrometer (Billerica, MA, USA) at a resonance frequency of 400 MHz. Mestrenova 10.0 software (Mestrelab Research, Santiago de Compostela, Spain) was used for spectral processing. Chitosan samples were dissolved in 0.056 M deuterated trifluoroacetic acid (TFA-d in D 2 O) at a concentration of 7 g/L.
The molecular weight (Mw) was measured by gel permeation chromatography (GPC). The chromatographic system consists of a Waters 625 LC System pump connected to an Ultrahydrogel column (300 × 7.8 mm, Waters, Milford, MA, USA), thermostated in an oven at 35 • C. The detectors were refractive index (Waters 2414), coupled in line with Evaporative Light Scattering (ELS Waters 2424). The mobile phase was 0.2M CH 3 COOH/0.15M CH 3 COONH 4 working under a flow rate of 0.5 mL/min and an injection volume of 20 µL. The chitosan samples were prepared at a concentration of 1 mg/mL in the same mobile phase solution. All of the solvents and solutions were filtered through 0.45 µm filters (Millipore, Billerica, MA, USA). The standard samples of chitosan were used to establish the molecular weight average of the analyzed samples.

Conclusions
The pilot plant production of chitin from P. vannamei by-products (cephalotorax shells) by a combination of enzymatic and chemical processes was performed. The characteristics of the chitin obtained were: 96% of acetylation degree; 88% of crystallinity; and a very low content of ashes, lipids, proteins, and heavy metals (trace in some cases). Using the response surface methodology, the joint effect of time and NaOH concentration were also studied. The optimal conditions of the deacetylation reaction were obtained at 53.8% of NaOH and 17.5 h of processing in order to produce chitosan with a DD of 92% and Mw of 82 kDa. This methodology and the predictive polynomial equation developed can be used for the production of chitosan with defined DD values. Further studies performing RSM should be done to produce chitosan from P. vannamei with tailored molecular weights.