Protein recovery by ultrafiltration during isolation of chitin from shrimp shells Parapenaeus longirostris

(1)

Protein recovery by ultra

ﬁltration during isolation of chitin from

shrimp shells Parapenaeus longirostris

M.S. Benhabiles

a

, N. Abdi

a

, N. Drouiche

a,c,*

_{, H. Lounici}

a

_{, A. Pauss}

b

_{, M.F.A. Goosen}

d

_{, N. Mameri}

b a_{National Polytechnic School of Algiers, B.P. 182-16200, El Harrach, Algiers, Algeria}

b_{University of Technology of Compiègne, Département Génie chimique, B.P. 20.509, 60205 Compiègne Cedex, France}

c_{Centre de Recherche en Technologie des Semi-conducteurs pour l’Energétique, 2 Bd Frantz Fanon, BP140, Alger, 7 Merveilles, 16000 Algeria} d_{Alfaisal University, Riyadh, Saudi Arabia}

a r t i c l e i n f o

Article history: Received 8 October 2012 Accepted 29 November 2012 Keywords: Chitin Protein hydrolysate Shell waste Ultraﬁltration

a b s t r a c t

In the food processing industry shrimp shells (Parapenaeus longorostris) have great commercial value because they are rich in chitin (24 wt%), protein (40 wt%), lipids, pigments andflavor compounds. In the present study protein recovery by ultrafiltration was examined during isolation of chitin from shrimp shell P. longirostris. Up to 96 wt% of the proteins could be removed (i.e. deproteinization) from the shrimp shells by incubating them in NaOH (2 N) over 2 h, at T¼ 45_{C, and solid to solvent ratio of 1:2 (w/v). A} solute rejection coefficient (R0) of 97% was obtained in the ultrafiltration process to recover proteins from

deproteinized shell waste water. The protein concentration process which was carried out beyond the criticalﬂux of 380 L/h.m2_{, at a trans-membrane pressure of 3 bars, and a tangential velocity of 5 m/s was}

found to reduce the hydrolysate volume by a factor of 2.4. Due to a reduction in organic matter in the efﬂuent, the chemical oxygen demand (COD) of the permeate was reduced by 87%.

1. Introduction

Total global production of captured and farmed shrimp reached 6 million tons in 2006 (FAO, 2009), with only 60% being used as food, leaving 2.3 million tones for non-food uses. Shrimp waste, which is rich in chitin, proteins, lipids, pigments and ﬂavor compounds, has potential commercial value in the food industry as well as a supplement in animal feed (Cira, Huerta, Hall, & Shirai, 2002). The primary reason for the limited commercial utilization of shrimp waste is its highly perishable nature; i.e. the waste quickly becomes colonized by spoilage organisms and can rapidly be transformed into both a nuisance and a public health hazard (Zakaria, Hall, & Shama, 1998). Under humid and high temperature conditions such as those in North Africa countries, decomposition of the waste occurs ever more rapidly. Hence, it is imperative that an inexpensive and effective method of preserving shrimp waste is found.

The economics of industrial processing of crustaceans can be improved by fully utilizing chitin and proteins which are found in

shell waste in amounts ranging from 14 to 32% and 18e42%, respectively (Heu, Kim, & Shahidi, 2003; Shahidi & Synowiecki, 1991). Chitin and its deacetylated derivative, chitosan, have unique properties, which make them useful for a variety of applications such as antibacterial agents and preservative coatings for fruits and vegetables (Aranaz et al., 2009;Goy, Britto, & Assis, 2009;Raafat & Sahi, 2009). Proteins extracted from shrimp waste have also been found to be an excellent source of animal feed (Alireza & Bhagya, 2009;Cao, Zhang, Hong, & Zhang, 2009;Meyers & Benjamin, 1987). For removal of proteins from crustacean shells (i.e. deproteini-zation) a base extraction procedure is usually employed. However, waste liquid is produced containing proteins and protein degra-dation products. In order to minimize the loss in nutritional quality, the recovery of the protein hydrolysate must include a concentra-tion step before the protein can be stored. High temperatures and long processing periods must also be avoided (Liaset, Norvedt, Lied, & Espe, 2002;Shahidi, Han, & Synowiecki, 1995). The use of ultra-ﬁltration, may be a promising approach in this regard (Benhabiles, Abdi, et al., 2012; Benhabiles, Salah, et al., 2012; Bourseau, Vandanjon, Jaouen, Chabeaud, & Johansson, 2009;Picot, Ravaller, Fouchereau-péron, Vandanjon, & Bourseau, 2010).

The use of membranes for the treatment of process and waste waters from marine sources was extensively reviewed byAfonso and Bórquez (2002)andMassé et al. (2008). They evaluated the application of membrane technology for treatment and processing

* Corresponding author. Centre de Recherche en Technologie des Semi-con-ducteurs pour l’Energétique, Department of Environmental Engineering, 2 Bd Frantz Fanon, BP140, Alger-7-Mervielles, 16000 Algeria. Tel.:þ213 21 279880x192; fax:þ213 21 279555.

E-mail address:nadjibdrouiche@yahoo.fr(N. Drouiche).

Contents lists available atSciVerse ScienceDirect

Food Hydrocolloids

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / f o o d h y d

(2)

of seafood waste waters and recovery of proteins. Interestingly, Massé et al. (2008)noted that some authors have also studied the recovery of proteins, and were able to incorporate them into the formulation of animal or human feedstuffs. A good review on the use of membrane separation techniques in theﬁelds of peptide or lipid fractionation, aroma up-grading and recovery of carbohy-drates from marine sources is provided byBourseau et al. (2009). Finally, it has been noted that inorganic membranes may also have a high application potential because of their chemical, mechanical, and thermal stability as well as high permeability (Kuca & Szaniawska, 2009; Perez-Galveza, Guadixa, Bergeb, & Guadixa, 2011).

In the present study protein recovery by ultrafiltration was examined during isolation of chitin from shrimp shell Parapenaeus longirostris. Specifically, the deproteinization of shrimp waste was optimized and then the protein was recovered from the shell wash water by concentration using ultrafiltration.

2. Materials and methods 2.1. Raw materials

Shrimp shells, obtained from a seafood restaurant, were from a single species P. longirostris which is a large species of shrimp commonly found in the Mediterranean Sea. The shells were washed under running warm tap water to remove soluble organics, adherent proteins and other impurities. They were then collected and boiled in water for 1 h to remove tissue. The shells were then dried in an oven (Prolabo, model Volca MC18, French) at 160C for 2 h to make them more brittle and to break down the crystalline structure of chitin (Mukherjee, 2002). At the end, the dried shells were crushed into aﬁne powder using a grinder (standard Model KU-2, PredomMesko, SkarzyskoKam., Poland). All chemicals (i.e. sulfuric acid H2SO4, ammonium sulfate (NH4)2SO4, sodium hydroxide NaOH, ammonia NH4OH, sodium carbonate Na2CO3, chloroform CHCl3, methanol CH3OH,. and citric acid C6H8O) used in this study were analytical grade and purchased from Sigma Chemical Co. (St. Louis Mo).

2.2. Analyses

Water content was determined by oven-drying of a 1 g sample at 105C until a constant weight was obtained (AOAC, 1990). Ash content (%) was determined by the AOAC standard method942.05 (AOAC, 1995). 2.0 g of shrimp shell were placed into previously ignited, cooled, and tarred crucibles (triplicate). The samples were heated in a mufﬂe furnace precheated to 600_{C for 6 h. Crucibles} were allowed to cool and then their weight, and respectively the ash content, were subsequently measured using the following calculation:

Weight of residue; g

Sample weight; g 100 ¼ %Ash

Two different methods were used to measure the protein concentration (Biuret, Kjeldahl). Although the Kjeldahl procedure (AOAC, 1990) is the most appropriate, the Biuret method is easy to use and relatively quick and inexpensive. The Kjeldahl method requires the prior mineralization proteins. In contrast it has the disadvantage of being very sensitive (to the milligram), and therefore requires that the test solution is concentrated enough (as is the case of the present work). Also the use the Biuret method allowed us to compare the results with those obtained with the results of work done previously available. In our case, the results

obtained by the two methods are very close (to close inaccuracies). To avoid equivocal, we cite only theﬁrst method in this work.

Total lipids were extracted from the shell wastes according to the method of Bligh and Dyer (1959) using a chloroforme methanolewater (1:2:0.8 v/v/v) system. Protein content was calculated (in three replicates) by extracting (2e3 g) samples with 10% (w/v) NaOH solution for 2 h at 90C, separating the insoluble matter on a coarse sintered-glass funnel and dilution to 100 mL with distilled water. The extract (5e10 mL) was used for protein determination (N 6.25) according to the Kjeldahl procedure (AOAC, 1990).

2.3. Deproteinization of shrimp shells

Dried shells were mixed with 0.5e5 M sodium hydroxide alkaline solution in ratios ranging from 1/10 to 1/40 (w/v). The reaction times were varied from 10 to 400 min, and the tempera-ture from 20 to 100C. At the end of this process the samples were ﬁltered, washed and dried. In order to evaluate the amount of protein removed from the shells (i.e. extent of deproteinization), the protein concentration in the wash solutions was determined by the Biuret method (Itzhaki & Gill, 1964).

2.4. Protein concentration by ultraﬁltration

Protein concentration experiments were carried out using a 130 S ultrafiltration pilot unit (Gamma filtration company, France) equipped with a Membralox-Ceraver module (Fig. 1). This module (P19-40) was a multi-channel ceramic membrane composed of ultrafine porous ZrO2 (0.05

m

m) supported on coarse porous alumina (15

m

m). The totalﬁltration area of this module was 0.2 m2_. The concentrate stream (i.e. retentate) was recirculated back to the initial feed solution. The process was operated at 20C and pH 7. For each experiment 5 L of deproteinized shell waste water (DSWW) was concentrated to 2.1 L on a batch basis over a 50 min

Fig. 1. Scheme of pilot ultrafiltration unit 1: feed tank; 2: feed pump; 3: ultrafiltration module; 4: valves; 5:flow-meter; 6: cooling system; 7: thermometer; 8: pressure regulator.

(3)

period. Solvent (i.e. permeate)ﬂux was measured as a function of trans-membrane pressure.

The inﬂuence of the hydrodynamic parameters, namely the tangential velocity and the trans-membrane pressure was investi-gated during the product concentration experiment. The solute rejection (Ro) of the ultraﬁltration (UF) membrane was assessed.

After each ultraﬁltration experiment, the following membrane cleaning operation was employed; First a pre-washing step was performed where distilled water at the temperature T¼ 40 2_{C was circulated in a closed loop in the apparatus for} 10 min. The temperature used during the cleaning experiment was chosen by taking into account the membrane operating temperature indicated by the supplier. The membrane was then washed with detergent for 20 min (10 min without any trans-membrane pressure (

D

P ¼ 0 Pa) and 10 min with

D

P¼ 0.5 105_{Pa at 50}₂_{C). An ethanolewater mixture (10/} 90; v/v) was used to rinse the membrane without any applied pressure for 15 min. Finally, pre-ﬁltered water was used to rinse the membrane at room temperature at an applied pressure of

D

P¼ 1 105_Pa.

Membrane fouling was assessed by running the UF process with pure solvent (i.e. distilled water) before the deproteinized shell waste water (DSWW) concentration step (i.e. with a clean UF membrane) and then repeating the process with pure solvent after the UF membrane had been used for DSWW concentration. Solvent ﬂux (Jv) was measured as function of trans-membrane pressure (

D

P) as an indication of membrane fouling.

2.5. Determining membrane performance indicators

To measure the efﬁciency of the UF in concentrating the deproteinized shell waste water (DSWW), an apparent rejection coefﬁcient (Ro) was calculated (Goosen, Sablani, Dal-Sin, & Wilf,

2011). This parameter using the initial (Co) and permeate (Cp) solute concentration was determined by the following equation:

R0 ¼ 1CP C0 *100 (1)

To assess the impact of the permeate on the environment; the chemical oxygen demand (COD) was assessed by standard method using uvevisible spectrophotometer at long wave

l

¼ 600 nm. The COD test is used to indirectly measure the amount of organic compounds in water. It is expressed in mg/L which indicates the mass of oxygen consumed per liter of solution. The higher the COD, for example, the more polluted the water. All experiments were run in triplicate.

The hydraulic permeability of the membrane before and after ultrafiltration of DSWW was also determined. The membrane permeability, Lp, was calculated by deducing the slope of the lines obtained by plotting the change of the permeateflux against the average trans-membrane pressure during the filtration of pre-filtered water (Goosen et al., 2011). Assuming that solute adsorp-tion onto an ultrafiltration membrane leads to a permeability variation due to the average pore radius alteration from rp(0) to rp(1), then the Poiseuille’s equation (Fane, Fell, & Waters, 1981) can be used:

LP ¼ NP

p

ðrPÞ 4

8$

m

$e

(2)

where Npis the number of pores per unit membrane area,“e” is the membrane thickness (m),

m

is the solvent kinematic viscosity (Poiseuille) and the reduced pore radius rp(1) may be expressed as:

Jv ¼

Np rp

4

8

m

e

D

P ¼ Lp

D

P (3)

3. Results and discussion

3.1. Characterisation of shrimp shells & optimization of deproteinization step

Chemical analyses (Table 1) showed that shrimp shell had a protein content of 40 w/w% which was similar to other data cited in the literature.Shahidi & Synowiecki (1991);Heu et al. (2003), for example, reported protein contents in shrimp shell ranging from 14 to 32 w/w%. The shrimp shells contained 24% of dry weight chitin. Due to the high protein content an appropriate procedure was needed to remove and recover protein in the course of chitin isolation.

When the sodium hydroxide concentration increased from 0.5 to 2 M in the deproteinization step, the extent of protein removal increased from 12 to 40 wt%, respectively (Fig. 2a). However, an increase in sodium hydroxide concentration from 2 M to 5 M did not increase the degree of protein removal beyond 40 wt%. A similar effect was observed with the deproteinization time. There was an increase in degree of deproteinization from 11 to 39 wt% as the deproteinization time increased from 10 to 120 min (Fig. 2b). Though beyond 2 h there was no further apparent increase in protein removal. On the other hand the sodium hydroxide solution temperature had a signi_{ﬁcant effect on the extent of} deproteini-zation. When the temperature was increased from 20C to 50C the degree of protein removal from the shrimp shells increased from 39 wt% to 55 wt% (Fig. 2d). This rise may have been due to an increase in the solubilty of protein in the alkali solution with higher temperature. Beyond 50C the solution saturation may have been reached. Similar results were obtained by several authors (Khanafari, Marandi, & Sanatei, 2008;Sunita & Ganesh, 2009).

The solid to solvent ratio appeared to be the most signiﬁcant parameter in the removal of protein from shrimp shells using sodium hydroxide solution (Fig. 2c). Using the optimum parame-ters determined in the previous experiments (i.e. 2 M sodium hydroxide concentration (Fig. 2a), 2 h deproteinization time (Fig. 2b) and T¼ 45_{C (}_{Fig. 2}_{d)), the solid to solvent ratio was varied} from 1/10 to 1/40 (w/v) (Fig. 2d). The extent of deproteinization increased from 30 wt% at a solidesolvent ratio of 1/10 (w/v) to 96 wt% with a solidesolvent ratio of 1/20 (w/v); the highest value obtained. There was no further rise in degree of protein removal when the solidesolvent ratio was increased from 1/20 to 1/40 (w/ v). We can assume that the rapid increase in extent of deproteini-zation from 30 wt% at a solidesolvent ratio of 1/10 (w/v) to 96 wt% with a solidesolvent ratio of 1/20 (w/v) may have been the result of an enhanced accessibility of the sodium hydroxide molecules to the protein bound to chitin. Complete removal of protein was not ob-tained even at a high solidesolvent ratio (Fig. 2c). The protein is bound by covalent bonds to chitin forming a stable complex which made dif_{ﬁcult to obtain a 100 wt% removal. A complete removal of} protein would have been even more suitable because it would have

Table 1

Chemical composition of shrimp shell waste (w/w%).

Parametre Value

Moisture 2

Ash 20

Protein 40

Chitin 24

(4)

allowed for a higher solubility value of chitin derivative“chitosan”. In a related study by Toan (2011), it was shown that residual content protein was about 0.66% during extraction of chitin from black tiger shrimp shells.

3.2. Protein concentration and assessment of ultraﬁltration membrane performance

During the concentration of proteins from shell waste water by ultrafiltration (UF) it was observed that the permeate flux (Jv) increased with a rise in both pressure (P) and tangential flow velocity (U) (Fig. 3a). For example, at a velocity of U¼ 3 m/s as the trans-membrane pressure, P, increased from 0 to 3 bars, the permeateflux Jvincreased from 0 to 250 L/h.m2. Beyond P¼ 3 bars the flux became independent of trans-membrane pressure. At a tangentialflow velocity U ¼ 6 m/s a maximum permeate flux, Jv, of 380 L/h.m2occurred at P¼ 3 bars (i.e. 3 105_{Pa). In fact the}_flux started to level of at P¼ 2 bars. Past 3 bars the flux became inde-pendent of pressure. For all four tangentialflow velocities exam-ined (i.e. U ranging from 1.5 to 6 m/s) the_{flux eventually became} independent of pressure. However, as thefluid velocities at the membrane surface increased from 1.5 m/s to 6 m/s, the maximum permeateflux, Jv, increased from 230 L/h/m2to 380 L/h.m2.

The point at which the permeateﬂux becomes independent of the trans-membrane pressure, normally termed the critical ﬂux, Jv,crit, and the critical pressure, Pcrit, is due to concentration polari-zation of solute (i.e. proteins) at the membrane surface (Goosen, Sablani, Al-Hinai, Al-Belushi, & Jackson, 2004,2011;Song, 1998; Yahiaoui et al., 2011) (Fig. 4). A fouling layer of proteins will form

between the polarization layer and the membrane surface when the applied pressure exceeds the critical value (e.g. at U¼ 3 m/s, Jv,crit¼ 250 L/h.m2and Pcrit ¼ 3 bars; seeFig. 3a). At a constant velocity, U, any increases in pressure beyond Pcrit results in a temporary increase in permeateflux. However, at the same time there is a precipitation of solute at the membrane surface which increase the overall membrane resistance and thus lowers the permeateflux. Therefore the flux becomes independent of trans-membrane pressure beyond this critical point. This effect was observed for four studies shown inFig. 3a. It should be noted that Fig. 4is schematic representation of concentration polarization and fouling at the membrane surface, which was retrieved from the literature (Goosen et al., 2004) and describes the phenomenon. For further details readers should look into respective works by previous researchers.

Membrane fouling was assessed by running the UF process with pure solvent (i.e. distilled water) before the deproteinized shell waste water (DSWW) concentration step (i.e. with a clean UF membrane) and then repeating the process with pure solvent after the UF membrane had been used for protein concentration. The results showed that while the clean unused UF membrane had a membrane permeability, Lp, of 302 L/h.m2bar the used membrane after the cleaning procedure had an Lpof only 33 L/h.m2bar (Fig. 3b, top and bottom lines respectively). This was a 90% reduction in permeability and suggested the formation of an irreversible bound (i.e. permanent) fouling or gel layer on the membrane surface. For example, at P¼ 3 bars, the pure solvent ﬂux (Jv), for the clean UF membrane was 900 L/h.m2. At the same pressure the pure solvent for a fouled membrane was only 105 L/h.m2. The reduction inﬂux

Fig. 2. Effect of operative parameters on deproteinization of shrimp shells a) NaOH at: ambient temperature; t¼ 24 h; solid/solvent ¼ 1:15 g/mL. All values presented correspond to % deproteinization measured after 2 h ofﬁltration. b) reaction time at: room temperature; NaOH (2 N); solid/solvent ¼ 1/15 g/mL c) reaction temperature at; NaOH (2 N); solid/ solvent¼ 1/15 g/mL; t ¼ 2 h d) Solidesolvent ratio at : NaOH (2 N); T ¼ 45_{C; t}_{¼ 2 h. 1 g of dried shrimp.}

(5)

was presumably due to an additional resistance caused by the irreversible fouling layer. Furthermore, the schematic representa-tion of concentrarepresenta-tion polarizarepresenta-tion and fouling at the membrane surface inFig. 4shows an irreversibly bound fouling layer on the membrane surface. However, we can assume that irreversible pore blockage in the membrane has also taken place.

The solute rejection coefficient (R0) should be included in defining the optimum conditions for an ultrafiltration process. There was variation in Jvand R0as a function of time (Fig. 3c and d respectively). The permeateflux decreased from 163 L/h/m2_and reached a steady state value of 70 L/h.m2_{, after an equilibrium time} estimated about 25 min (Fig. 3c). This again suggests the formation of a fouling layer at the membrane surface. The thickness of the layer increases with time, thus increasing the resistance to mass transfer as shown by the decrease in permeate flux (Jv). After 25 min a criticalflux was reached at which point the flux became independent of both time and pressure (Chabeaud, Vandonjon, Bourceau, Jaouen, & Guérart, 2009).

The rejection coefﬁcient, R0, increased in value during the concentration experiment reached a maximum of R0¼ 97% after

25 min (Fig. 3d). This high R0was probably due to the fact that Jv,crit had been reached (Fig. 3c at 25 min). There was also a reduction in retentate (i.e. DSWW) volume of a factor of 2.4 by the end of the UF protein concentration experiment. In a related study,Watanabe, Ohtani, Horikita, Ohya, and Kimura (1986)examined the effect of pore size of ceramic support on the self-rejection characteristics of what they termed a dynamic membrane formed with water-soluble proteins in waste water. Microscopic observations showed that a large part of this membrane contained a fouling layer on the uneven surface of the ceramic support. This concept is similar to the concentration polarization and gel layer formation on the membrane surface (Goosen et al., 2011).

Using the change in the value of membrane permeability Lp, before and after protein concentration, it was possible to estimate the change in the apparent membrane pore radius, rp, as a result of membrane fouling. Assuming a membrane pore radius rp(0)¼ 500 A before ultrafiltration, using Equation(3)the calcu-lated value after ultrafiltration rp(1) ¼ 300 A. This indicated a significant decrease (i.e. 40%) in pore size which would increase solute rejection as well as decrease the permeate flux. The

Fig. 3. Effect of operating parameters on ultraﬁltration process a) effect of pressure and velocity on permeate ﬂux b) effect of membrane fouling on membrane permeability Lpto

pure solvent: (:) solvent flux line for clean membrane before ultrafiltration (-) solventflux line for fouled membrane after ultrafiltration of proteins and cleaning procedure. c) flux of permeate plotted against time, T ¼ 20_{C, U}_{¼ 6 m/s,}_D_P_{¼ 3 bars, pH ¼ 7. d) rejection coefficient plotted against time; T ¼ 20}_{C, U}_{¼ 6 m/s,}_D_P_{¼ 3 bars, pH ¼ 7 e) variation of}

(6)

difference between rp(1) and rp(0) can be explained by the fact of adsorption of various solutes such as proteins on the surface of the pores of the membrane.Meireles et al. (1992), demonstrated that pore radius of membrane with molecular weight cut-off (MWCO) of 40,000 Da, decreased from 70 A to 36 A when BSA (Bovin Serum Albumin) was studied.

Assessment of the chemical oxygen demand (COD) of the permeate showed a gradual decrease of this parameter which decreased from 1550 mg/L at t¼ 0 min to 200 mg/L at t ¼ 50 min (Fig. 3e). This indicated a decrease in the organic compounds molecules in the permeate. We can speculate that this decrease was due to the development of a reversible concentration polarization layer and an irreversible fouling layer on the retentate side of the UF membrane which increased the resistance to solute movement across the membrane. A low permeate COD is important for minimizing the impact of the shell waste water (i.e. permeate) on the environment. COD studies on permeate waste streams have also been done byAloulou, Walba, Ben Amar, and Jaouen (2006) who found a 65% COD reduction. While,Shiau and Chai (1999) found that the value of chemical oxygen demand (COD) in the permeate was reduced by 47%.

4. Conclusions

The solution temperature and solidesolvent ratio were found to be the most critical in the removal of protein from shrimp shell waste. Up to 96 wt% of the proteins could be removed (i.e. depro-teinization) from the shrimp shells by incubating them in NaOH (2 N) over 2 h, at T¼ 45_{C, and at a solid to solvent ratio of 1:2 (w/} v). In the ultrafiltration process to recover proteins from deprotei-nized shell waste water a solute rejection coefficient (R0) of 97% was obtained. The protein concentration process which was carried out beyond the criticalflux, Jv,crit, of 380 L/h.m2, at a trans-membrane pressure of 3 bars, and a tangential velocity of 5 m/s was found to reduce the hydrolysate volume by a factor of 2.4. Due to a reduction in organic matter in the effluent, the chemical oxygen demand

(COD) of the permeate was reduced by 87%. However, the results also show that irreversible membrane fouling is occurring which may limit the usefulness of UF unless effective membrane cleaning methods can be found. Further work is needed in this area. The present study highlights the opportunities to recover added value products.

References

Afonso, M. D., & Bórquez, R. (2002). Review of the treatment of seafood processing wastewaters and recovery of proteins therein by membrane separation processes d prospects of the ultraﬁltration of wastewaters from the ﬁsh meal industry. Desalination, 142, 29e45.

Alireza, S. M., & Bhagya, S. (2009). Effect of recovery method on different property of mustard protein. World Journal of Dairy & Food Science, 4, 100e106. Aloulou, I., Walba, K., Ben Amar, R., & Jaouen, P. (2006). Preliminary study of the

treatment of effluents containing cuttlefish ink by centrifugation and membrane processes. Journal of Water Science (Revue des Sciences de l’eau), 19, 383e392. AOAC. (1990). In E. Horritz (Ed.), Official methods of analysis (15th ed.).

AOAC. (1995). Association of ofﬁcial analytical chemists (15th ed.).In Ofﬁcial methods of Analysis, Vol. 1 Arlington, VA: AOAC.

Aranaz, I., Mengibar, M., Harris, R., Panos, I., Mirrales, B., Acosta, N., et al. (2009). Functional characterization of chitin and chitosan. Current Chemical Biology, 3, 203e230.

Benhabiles, M. S., Abdi, N., Drouiche, N., Lounici, H., Pauss, A., Goosen, M. F. A., et al. (2012). Fish protein hydrolysate production from sardine solid waste by crude pepsin enzymatic hydrolysis un a bioreactor coupled to an ultraﬁltration unit. Materials Science and Engineering C, 32, 922e928.

Benhabiles, M. S., Salah, R., Lounici, H., Drouiche, N., Goosen, M. F. A., & Mameri, N. (2012). Antibacterial activity of chitin, chitosan and its oligomers prepared from shrimp shell waste. Food Hydrocolloids, 29, 48e56.

Bligh, E. G., & Dyer, W. J. A. (1959). A rapid method of total lipid extraction and puriﬁcation. Canadian Journal of Biochemical Physiology, 37, 911e917. Bourseau, P., Vandanjon, L., Jaouen, P., Chabeaud, A., & Johansson, I. (2009).

Frac-tionation of fish protein hydrolysate by ultrafiltration and nanofiltration: impact on peptidics population. Desalination, 244, 303e320.

Cao, W., Zhang, C., Hong, P., & Zhang, J. (2009). Autolysis of shrimp head by gradual temperature and nutritional quality of the resulting hydrolysate. LwteFood Science and Technology, 42, 244e249.

Chabeaud, L., Vandonjon, V., Bourceau, P., Jaouen, P., & Guérart, F. (2009). Frac-tionation by ultraﬁltration of a saithe protein hydrolysate (Pollachius virens): effect of material and molecular weight cut-off on the membrane perfor-mances. Journal of Food Engineering, 91, 408e414.

Cira, L. A., Huerta, S., Hall, G. M., & Shirai, K. (2002). Pilot scale lactic acid fermen-tation of shrimp waste for chitin recovery. Process Biochemistry, 37, 1359e1366. Fane, A. G., Fell, C. J. D., & Waters, A. G. (1981). The relationship between membrane surface pore characteristics and ﬂux ultraﬁltration membrane. Journal of Membrane Science, 9, 245e262.

FAO STAT. 2009. <HTTP://FAOSTAT.FAO.ORG/SITE/339/DEFAULT.ASPX> Accessed 02.12.10.

Goosen, M. F. A., Sablani, S., Al-Hinai, H., Al-Belushi, R., & Jackson, D. (2004). Fouling of reverse osmosis and ultraﬁltration membranes: a critical review. Separation Science and Technology, 39, 1e37.

Goosen, M. F. A., Sablani, S., Dal-Sin, M., & Wilf, M. (2011). Effect of cyclic changes in temperature and pressure on permeation properties of composite polyamide seawater reverse osmosis membranes. Separation Science and Technology, 46, 14e26.

Goy, R. C., Britto, D., & Assis, O. D. G. (2009). A review of the antimicrobial activity of chitosan. Polimeros Gienca e Tecnologia, 9, 241e247.

Heu, M. S., Kim, J. S., & Shahidi, F. (2003). Component and nutritional quality of shrimp processing by-products. Food Chemistry, 82, 235e242.

Itzhaki, R. F., & Gill, D. M. (1964). A micro-biuret method for estimating proteins. Analytical Biochemistry, 9, 401e410.

Khanafari, A., Marandi, R., & Sanatei, S. (2008). Recovery of chitin and chitosan from shrimp waste by chemical and microbial methods. Iranian Journal Environment Health Sciences Engineering, 5, 19e24.

Kuca, M., & Szaniawska, D. (2009). Application of microfiltration and ceramic membranes for treatment of salted aqueous effluents from fish processing. Desalination, 241, 227e235.

Liaset, B., Norvedt, R., Lied, E., & Espe, M. (2002). Studies on the nitrogen recovery in enzymatic hydrolysis of Atlantic salmon (Salmo salar, L.) frames by Protamex TM proteas. Process Biochemistry, 37, 1263e1269.

Massé, A., Vandanjon, L., Jaouen, P., Dumay, J., Kéchaou, E., & Bourseau, P. (2008). Upgrading and pollution reduction of ﬁsh industry process-waters by membrane technology. In J.-P. Bergé (Ed.), Added value toﬁshery wastes (pp. 81e 100). Kerala: Transworld Research Network.

Meireles, M., Aimer, P., & Sanchez, V. (1992). Les techniques à membrane: Micro et Ultraﬁltration. Le technoscope de biofutur, 111, 1e18.

Meyers, S. P., & Benjamin, G. (1987). Feeding value of crustacean wastes can be improved through proper silage treatment. Feedstuffs, 30(March), 12. Mukherjee, D. P. (2002). Method for production chitin or chitosan. United State

Patent, 6.310.188. Fig. 4. A schematic representation of concentration polarization and fouling at the

(7)

Perez-Galveza, R.,M., Guadixa, E., Bergeb, J. P., & Guadixa, A. (2011). Operation and cleaning of ceramic membranes for theﬁltration of ﬁsh press liquor. Journal Of Membrane Science, 384, 142e148.

Picot, L., Ravaller, R., Fouchereau-péron, M., Vandanjon, L., & Bourseau, P. (2010). Impact of ultrafiltration and nanofiltration of an industrial fish protein hydro-lysate on its bioactive properties. Journal of Food & Agriculture, 90, 1819e1826. Raafat, D., & Sahi, H. G. (2009). Chitosan and its antimicrobial potentialea critical

literature survey. Microbial Biotechnology, 2, 186e201.

Shahidi, F., Han, X. Q., & Synowiecki, J. (1995). Production and characteristics of protein hydrolysates from capelin (Mallotus villosus). Food Chemistry, 53, 285e293.

Shahidi, F., & Synowiecki, J. (1991). Isolation and characterization of nutrients and value-added products from snow crab (Chionoecetes opilio) and shrimp (Pan-dalus borealis) processing discards. Journal of Agricultural and Food Chemistry, 39, 1527e1532.

Shiau, C. Y., & Chai, T. J. (1999). Protein recovered from oyster wash water by ultraﬁltration and their utilization as oyster sauce through fermentation. Jour-nal of Marine Science and Technology, 7, 110e116.

Song, L. (1998). A new model for the calculation of the limitingﬂux in ultraﬁltra-tion. Journal of Membrane Science, 144, 173e185.

Sunita, D., & Ganesh, A. (2009). Extraction of chitin from trash crabs (Podophthalmus vigil) by an eccentric method. Journal of Biological Sciences, 272e275. Toan, N. V. (2011). Improved chitin and chitosan production from black tiger shrimp

shells using salicylic acid pre-treatment. The Open Biomaterials Journal, 3, 1e3. Washington, DC. Association of Ofﬁcial Analytical Chemists, method#930.15 P69,

method #942.05 P70, method #962.09 p80.

Watanabe, A., Ohtani, T., Horikita, H., Ohya, H., & Kimura, S. (1986). Recovery of soluble proteins from ﬁsh ﬁlly processing with self-rejection dynamic membrane. Food Engineering & Process Application, 2, 225e336.

Yahiaoui, O., Lounici, H., Abdi, N., Drouiche, N., ghaffour, N., Pauss, A., et al. (2011). Treatment of olive mill wastewater by the combination of ultraﬁltration and bipolar electrochemical reactor processes. Chemical Engineering and Processing B, 50, 37e41.

Zakaria, Z., Hall, G. M., & Shama, G. (1998). Lactic acid fermentation of scampi waste in a rotating horizontal bioreactor for chitin recovery. Process Biochemistry, 33, 1e6.