Introduction

A light-emitting diode (LED) is a semiconductor device which can emit very narrow visible light wavelengths (Kim et al., 2016; Mori et al., 2007). LEDs have merits, including lower energy consumption, long life, reduced heat output, reduced heat, and easy manipulation in small sizes and various shapes over the traditional visible light sources (Ghate et al., 2013; Mori et al., 2007). Inactivation of foodborne pathogens using a LED technology has recently received the interest of researchers due to its advantages and antimicrobial effect. Previous studies demonstrated that 405 and 460 nm LEDs inactivated Escherichia coli O157:H7, Listeria monocytogenes, Salmonella Typhimurium, and Staphylococcus aureus in phosphate buffered saline (PBS) or tryptic soy broth (TSB) at refrigeration temperatures (Ghate et al., 2013; Kim et al., 2015, 2016). In addition, blue LEDs inactivated a variety of foodborne pathogens on food matrices such as cooked chicken, salmon, mango, papaya, pineapple, and orange juice without the aid of photosensitizers (Ghate et al., 2016; Ghate et al., 2017; Kim et al., 2017a; b; c; Li et al., 2018).

The antimicrobial mechanism of blue LEDs could be explained by photodynamic inactivation (PDI) requiring photosensitizers and visible lights in the presence of oxygen (Luksienė and Zukauskas, 2009). Once bacterial cells are exposed to light energy, the photosensitizer, such as intracellular porphyrin molecules, is excited and then produced reactive oxygen species (ROS), such as hydrogen peroxide, hydroxyl radicals, superoxide, and singlet oxygen. The ROS can attack cellular components, including DNA, lipids, and proteins, causing oxidative damage and eventually bacterial death (Kim et al., 2017a, c; Luksienė and Zukauskas, 2009). Recent studies have focused on LED illumination with exogenous photosensitizers such as chlorophyllin and hypericin to enhance the antimicrobial effect on foods (Aponiene et al., 2015; Luksienė and Paskeviciute, 2011). Our previous study has also demonstrated that 460 nm LED illumination with the aid of riboflavin resulted in 0.7–1.2 log CFU/cm2 reductions of L. monocytogenes cells on smoked salmon at chilling temperatures (Josewin et al., 2018). However, no study has been conducted with 405 nm LED illumination with exogenous riboflavin on L. monocytogenes on smoked salmon.

Smoked salmon as one of ready-to-eat (RTE) foods is commonly purchased vacuum-packed and stored at refrigeration temperature for weeks to extend the shelf-life (Rørvik, 2000). L. monocytogenes have recently been associated with several foodborne illness outbreaks due to cold-stored RTE foods, such as meat, soft cheese, and especially cold-smoked fishery products, since they have the ability to grow 4 °C (Heir et al., 2019; Kim et al., 2015; Ye et al., 2008). In salmon processing industries, L. monocytogenes existing in raw salmon can be contaminated to food contact surface, such as conveyer belts, slicing machines, and knives (Pang et al., 2019). RTE fish products can be also cross-contaminated with L. monocytogenes during handling and further processing at food establishments, although smoked salmon may involve other processes, such as drying, salting, smoking, and packaging (Ricci et al., 2018; Rørvik, 2000). For this reason, human listeriosis outbreaks are frequently associated with smoked salmon due to eating it without the additional cooking step (Ye et al., 2008).

To control the growth of L. monocytogenes in RTE fisher products, several physical intervention technologies have been applied, including irradiation, pulsed UV-light treatment, high pressure processing, and microwave heating (Tocmo et al., 2014). However, these technologies have shortcomings, such as safety concerns, high cost, or changes in the food quality (Josewin et al., 2018). Thus, the implementation of suitable preservation technologies in fresh fishery products should be needed to minimize the risk of listeriosis. In this study, the efficacy of 405 nm LED illumination with and without riboflavin on L. monocytogenes on the surface of smoked salmon at different storage temperatures was evaluated and the quality changes of smoked salmon after long-term LED illumination were investigated.

Materials and methods

Bacterial strains and culture conditions

Four L. monocytogenes serovars were used in this study: 1/2a (SSA81) and 3a (LM8) isolated from pre-paced smoked salmon from supermarkets and 1/2b (LM4) and 4b (LM2) isolated from smoked salmon in salad bar obtained from the National Environment Agency (Chau et al., 2017). All L. monocytogenes serovars were preserved on beads into cryoinstant vials (DeltaLab, Barcelona, Spain) at − 70 °C. Frozen cultures were activated in 10 mL of sterile tryptone soy broth (TSB; Oxoid, Basingstoke, UK) for 18–24 h at 37 °C. After two consecutive transfers for 18–24 h at 37 °C, a cocktail culture of L. monocytogenes was prepared by combining equal portions (0.25 mL) of each strain to be a total of 1 mL and centrifuged at 6000×g for 10 min at 4 °C, followed by washing twice with sterilized phosphate buffered saline (PBS; Vivantis Technologies Sdn Bhd, Malaysia). The resultant pellets were resuspended in 1 mL of PBS and serially diluted to ca. 106 CFU/mL for inoculation.

Light-emitting diode (LED) source and illumination system

LEDs (8 × 8 mm) at 405 nm wavelength were purchased from Shenzhen Getian Opto-Electronics Co., Ltd. (Shenzhen, Guangdong, China) and attached to a fan and a heat sink to reduce heat transfer to samples. The irradiance of a 405 nm LED unit was 16 ± 1 mW/cm2 at the surface of samples, which was measured with a Compact Power and Energy Meter Console (PM100D; Thorlabs GmbH, Dachau, Germany). The dose applied to each sample was calculated with the following equation (Ghate et al., 2013; Kim et al., 2015):

$$E = Pt$$

where E = dose (energy density) in J/cm2, P = Irradiance (power density) in W/cm2, and t = time in s.

To adjust the height of each LED system, the system was set up in an acrylonitrile butadiene styrene (ABS) housing. To illuminate the entire surface of bacterial suspension and smoked salmon, 10 mL bacterial suspension (ca. 1 cm depth) and smoked salmon (ca. 0.2 cm depth) were deposited in a sterile Petri dish (60 mm and 90 mm diameter, respectively) and then placed directly below the LED at distance of 5.5 cm and 4.5 cm, respectively (Fig. 1A and B). The surface temperatures of smoked salmon were monitored using a Fluke 5.4 thermocouple thermometer (Everett, WA, USA) during exposure to LED illumination at doses up to 288 J/cm2 (for 300 min).

Fig. 1
figure 1

Schematic diagrams of 405 nm LED illumination systems in phosphate-buffered saline (PBS) solution (A) and on the surface of smoked salmon samples (B)

Full size image

Riboflavin-mediated LED illumination in PBS

To evaluate the antibacterial efficacy of 405 nm LED illumination in the presence of riboflavin, a 10 mL bacterial suspension (ca. 106 CFU/mL in PBS) containing 0, 50 and 100 μM riboflavin (RBF; Sigma-Aldrich, St. Louis, MO, USA) was placed in the LED system as described above and illuminated up to 57.6 J/cm2 (for 1 h) at 25 °C in a temperature-controlled incubator (MIR-154, Panasonic Healthcare Co., Ltd., Osaka, Japan). Non-illuminated control samples were placed in the incubator in the dark. A 0.1 mL aliquot was withdrawn at selected time intervals and serially diluted with 0.1% (w/v) sterile peptone water (PW; Oxoid), if necessary. The diluents were pour-plated in tryptone soy agar (TSA; Oxoid). After incubation at 37 °C for 48–72 h, the number of surviving cells was enumerated with a colony counter (Rocker Scientific Co. Ltd., Taipei, Taiwan) and reported as log CFU/mL.

Preparation of smoked salmon and inoculation

Norwegian smoked salmon (Norlax, Outrup, Denmark) in the absence of preservation and coloring agent was purchased from a local supermarket in Singapore and used for the experiments within two days. Smoked salmon was cut into ca. 4.0 ± 0.5 g slices (4 × 4 cm) in a biosafety cabinet. A 20 µL aliquot of the diluent (ca. 106 CFU/mL) as previously described was spot inoculated at 5 sites on the salmon surface to reach a final concentration of ca. 5 × 103 CFU/cm2, followed by dry for 20 min in a biosafety cabinet.

To prepare non-illuminated control and illuminated salmon samples with riboflavin, riboflavin was added on the surface of inoculated salmon samples and spread evenly using a sterile spreader, followed by dry for 10 min in a biosafety cabinet. All samples were individually wrapped with a polyethylene film to keep moisture of salmon samples during storage. No significant (p < 0.05) difference in the intensity of LED was observed when the intensity sensor was applied with/without the film in the preliminary study (data not shown).

Riboflavin-mediated LED illumination on smoked salmon

Two inoculated samples with/without riboflavin were placed in each LED illumination system at 4, 12, and 25 °C in an incubator at doses of 1.27–2.76 kJ/cm2 (for 22–48 h). Non-illuminated control samples with/without riboflavin were also placed in an incubator without LED illumination (dark condition). Each of two samples were withdrawn at a selected time interval, immediately placed in a sterile stomacher bag containing 36 mL of 0.1% PW, and homogenized for 2 min with a Paddle blender (Silver Masticator, IUL Instruments GmbH, Königswinter, Germany). The homogenized sample was serially diluted if necessary. A 100-μL diluent was spread on a thin agar layer (TAL) medium consisting of top thin layer of TSA (non-selective agar) and bottom layer of PALCAM Medium Base supplemented with PALCAM Antimicrobial Supplement (selective agar; Difco™, Becton, Dickinson and Co., Sparks, MD, USA) in order to recover injured cells and to selectively enumerate L. monocytogenes cells, respectively, followed by incubation for 48–72 h at 37 °C. The number of colonies was enumerated with a colony counter (Rocker Scientific Co. Ltd., Taipei, Taiwan) and expressed in log CFU/cm2.

Thiobarbituric acid reactive substance (TBARS) assay

A thiobarbituric acid reaction substance (TBARS) assay was carried out in order to analyze lipid peroxidation of smoked salmon by 405 nm LED illumination with or without riboflavin. Malondialdehyde (MDA), an indicator of lipid peroxidation, is produced naturally when ROS attack lipids containing a carbon–carbon double bond, in particular polyunsaturated fatty acids (PUFAs) (Ayala et al., 2014; Kim et al., 2017a).

Smoked salmon samples (ca. 2 g) containing 0–50 μM riboflavin were illuminated to 405 nm LED at 4 °C at a total dose of 2.76 kJ/cm2 in an incubator. Non-illuminated control samples were stored in the dark. To quantify lipid peroxidation, each of two non-illuminated and illuminated samples with or without riboflavin was withdrawn at a selective time interval. Each sample was mixed with 4 mL of deionized water, 40 μL of 100 × butylated hydroxytoluene (BHT; Sigma-Aldrich, St. Louis, MO, USA), and 4 mL of 10% trichloroacetic acid (TCA; Sigma-Aldrich) in a 50 mL Falcon tube, followed by vigorous vortexing. The mixture was centrifuged at 3,000 × g for 2 min. One milliliter of the supernatant was transferred to a new 15 mL Falcon tube and subsequently added to 1 mL of thiobarbituric acid (TBA; Sigma-Aldrich) reagent freshly prepared (5.2 mg TBA per 1 mL of 5% TCA). The mixture was incubated at 95 °C for 30 min and cooled to room temperature in an ice bath for 10 min. Afterwards, 200 μL of the mixture was transferred to a 96-well plate in duplicate and measured at 532 nm with a Synergy HT multi-detection Microplate reader (Bio-Tec Instruments Inc., Winooski, VT, USA). The degree of lipid peroxidation was calculated using a standard curve between 0 and 100 µM MDA (Merck, Darmstadt, Germany) and expressed as μg MDA per g fresh weight (FW).

Color analysis

The color of non-illuminated and illuminated salmon samples with or without riboflavin (0–100 μM) was measured with a reflectance spectrometer (CM-3500d; Konica Minolta Sensing Inc., Osaka, Japan) by set to a D65 illuminant and an observation angle of 10°. The color parameters of L* (lightness), a* (red to green), and b* (yellow to blue) were measured via reflectance values. The values were used to calculate the white index (WI) between non-illuminated control and illuminated samples with or without riboflavin using the following equations (Bermúdez-Aguirre and Barbosa-Cánovas, 2013):

$${\text{WI}} = 100 - \sqrt {\left( {100 - L^{*} } \right) + a^{* 2} + b^{* 2} }$$

Statistical analysis

All experiments were carried out in independent triplicate with duplicate sampling (n = 6). Data were expressed by mean ± standard deviation. Significant differences in the mean values were calculated at the 95% confidence interval (p < 0.05) using one-way analysis of variance (ANOVA) and means were separated by least significant difference (LSD) with the IBM SPSS statistical software (version 17.0; SPSS Inc., IBM Co., Armonk, NY, USA).

Results and discussion

Change in temperature during 405 nm LED illumination

The temperature fluctuation on the surface of smoked salmon was monitored for 300 min with 1-min interval during 405 nm LED illumination at different storage temperatures to determine the heat transfer from the LED to salmon samples. To simulate the ideal refrigeration condition, temperature abuse at food establishments, and the room temperature condition, the different storage temperatures at 4, 12, and 25 °C, respectively, were chosen in this study. Results show that the surface temperature of smoked salmon increased by approximately 4.5 °C within 30 min during LED illumination, irrespective of set temperatures, and the elevated temperature was maintained until the end of measurement when the set temperatures of the incubator were 0, 7.5, and 20.5 °C (Fig. 2). Hence, non-illuminated control salmon samples were performed at the incubator temperatures of 4, 12, and 25 °C, respectively, to eliminate the effect of temperature difference under non-illuminated and illuminated conditions during long-term storage. On the other hand, the bacterial suspension in PBS was held at the incubator temperature of 25 °C for both non-illuminated control and illuminated suspension with/without riboflavin due to short-time exposure.

Fig. 2
figure 2

Temperature profile on the surface of smoked salmon at 4 °C (A), 12 °C (B) and 25 °C (C) during 405 nm LED illumination

Full size image

Antibacterial effect of riboflavin mediated 405 nm LED illumination in PBS

Riboflavin, known as a water-soluble vitamin B2, is essential for the metabolism of nutrients and antioxidant protection for human health (Ashoori and Saedisomeolia, 2014). However, riboflavin is sensitive to lights in aqueous solution and its absorbance spectrum are a sharp peak at 270 nm and shorter peaks at 380 and 445 nm (Ashoori and Saedisomeolia, 2014; Josewin et al., 2018; Liang et al., 2013). Riboflavin could be photodegraded by several pathways including the excited states upon light exposure such as ultra-violet (UV) and visible light at blue wavelengths (Ashoori and Saedisomeolia, 2014; Liang et al., 2013), generating ROS as previously described. Thus, in this study, the antimicrobial activity of 405 nm LED illumination with the aid of exogenous riboflavin against L. monocytogenes in PBS at doses as high as 57.6 J/cm2 at 25 °C was evaluated. Two concentrations of riboflavin (50 and 100 μM) were selected in this study since the maximum water solubility of riboflavin is about 120 μM (Liang et al., 2015) and our previous study also showed that 460 nm LED illumination with 25 μM of riboflavin was less effective than 50 and 100 μM in inactivating L. monocytogenes on smoked salmon (Josewin et al., 2018). The present results show that LED illumination alone inactivated 1.9 log CFU/mL of L. monocytogenes cells at 57.6 J/cm2, while riboflavin-mediated LED illumination reduced the populations by 6.2 log CFU/mL (below the detection limit) at 19.2 J/cm2, revealing that the antibacterial effect of 405 nm LED illumination could be significantly (p < 0.05) enhanced in the presence of riboflavin (Fig. 3). There was no significant (p ≥ 0.05) difference in the inactivation of L. monocytogenes between 50 and 10 μM of riboflavin.

Fig. 3
figure 3

Antimicrobial activity of 405 nm LED illumination with different concentrations of riboflavin (RBF) on L. monocytogenes in phosphate-buffered saline (PBS) at 25 °C. Asterisk (*) indicates significant (p < 0.05) difference in cell counts between non-illuminated control and LED illumination with/without riboflavin. Hash (#) indicates significant (p < 0.05) difference in cell counts between LED illumination alone and riboflavin-mediated illumination

Full size image

Similarly, a previous study demonstrated that 400 nm LED irradiation (total irradiance of 260 mW/cm2) in pure water for 10 min significantly (p < 0.05) reduced 0.9 to 3.0 log CFU/mL of Streptococcus mutants, Escherichia coli, and Pseudomonas aeruginosa, whereas PDI in the presence of polyphenols, such as caffeic acid and chlorogenic acid, was enhanced by ≥ 5 log reduction (Nakamura et al., 2015). Another study conducted by Luksienė et al. (2013) showed that chlorophyllin-based 405 nm LED illumination in PBS for 5 min (at 3.6 J/cm2) inactivated by 7 log CFU/mL of L. monocytogenes cells, while aminolevulinic acid-based photosensitization caused less than 4 log reduction of the populations for 20 min (14.4 J/cm2). Wong et al. (2016) reported that 405 and 465 nm LED irradiation at 9 ± 1 °C enhanced the inactivation rate of S. aureus in the presence of riboflavin-5′-phosphate (FMN), but S. aureus populations were susceptible to 405 nm LED illumination than 465 nm LED, regardless of FMN. Unlike the present study, 465 nm LED irradiation with 60 and 120 μM FMN for 2 h (total doses of 14.4 J/cm2) inactivated by 41.1% and 94.9% of S. aureus populations, respectively, exhibiting that the higher concentration of FMN could cause the better inactivation rate of S. aureus under 465 nm LED exposure.

Behavior of L. monocytogenes on smoked salmon surface during LED illumination with/without riboflavin

For food application, the efficacy of 405 nm LED illumination with and without riboflavin (0–100 μM) on L. monocytogenes on the surface of smoked salmon was investigated at 4, 12, and 25 °C, respectively, until total doses of 1.27–2.76 kJ/cm2. As shown in Fig. 4, non-illuminated control cells on smoked salmon remained unchanged during storage at 4 and 12 °C, whereas riboflavin-mediated LED illumination inactivated L. monocytogenes populations on smoked salmon by 1.7 to 2.0 log CFU/cm2 until total doses of 2.76 kJ/cm2. But no further inactivation was observed at doses greater than 1.27 kJ/cm2, regardless of the riboflavin concentration. Interestingly, L. monocytogenes cells on smoked salmon were found to be the most (p < 0.05) susceptible to LED illumination alone after 1.96 kJ/cm2 at chilling temperatures, compared to riboflavin-mediated LED illumination. At 25 °C, non-illuminated control cells on smoked salmon grew to 6.1 log CFU/cm2 at the end of storage, while LED illumination reduced the populations of L. monocytogenes to 2.2–2.8 log CFU/cm2 at 1.27 kJ/cm2, regardless of the presence of exogenous riboflavin (Fig. 4c).

Fig. 4
figure 4

Effects of 405 nm LED illumination with different concentrations of riboflavin (RBF) on survival of L. monocytogenes on the surface of smoked salmon at 4 °C (a), 12 °C (b), and 25 °C (c). Uppercase letters (A–B) for the same exposure dose and lowercase letters (a–d) for the same bar indicate significant (p < 0.05) difference

Full size image

The antibacterial effect of 405 nm LED with and without the additional riboflavin on smoked salmon was considerably decreased when compared with that in PBS. Similarly, our previous study has shown that S. Enteritidis in PBS were more susceptible to 405 nm LED illumination than that on cooked chicken (Kim et al., 2017b). Another study conducted by Ghate et al. (2016) reported that 460 nm LED illumination inactivated by 5 log CFU/mL of Salmonella populations in orange juice. In contrast, 460 nm LED illumination alone on smoked salmon did not inactivate L. monocytogenes populations (Josewin et al., 2018). This might be because liquid samples can be evenly illuminated by LEDs, while the efficacy of LED illumination can be limited on the surface of solid food matrices due to uneven and opaque surface such as shadow effect.

Furthermore, unlike PBS, no synergistic effect was observed under 405 nm LED illumination combined with riboflavin on smoked salmon. Food components could also affect bacterial behaviors on smoked salmon by LED illumination. Salmon has high levels of fat contents (5.7 to 17.6%), especially polyunsaturated lipids (~ 38% of total fat contents) (Nicorescu et al., 2014; Wold et al., 1996). The double bonds in polyunsaturated lipids are susceptible to deterioration by oxidation, resulting in lipid peroxidation (Ma et al., 2018). Lipid peroxidation products, such as reactive aldehydes (MDA and 4-hydroxynonenal), can also react with DNA directly (Catalá, 2009; Ma et al., 2018). On the other hand, as described above, riboflavin is sensitive to light exposure, but also act as antioxidants under oxidative stress, particularly lipid peroxidation and oxidative injury (Ashoori and Saedisomeolia, 2014; Nicorescu et al., 2014), potentially playing a crucial role in preventing lipid peroxidation (Ma et al., 2018). Thus, it is postulated that some of unexcited riboflavin due to shadow effect on smoked salmon might act as antioxidant protection by quenching ROS generated by LED illumination.

In our previous study, 460 nm LED illumination with riboflavin on smoked salmon effectively inactivated L. monocytogenes by 0.5–1.6 log CFU/cm2 at 2.4 kJ/cm2 (Josewin et al., 2018), while 405 nm LED alone on smoked salmon nearly reached below the detection limit (< 1.4 log CFU/cm2) at chilling temperatures in this study. These findings indicate that 405 nm LED could be much more effective in inactivating L. monocytogenes than 460 nm LED. Amit et al. (2016) also reported that the greater antibacterial effect was observed in 405 nm LED than 460 nm LED. Bacterial inactivation by visible light can differ depending on the absorption spectrum of photosensitizers. For instance, porphyrin compounds naturally produced by some bacteria have a strong absorption band at around 400 nm (Kim and Yuk, 2017; Redi and Jori, 1988), resulting in the production of ROS by 405 nm LED illumination in the presence of oxygen and then indirectly damages to bacterial components, such as DNA, lipids, proteins, and membrane (Amit et al., 2016; Kim and Yuk, 2017). Another possible explanation could be because some portions of 405 nm LED spectrum were ranged in the UV which could directly cause DNA damage (Amit et al., 2016). For these reasons, 405 nm LED itself might enhance the bactericidal effect by generating much more ROS than 460 nm LED.

The present results show that storage temperatures did not affect the bactericidal efficacy of 405 nm LED illumination on smoked salmon, irrespective of the presence of exogenous riboflavin. Unlike this, our previous studies using 405 nm LED on Salmonella on fresh-cut papaya and cooked chicken proved the greater antibacterial effect at 4 °C than those of 10 and 20 °C (Kim et al., 2017a, 2017b). Kim et al. (2017c) also showed the greater antibacterial effect against L. monocytogenes on fresh-cut mango at lower temperatures of 4 and 10 °C than that of 20 °C, whereas Salmonella on mango were susceptible to LED illumination at all temperatures tested. On the contrary, the antibacterial effect of 460 nm LED illumination against Salmonella in orange juice was enhanced at room temperature rather than chilling temperatures (Ghate et al., 2016). These findings indicate that the different bacterial susceptibility to LED illumination might be dependent on species as well as be affected by experimental design, LED wavelengths, illumination temperatures, and food components such as antioxidants, lipids, and proteins (Kim et al., 2017a, b, c).

Lipid peroxidation and color change of smoked salmon by LED illumination with/without riboflavin

Lipids are known as one of chemically unstable food components and easily undergo oxidation and autoxidation processes (German, 1999; Nicorescu et al., 2014). Rancidity and off-flavor in salmon are caused by lipid peroxidation, being able to impact the sensory quality (Refsgaard et al., 1998). Therefore, chemical quality of illuminated smoked salmon with and without riboflavin should be evaluated since polyunsaturated lipids and color are known to be susceptible to light. In this work, no significant (p ≥ 0.05) changes in TBARS values were observed in non-illuminated smoked salmon with and without riboflavin during storage at 4 °C (Fig. 5). On the other hand, TBARS values for illuminated smoked salmon significantly (p < 0.05) increased with the increase in exposure doses, regardless of the presence of riboflavin, but the mean values of illuminated smoked salmon without riboflavin were higher than riboflavin-mediated illuminated smoked salmon. A previous study reported that pulsed light (PL) treatment at 30 J/cm2 induced lipid peroxidation in raw salmon samples (Nicorescu et al., 2014). Another study conducted by Pedrós-Garrido et al. (2018) revealed that PL and UV-C treatments for 30–90 s also caused the lipid peroxidation in salmon samples. In contrast, Colejo et al. (2018) demonstrated that lipid peroxidation in UV-C treated smoked salmon samples remained unchanged until 900 mJ/cm2, while TBARS values for smoked salmon treated with non-thermal atmospheric plasma (NTAP) after 4 min were higher than those for control untreated samples. However, lipid peroxidation observed in these previous studies did not exceed the consumption limit of 7 to 8 mg MDA/kg (Coleio et al., 2018; Oğuzhan, 2013), whereas our results were over the consumption limit probably due to the long-term exposure. Furthermore, the lipid peroxidation may induce color changes due to the co-oxidation of water-soluble and fat-soluble pigments such as carotenoids and myoglobin in food products (Nicorescu et al., 2014). Thus, color modification on illuminated smoked salmon samples was also investigated for further experiment.

Fig. 5
figure 5

Lipid peroxidation of smoked salmon by 405 nm LED illumination with different concentrations of riboflavin (RBF) at 4 °C. Uppercase letters (A–C) for the same exposure dose and lowercase letters (a–c) for the same bar indicate significant (p < 0.05) difference

Full size image

Whiteness index (WI) is useful to quantify the degree of discoloration during processing, which is generally used for consumer’s preference of white colors (Anyasi et al., 2015). In this study, changes in WI of smoked salmon illuminated by 405 nm LED illumination with and without riboflavin were measured to determine if ROS generated by LED illumination causes discoloration. As shown in Fig. 6, the WI values of non-illuminated smoked salmon with and without riboflavin remained unchanged (p ≥ 0.05) during storage in the dark, regardless of storage temperatures, although the WI values of non-illuminated smoked salmon with exogenous riboflavin were lower than those of non-illuminated control smoked salmon. On the other hand, LED illumination significantly (p < 0.05) increased by 8–14 of the WI values until 1.27–2.76 kJ/cm2, irrespective to storage temperatures and the presence of riboflavin, but the mean values of WI in illuminated smoked salmon were statistically (p < 0.05) higher than those of riboflavin-mediated illuminated samples. In contrast to the present results, our previous study has reported that no color change in illuminated smoked salmon with 460 nm LED alone was observed, while LED illumination with 100 mM riboflavin significantly (p < 0.05) increased the WI (Josewin et al., 2018). The color of salmon ranges from pink to orange-red and might be correlated to the content of astaxanthin that is one of the predominant carotenoids (Yagiz et al., 2010). Astaxanthin is sensitive to oxidation and irradiation due to its highly unsaturated structures (Takeungwongtrakul and Benjakul, 2016; Yagiz et al., 2010). For example, the amount of astaxanthin was decreased with the increase in irradiation dose, resulting in a loss of redness of smoked salmon (Yagiz et al. 2010). Unlike these findings, our previous studies have demonstrated that 405 nm LED illumination on fresh-cut mango and papaya did not adversely influence the color as well as contents of β-carotene, flavonoids, antioxidant capacity, and ascorbic acid compared to non-illuminated control fruits during long-term exposure, regardless of storage temperatures (Kim et al., 2017a; 2017c). The discrepancy in our previous and present studies may be attributed to types and contents of nutrition and bioactive compounds in food matrices.

Fig. 6
figure 6

Comparison of white index (WI) between 405 nm LED illuminated and non-illuminated smoked salmon with/without RBF at different temperatures of 4 °C (a), 12 °C (b), and 25 °C (c). Uppercase letters (A–D) for the same exposure dose indicate significant (p < 0.05) difference

Full size image

In conclusion, the results in this study demonstrate that 405 nm LED illumination in the presence of riboflavin could accelerate the inactivation rate of L. monocytogenes in aqueous solution, while the effect of riboflavin-mediated LED illumination was highly diminished on the surface of smoked salmon that in PBS. Interestingly, LED illumination alone could more effectively reduce L. monocytogenes populations on smoked salmon at chilling temperatures than riboflavin-mediated LED illumination. Although illumination with and without riboflavin caused lipid peroxidation and discoloration of smoked salmon, this study proved the potential of 405 nm LED to play a role in a food preservation technology for RTE fishery products. Such technology can be applied to preserve smoked fish susceptible to post-processing contamination and can be also applicable to food establishments where smoked salmon should be sold within 2–3 days. However, the optimization of LED illumination conditions is crucial to minimize the quality changes of smoked salmon during storage. Furthermore, a scale-up study is required to be applied to commercial food systems under actual operating conditions.