Effect of cold chain temperature abuses on the quality of frozen watercress (Nasturtium officinale R. Br.)

* Corresponding author. Tel.: +351 22 5580058; fax E-mail addresses: [email protected] (R.M.S. Cruz), m [email protected] (C.L.M. Silva). t of temperature abuses on the colour and vitamin C conturtium officinale R. Br.). c (AA) and dehydroascorbic (DHAA) acids, and colour, evaluated along a plan of temperature abuses, based on e period. A comparison between the hue angle and AA kinetic parameters determined under isothermal condithe parameters L, a, TCD (Total Colour Difference) and uses, some fluctuation was observed, no vitamin C degro-order and a first-order prediction models fitted well ehaviour with temperature abuses, respectively. In gencontent were not impaired by the imposed temperature stand the sensory and nutritional quality changes of frotorage and distribution in the cold chain under the used Watercress (Nasturtium officinale R. Br.) is a leafy vegetable of the family Cruciferae that grows in and around water. It is a good source of essential vitamins and minerals and beneficial phytochemicals, such as lutein and zeaxanthin (U.S. Department of Agriculture, 2003). Normally consumed fresh, it has a short shelf life (approximately seven days) that can be extended throughout freezing, allowing a longer period for distribution and storage. Blanching is one of the pre-treatments used before freezing, to inactivate the enzymes and prevent biochemical reactions in the frozen product that might contribute for the development of offflavours and discoloration (Mountney and Gould, 1988). However, blanching causes undesirable changes in the food properties that may result in loss of colour, flavour, texture and nutrients (Pala, 1983; Pizzocaro et al., 1995). The freezing process is an excellent method to preserve taste, texture and nutritional value of foods, and better than most other preservation methods. Nevertheless, such qualities depend upon the careful choice of food materials, suitable pre-treatments, : +351 22 5090351. [email protected] (M.C. Vieira), choice of freezing process, and appropriate packaging and frozen storage options (Flair flow Europe, 2000; Blond and Le Meste, 2004). Exposure to higher temperatures and/or fluctuations of storage temperature produces cumulative adverse effects on the quality of stored foods, which is the primary cause of damage to food marketed through retail channels (Blond and Le Meste, 2004). The loss of quality is caused by physical and chemical changes taking place in the product. During the storage period, physical changes that affect frozen vegetables and fruits quality are a result of recrystallization and sublimation phenomena, related to the stability of the ice crystals inside and on the surface of the products. Recrystallization and surface drying are accelerated by temperature fluctuations during freezing, although the importance of these physical changes decreases at lower storage temperatures (Canet, 1989; Alvarez and Canet, 1998; Sousa et al., 2005). New frozen product formulations and new freezing technologies benefits should be assessed taking into account the frequent temperature abuses occurring in commercial and domestic storage (Estrada-Flores, 2002). In the distribution of frozen food, the cold chain starts at the raw material supplier and continues until the consumers’ freezer, playing a very important role in the maintenance of food safety and quality. Therefore, in each major stage of the cold chain, Nomenclature a colour co-ordinate, represents red to green b colour co-ordinate, represents blue to yellow C parameter concentration or value Ea activation energy (kJ mol ) k reaction rate (day ) L colour co-ordinate, represents black to white R universal gas constant (8.314 J mol 1 K ) t time (days) T absolute temperature (K) TCD total colour difference parameter Subscripts T relative to a given temperature 0 initial value at time equal to zero ref at the reference temperature Factory (product at -21 oC) 1 to 8 selling point* (unload the truck: 15 min/4 oC) (transportation: 30 min/-15 oC) Transportation (15 min/4 oC; 45 min/-10 oC; 1h/-18 oC) 1 distribution center (unload the truck: 15 min/4 oC) (transportation: 15 min/-10 oC; 1h/-18 oC) 2 distribution center (unload the truck: 15 min/4 oC) (transportation: 15 min/-10 oC; 1h/-18 oC) 3 distribution center (unload the truck: 20 min/-10 oC) (distribution chamber: 3 days/-21 oC) S3 Load the retail car (1 h at 25 oC) S4 Transportation (15 min/4 oC; 15 min/-10 oC) Consumer (product purchase: 30 min at 25 oC) S7 (product storage: 3 months/at least -18 oC; plus 5 door openings of 3 min and 1 door opening of 30 min per week (shopping simulation)) S8-S20 Load the Truck (1 h at 25 oC) S2 Last customer (10 min at 25 oC) S5 (10 days/-15 oC, night shutdown (12 h)) S6 S1 Fig. 1. Plan of temperature abuses. S1 to S20steps of analysis; * These two steps were repeated eight . temperature requirements must be taken into account. Usually, the suggested storage temperature is below 18 C, in which the microbial growth is completely stopped, and both enzymatic and non enzymatic changes continue but at much slower rates during frozen food shelf life, although temperature of 15 C is permitted for short periods during transportation or local distribution. Moreover, food retail display cabinets should be at 18 C with good storage practice, but not warmer than 12 C. At the consumer’s domestic freezers, storage temperature is also important, being the freezer ‘star rating’ related to its freezing capacity. A one-star freezer is capable of temperatures below 6 C, a two-star freezer of temperatures below 12 C, and a three-star freezer is capable of temperatures below 18 C (Flair flow Europe, 2000). The objective of this work was to study the effect of temperature abuses on colour and vitamin C content of frozen watercress and with the findings help to understand the sensory and nutritional quality changes that might occur during frozen storage and distribution in the food chain. Raw watercress (N. officinale R. Br.) was kindly supplied by Vitacress, a company that grows watercress in Almancil, Algarve-Portugal. The leaves were directly harvested from the producer and, on the same day, processed (selected, washed, blanched and frozen). The leaves of watercress were blanched in a boiling pan (Armfield 45 L, Hampshire, England) in a proportion of 30 g per litre of water, set at 95 C during 20 s, and cooled in an iced water bath (Cruz et al., 2006). After the blanching process, the leaves (30 g) were pressed in a mould into a slab shape (4.6 3.3 1.8 cm). The slabs were frozen in an air blast freezer (Armfield FT 36, Hampshire, England) at an average temperature of 25 C and an air velocity of 8 m s , until the temperature of 18 C was reached at the centre of the sample. Samples triplicates for each step of analysis were packed in low density polyethylene bags and stored at 21 C (Haier HF-248, Germany). All temperatures were monitored with a digital thermometer (Ellab ctd 87, Roedovre, Denmark) and a thermocouple (1.2 mm needle dia; constantan-type T). A plan of temperature abuses was established based on a real situation for a four month period (Oliveira et al., 2009) and applied to the frozen watercress packages (Fig. 1). The samples were stored in freezers (Haier HF-248, Germany) and analysed in each step of the plan (S1–S6 corresponds to production and distribution; S7– S20 corresponds to consumer product purchase and storage). The watercress temperature was measured with a type T thermocouple placed in the slab thermal centre and recorded with a data acquisition system (Delta-T Devices DL2 e, Cambridge, England). Colour was evaluated in terms of Lab values, with a tristimulus colorimeter (Dr Lange Spectro-colour, Berlin, Germany) in the Hunter system (HunterLab, 2000). The colorimeter (d/8 geometry, illuminant D65, 10 observer) was calibrated against a standard ceramic white tile (X = 84.60, Y = 89.46 and Z = 93.85) and a standard ceramic black tile (X = 4.12, Y = 4.38 and Z = 4.71). The colour changes were interpreted by calculating the TCD (Total Colour Differences) (Eq. (1)) where DL, Da and Db are determined by Eq. (2). The hue angle (tone or tint) was determined by Eq. (3) and chroma (saturation) by Eq. (4). Measurements were taken in triplicates. TCD 1⁄4 ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðDLÞ þ ðDaÞ þ ðDbÞ q ð1Þ DL 1⁄4 L L0 Da 1⁄4 a a0 Db 1⁄4 b b0 ð2Þ where L0, a0 and b0 are the readings of the frozen samples before temperature abuses, and L, a and b the individual readings at each step of the temperature abuses plan. Hue angle 1⁄4 ðtan 1 b=aÞ ð3Þ Chroma 1⁄4 ða þ bÞ ð4Þ Table 1 Watercress colour and vitamin C content before the plan of temperature abuses. Fresh Blanched Frozen L 35.36 ± 0.37 31.02 ± 0.91 29.05 ± 2.53 a 10.35 ± 0.03 10.47 ± 0.17 4.58 ± 0.74 b 12.30 ± 0.16 9.05 ± 0.21 3.97 ± 1.31 Chroma 16.08 ± 0.14 13.84 ± 0.25 6.08 ± 1.39 Hue angle ( ) 130.07 ± 0.29 139.16 ± 0.43 139.94 ± 5.89 AA (mg 100 g fw ) 35.66 ± 1.42 4.51 ± 0.48 4.63 ± 0.77 DHAA (mg 100 g fw ) 29.17 ± 8.31 6.32 ± 1.06 8.61 ± 5.17 Total vitamin C (mg 100 g fw ) 64.83 ± 7.85 10.83 ± 0.58 13.25 ± 4.43 Ascorbic acid (Riedel-de Haën) content was determined based on a method previously reported (Zapata and Dufour, 1992), by reverse phase ion interaction high performance liquid chromatography HPLC UV detection using isoascorbic acid (IAA) (Fluka) as internal standard. Dehydroascorbic acid (Sigma) was as well detected as fluorophore 3-(1,2-dihydroxyethyl)furo[3,4-b]quinoxaline-1-one (DFQ), after pre-column derivatization with 1,2-phenylenediamine dihydrochloride (OPDA) (Sigma). For the mobile phase, 13.61 g of potassium dihydrogen phosphate (Merk) and 3.64 g of cetrimide (Fluka) were added to 2 l of methanol–ultrapure water (5:95, v/v). The eluent was filtered in a 0.45 lm membrane (Macherey-Nagel, Porafil) and degassed in an ultrasonic bath for 15 min. The HPLC system consisted of a controller (LKB-2152, Bromma, Sweden), a solvent pump (LKB-2150, Bromma, Sweden), an injection valve with a 20 ll sample loop, a guard pre-column (Macherey-Nagel, Chromcart Nucleosil 100-10 C18) followed by a reversed phase column (Macherey-Nagel, Chromcart 100-10 Nucleosil, 250 4.6 mm) and a UV detector (LKB-2153, Bromma, Sweden). Each sample was homogenized with an Ultra-turrax (IKA T25 Janke & Kunkel, Staufen, Germany) in 20 ml of methanol–ultrapure water (5:95, v/v) for 5 min at 8000 rpm. Then, 5 ml were transferred to a 20 ml volumetric flask and 1 ml of IA standard solution (0.03 g/50 ml) was added. The pH was adjusted with HCl (Merk) to obtain final values between 2.20 and 2.45. The volume was completed to 20 ml with methanol–ultrapure water (5:95, v/v). The content was centrifuged (Sigma 3 K20, Osterode, Germany) for 5 min at 8.720 g and 4 C. Afterwards, 3 ml were transferred to another tube with 1 ml of OPDA (Sigma) (0.03 g/50 ml) daily prepared and maintained in dark. The mixture was then vortexed and placed in the dark at room temperature for 40 min. Then the mixture was filtered in a 0.45 lm membrane (Millipore), the first millilitre was discarded and 20 ll were injected in the HPLC. The wavelength detector was set to 348 nm and after elution of DHAA, the wavelength was shifted to 262 nm for AA and IA detection. The experiments were run in triplicate. Tissue microstructure was only examined to evaluate the effect of the blanching process. Thus, after blanching, the leaves were stored in a vertical freezer at 80 C (Snijders Scientific, Tilburg, The Netherlands). Afterwards, based on the method reported by Fonseca et al. (2005), the samples were transversally cut with a surgery thin blade and observed in a scanning electron microscope (SEM) (JEOL JSM-5600 LV, Tokyo, Japan) at low vacuum with an acceleration voltage of 15 kV and a cryo-chamber set at 25 C. The experiments were run with six replicates. The ascorbic acid and hue angle experimental data were compared with prediction models (using the kinetic parameters estimated under isothermal conditions (Table 2) (Gonçalves et al., 2009), in which the integration of the temperature variability with time was required. The reaction rate kT, for which the temperature dependence followed the Arrhenius behaviour, was calculated using Eq. (5): kT 1⁄4 kref exp Ea R 1 T 1 Tref ð5Þ Where kref is the reaction rate at the reference temperature, Ea, the activation energy (previously determined under isothermal conditions), R, the universal gas constant, T, the absolute temperature, and Tref, the reference temperature (258.15 K). In the isothermal study, a first-order reaction (Eq. (6)) fitted well the ascorbic acid experimental data. While in the colour evaluation, the parameters were estimated from a zero-order reaction (Eq. (7)) (Gonçalves et al., 2009). CðtÞ C0 1⁄4 expð kTtÞ ð6Þ CðtÞ 1⁄4 C0 kTt ð7Þ Under non-isothermal conditions, since T = f (t), the ascorbic acid and hue angle values can be predicted, for a given time, by Eqs. (8) and (9), respectively. CðtÞ 1⁄4 C0 exp kref Z t 0 exp Ea R 1 TðtÞ 1 Tref dt ð8Þ CðtÞ 1⁄4 C0 kref Z t 0 exp Ea R 1 TðtÞ 1 Tref dt ð9Þ Calculations were performed using the statistical software SPSS 17.0. One-way analysis of variance (ANOVA) was carried out to determine if there were significant differences between each step of the plan of temperature abuses. The Least Significant Difference (LSD) test was run to determine the significant differences between Table 2 Published watercress AA and hue value kinetic parameters under isothermal conditions (Gonçalves et al., 2009). Kinetic parameters AA Hue angle ( ) Ea (kJ mol ) 24.73 174.71 kref 15 C (d ) 0.00432 0.00287 C0 32.62 mg 100 g 1 132.35 Kinetic model First-order Zero-order Fig. 3. SEM micrographs of watercress cross section tissue at different magnification levels (300 and 800 ): (a and a’) Fresh watercress; (b and b’) Blanched watercress. P = parenchyma cells, Vt = vascular tissue, white arrow indicate the chloroplasts. the control (without abuses) and each treatment. Evaluations were based on a significance level of 5%. The experimental values for watercress colour and vitamin C content before the plan of temperature abuses are shown in Table 1. The applied blanching process, aiming 90% reduction of the enzyme peroxidase, caused some colour changes and degradation on the vitamin C content. From Fig. 2, it can be observed that, after blanching, the L and b parameters decreased about 12% and 26%, respectively, showing that the samples became darker and less yellow. On the other hand, the a parameter did not change after the blanching treatment. Colour changes after blanching could be related with the replacement of the gases inside the intercellular spaces by the blanching medium, altering light refraction from the cell surface (Bowers, 1992). In terms of microstructure quality, the fresh sample (Fig. 3a and a’) showed the parenchyma cells and vascular tissue turgid and well defined. The blanched watercress microstructure (Fig. 3b and b’) presented the parenchyma cells filled with spherical structures (probably chloroplasts released from inside the cells). The blanched samples also showed a more damaged microstructure, being the tissue firmness and cell wall strength rapidly lost due to the heat extent. Under thermal blanching, the cell wall membrane may be damaged and allow water to enter the cell. Internal organelles may be distorted and begin to leak their contents. A reduction of cell turgidity is caused by the loss of cell membrane function. Moreover, -20 -10 0 10 20 30 40 Fr es h B la nc he d Fr oz en S 1 W ith ou t a bu se s