Innovative Food Science and Emerging Technologies 21 (2014) 90–98

Contents lists available at ScienceDirect

Innovative Food Science and Emerging Technologies

j ourna l homepage: www.e lsev ie r .com/ locate / i fset

Microbiological efficacy in liquid egg products of a UV-C treatment in acoiled reactor

Poliana Mendes de Souza a,b,⁎, Alexandra Müller b, Avelina Fernández a, Mario Stahl b

a Instituto de Agroquímica y Tecnología de Alimentos, Consejo Superior de Insvestigaciones Científicas, Department of conservation and quality, Avda. Agustín Escardino, 7, 46980 Paterna, Spainb Max Rubner-Institut, Federal Research Institute of Nutrition and Food, Department of Food Technology and Bioprocess Engineering, Haid-und-Neu-Str. 9, 76131 Karlsruhe, Germany

⁎ Corresponding author at: Universidade Federal dos VInstituto de Ciência e Tecnologia, Campus JK, RodoviaMGT3Diamantina, Brazil. Tel.: +55 38 98070903.

E-mail address: poliana.souza@ict.ufvjm.edu.br (P.M. d

1466-8564/$ – see front matter © 2013 Elsevier Ltd. All rihttp://dx.doi.org/10.1016/j.ifset.2013.10.017

a b s t r a c t

a r t i c l e i n f o

Article history:Received 8 August 2013Accepted 30 October 2013

Editor Proof Receive Date 3 December 2013

Keywords:UV-CNon-thermal pasteurizationEgg whiteWhole eggEgg yolkListeria innocuaEscherichia coliSalmonellaShelf lifeUltraviolet treatment

The feasibility of ultraviolet processing as a non-thermal pasteurization technique for liquid egg products was in-vestigated. Inactivation of Salmonella subterranea DSM 16208 followed comparable kinetics than Salmonellaenteritidis, therefore it was used as a non-pathogenic surrogate. The influence of the dose on the inactivation ofS. subterranea DSM 16208, Escherichia coli DH5α and Listeria innocuaWS 2258 in liquid egg fractions was evalu-ated on a laboratory scale recirculating UV-C treatment coiled reactor UVivatec® Lab (Bayer Technology ServicesGmbH, Leverkusen, Germany) and a higher scale laboratory device (coiled tube reactor MRI2010, designed andassembled at the Max Rubner-Institut, Karlsruhe, Germany). Analogous inactivation kinetics were observed forall microorganisms, in which inactivation curves followed a logarithmic decay and changed over to a tailingphenomenon after some time of UV-C exposure. Comparable inactivation behavior and doses to inactivatethe selectedmicroorganismswere detected for whole egg, flowing from 9.5 to 20 L h−1. Highest doses werenecessary to attain a 5 Log reduction on egg yolk. UV-C processing resulted in a shelf life extension of liquid egg upto 8 weeks at 4 °C, which corresponds to a storage period 4 times higher than the recommended for pasteurizedLWE.Industrial relevance: This novelUV-C systemcanbe applied successfully to the Food Industry. UV-C can be effectivelyused to reduce thenumber of spoilage andpathogenic bacteria, aswell as yeasts andmolds indifferent kinds of foodproducts.

© 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Novel food processing technologies include the use of physical factorsto process and preserve food. Most of them, such as high hydrostaticpressure, pulsed electric fields, cold plasma, ultrasound/cavitation, andultraviolet light, are non-thermal or operate at temperatures belowconventional heat treatments. Among them, ultraviolet technology atgermicidal wavelengths (UV-C) has shown perspectives for thepasteurization of liquid food products in appropriate designed reactors(e.g. Choudhary et al., 2011; Oteiza, Giannuzzi, & Zaritzky, 2010). Themain principle of UV-C decontamination is based on the formation ofphotoproducts in the DNA of microorganisms, which prevents theirDNA replication (Bintsis, Litopoulou-Tzanetaki, & Robinson, 2000).The UV absorbance spectrum of DNA has a maximum around 260 nm(Koutchma, 2009) being lowpressuremercury lampswith peak emissionat 254 nm extremely effective to inactivate microorganisms. In liquids,UV absorption and scattering due to solutes and particles are the mostimportant limiting factors which determine the UV-C penetration

ales do Jequitinhonha e Mucuri,67, km583, no 5000, 39100-000

e Souza).

ghts reserved.

depth (Koutchma, 2004). This limitation can be compensated by tech-nologies using centrifugal forces to provide thin films (Geveke &Torres, 2013) or by effective mixing to achieve Reynolds numbers inthe turbulent range (Franz, Specht, Cho, Graef, & Stahl, 2009; Müller,Stahl, Graef, Franz, & Huch, 2011).

Eggs areworldwide consumedproducts because of their high nutrientcontent. It is estimated that in 2015 the Global Eggs Consumption willreach 1154 billion units (Global Industry Analysts, 2011). Liquid eggproducts are highly susceptible to microbial contamination; howeverthey ought to be pasteurized at mild conditions due to the rheologicalchanges induced by heat. As a result, the shelf life of eggs and eggproducts, including liquid eggs, is limited to a few days even under re-frigeration. Several studies have shown that the organoleptic propertiesof UV treated liquid egg products are comparable to the untreated,being therefore excellent candidates for UV-C decontamination(Geveke & Torres, 2012; Souza & Fernandez, 2012). Challenges for theUV-C decontamination of egg products are in particular the high opticaldensity of liquid egg fractions, the light scattering due to the Tyndalleffect, and the high viscosities. Some microbiological studies havebeen performed on liquid eggs with successful results (Geveke, 2008;Ngadi, Smith, & Cayouette, 2003; Souza & Fernandez, 2011; Ünlütürk,Atilgan, Baysal, & Tari, 2007; Ünlütürk, Atilgan, Baysal, & Ünlütürk,2010). The effects of pH, depth of food medium and ultraviolet (UV-C)

Fig. 1. Principle of the UVivatec® Lab reactor and MRI2010 reactor UV-C modules (withkind permission of Bayer Technology Services).

91P.M. de Souza et al. / Innovative Food Science and Emerging Technologies 21 (2014) 90–98

light dose on the inactivation of Escherichia coli O157:H7 in egg whitewere studied by Ngadi et al. (2003). In addition, the efficiency of UV-Cirradiation as a non-thermal pasteurization process for liquid eggwhite was investigated by Ünlütürk et al. (2007, 2010) using a benchcollimated beam apparatus, and by Souza and Fernandez (2011) using abench apparatus with reflector, in both works the target microorganismwas Salmonella enteritidis. Although, themain challenge in the applicationof UV-C technology is the scale-up of the process at feasible treatmentconditions.

In this study, themicrobial efficacy of UV-C treatments in coiled tubereactors up to 20 L h−1 has been investigated using inoculated non-pathogenic Salmonella subterranea, Listeria innocua and Escherichia colistrains. The efficacy was evaluated in whole egg, egg white and eggyolk fractions. In addition, the evolution of the natural egg microfloraof the UV-C treated liquid egg has been studied and compared withheat treated samples during storage at 4 °C.

2. Material and methods

2.1. Egg samples

Fresh eggs were purchased from Gutshof-Ei GmbH (Schackendorf,Germany). They were of yellow shell, and weighted between 55 and61 g. After reception, eggs were inspected for shell integrity and storedunder refrigeration at 4 °C. Just before the experiments were carried out,the egg content (separately, egg whites and egg yolks) was removedunder aseptic conditions, and collected in 1 L sterile flasks. The chalazawas removed and the separated egg fractions were then homogenizedfor 1 min using a commercial blender (31BL44, Waring, USA) at maxi-mum speed. To prepare the whole egg samples, 13.3 mL of egg yolkwas mixed with 26.7 mL of egg white. This proportion is the normalaverage composition of whole egg and was used to attain a constantwhite to yolk relationship. Physical parameters of the used egg fractionsare shown in Table 1.

2.2. UV-C technology

2.2.1. UV-C chamber for surface treatmentThe UV-C chamber constructed by UV-Consulting Peschl® España

(Valencia, Spain), already described by Souza and Fernandez (2011) ismade of stainless-steel and provided with one low pressure mercurylamp (Heraeus Noblelight GmbH, Hanau, Germany), with maximumpeak radiation at 253.7 nm and energy output of 9 W. Chamber dimen-sions are 50 cm × 50 cm × 60 cm, the inner surface is flat black paintedto avoid light reflection at the walls. To enhance the amount of lightreaching the samples, an aluminium reflector surface covered the lamp.After UV emission stabilization, the lamp remained on; a shutter betweenthe lamp and the exposition chamber was used to protect the operatorwithout disturbing the operational conditions of the lamp. A magneticstirrer (Ovan MBG15, Barcelona, Spain) was installed at the central partof the lamp at the level of the sample container.

2.2.2. UVivatec® Lab reactorFor laboratory scale UV-C treatment the UVivatec® Lab reactor

(Bayer Technology Services GmbH, Leverkusen, Germany) (Fig. 1a), de-scribed among others by Schmidt and Kauling (2007) and Müller et al.

Table 1Physical properties of the egg fractions.

Egg white Whole egg Egg yolk

Optical densityat 254 nm

42.03 ± 3.67 730.67 ± 65.85 1266.67 ± 135.77

Turbidity (NTU) 276.24 ± 22.64 9128.05 ± 1475.80 10827.89 ± 136.89Dynamic viscosity(mPa s)

3.90 ± 0.28 8.17 ± 0.04 86.95 ± 4.63

Density (g cm−3) 1.0414 ± 0.0015 1.0354 ± 0.0025 1.0270 ± 0.0050

(2011) was used. The 9 W UV-C lamp (UVP, Type CPQ-7731) is locatedinside a quartz glass tubewhich is wrapped around by a Teflon® helicoid(Fig. 1a) (Poggel, Wübben, Brod, Jenne, & Schmidt, 2008). The lamp usedis a low pressure mercury lamp with maximum peak radiation at254 nm. The liquids can be pumped through the device at flow rates be-tween 2 and 20 L h−1 by a peristaltic pump.

2.2.3. MRI2010 UV-C reactorThe coiled tube reactor MRI2010 (Fig. 1b) was designed and

assembled at the Max Rubner-Institut (Karlsruhe, Germany). Themain component is a module which consists of a PTFE envelope witha diameter of 3.7 mm (Adtech Polymer Engineering Ltd., Stroud, UK)which is wound around a 30 W low pressure mercury lamp with max-imum peak radiation at 254 nm (UVN 30, UV Technik SpeziallampenGmbH, Wümbach, Germany) (Fig. 1b). The liquids can be pumpedthrough the device atflow rates between 4 and 40 L h−1 by a peristalticpump (Pump drive Pd 5206, Heidolph, Schwabach, Germany).

2.2.4. UV-C dosimetrySeveral methods are used to compare the efficacy of UV-C treatments.

The first method relied on the fluence (Ds in J cm−2) defined as the con-stant fluence rate multiplied by the exposure time in seconds referring tothe available treatment surface (Bolton & Linden, 2003). TheUV-C dosageper liter (Dv in J L−1) of treated liquid for a reactorwith continuousflow iscalculated as the UV-C output of the lamp (W) per flow rate (L s−1)(Keyser, Müller, Cilleirs, Nel, & Gouws, 2008; Wright, Summer, Hackey,Pierson, & Zoecklein, 2000) and used as dose definition in this researchproject. To compare the inactivation method with others based on rele-vant energy input (Del in J L−1), the thirdmethod relied on the necessaryelectrical energy input of the lamp (W)perflow rate (L s−1) (Müller et al.,2011).

To specify the flow condition in the reactor at several flow ratesReynolds numbers (Re) were calculated.

Re ¼ u� dυ

¼ u� d� ρη

ð1Þ

where d is the diameter of the tube, u is the velocity (m s−1), υ is thekinematic viscosity (m2 s−1), η is the dynamic viscosity (Pa × s) andρ is the mass density (kg m−3).

Table 2 gives an overview of the different dose values and Reynoldsnumbers used in this study.

Table 2Used UV-C doses per cycle and Reynolds numbers.

Egg fraction Flow rate ElectricalenergyDel (J L−1)

SurfacedoseDs (mJ cm−2)

UV-C energy outputper volumeDv (J L−1)

Re

Egg white 14.5 2238 43.3 468 432.3Whole egg 9.5 3417 66.1 715 129.6Egg yolk 2.6 12,558 243.0 2626 2.6Whole egga 20.0 6480 n.a.b 3600 242.3

a MRI2010 reactor.b n.a. means not applicable.

Table 3Summary of the analyzed populations, together with their growth medium, platingtechnique and incubation conditions (Gonzalez et al., 2009).

Populations Growthmedium

Technique Incubationconditions

Total aerobic mesophiliccount (TAM)

Standard I Pour plate 30 °C, 3 days, aerobic

Total anaerobic mesophiliccount (TAnM)

DRCM Pour plate 30 °C, 3 days, anaerobic

Total aerobic psychrotrophiccount (TAP)

Standard I Pour plate 22 °C, 5 days, aerobic

Total anaerobic psychrotrophiccount (TAnP)

DRCM Pour plate 22 °C, 5 days, anaerobic

Total aerobic spore count (TAS) Standard I Pour plate 30 °C, 3 days, aerobicTotal anaerobic spore count(TAnS)

DRCM Pour plate 30 °C, 3 days, anaerobic

Lactic acid bacteria (LAB) MRS Pour plate 30 °C, 3 days, aerobicYeast and molds (YM) YGC Spread plate 30 °C, 3 days, aerobicPseudomonas spp. (PS) PAB-CFC Spread plate 30 °C, 2 days, aerobicEnterobacteriaceae (EN) VRBG Spread plate 37 °C, 1 day, aerobicSalmonella spp. (SA) BSA Spread plate 37 °C, 1 day, aerobic

92 P.M. de Souza et al. / Innovative Food Science and Emerging Technologies 21 (2014) 90–98

2.3. Microorganisms and culture media

S. subterranea DSM 16208 is a non-pathogenic salmonella strain (riskgroup 1, classification of DSMZ) isolated from uranium contaminatedsubsurface sediment (Shebolina, Sullivan, O'Neill, Nevin, & Lovley,2004) and selected for this study as substitute (Fig. 2) of the pathogenicstrain Salmonella enterica subsp. enterica ser. Enteritidis ATCC 13076.

L. innocuaWS 2258 isolated byWeihenstephan (Freising, Germany)was used in this study as substitute for cold-tolerant food spoilagemicroorganisms. Escherichia coliDH5α a commonK12-derived laboratorystrain was also used for the inactivation studies. Strains were stored inStandard I Broth (Merck, Darmstadt, Germany) with 20% glycerol at −80 °C. Stock cultures were kept by regular subculture on Standard IBroth. Prior to inoculation of egg fractions, an overnight culture was pre-pared with 5 mL of the subculture in 500 mL of Standard I and incubatedat 37 °C for 18 h.

For inactivation trials, the overnight culture was centrifuged at7500 rpm for 10 min at 4 °C (Sorvall RC-26 Plus, Thermo Fisher ScientificInc., Waltham, USA) and the pellet was resuspended in quarter strengthringer solution (QSRS) until the optical density equivalent to 8 to 9 LogCFU mL−1was obtained. The suspension ofmicroorganismswas decimaldiluted in fresh egg samples, to obtain a concentration of cells around 7 to8 Log CFU mL−1. For enumeration, decimal dilutions of the sampleswere made in QSRS and surface plated in duplicate on Standard I Agar(Merck, Darmstadt, Germany). The plates were incubated at 37 °C for24 h and counted. All experiments were performed in triplicate.

The egg naturally occurringmicroorganismswhichwere analyzed inthe shelf life study are total aerobic mesophilic count (TAM), totalanaerobic mesophilic count (TAnM), total aerobic psychrotrophiccount (TAP), total anaerobic psychrotrophic count (TAnP), total aerobicspore count (TAS), total anaerobic spore count (TAnS), lactic acid bacteria(LAB), yeasts andmolds (YM), Pseudomonas spp. (PS), Enterobacteriaceae(EN) and Salmonella spp. (SA). Table 3 presents the growth media, platetechnique and the incubation conditions used on this research project.

-5

-4

-3

-2

-1

0

0 2 4 6 8 10 12 14 16

Dose (J cm-2)

Log

(N

N0-1

)

S. enteritidis EW S. subterranea EWS. enteritidis WE S. subterranea WES. enteritidis EY S. subterranea EY

Fig. 2. Comparison of UV-C surface inactivation of Salmonella enteritidis ATCC 13076 andSalmonella subterranea DSM 16208 at a fluence of 3.94 mW cm−2 for egg white (EW),whole egg (WE) and egg yolk (EY). Study made at a UV-C chamber constructed byUV-Consulting Peschl® España (Valencia, Spain), described by Souza and Fernandez (2012).

2.4. UV-C inactivation of microorganisms in liquid egg

2.4.1. UV-C surface inactivation of Salmonella strainsA series of trials was conducted to evaluate the effect of UV-C on the

inactivation of S. subterranea DSM 16208 and S. enteritidis ATCC 13076population on liquid egg products. Previous work was conducted withS. enteritidis (Souza & Fernandez, 2011), a commonmicroorganism asso-ciated with eggs cause of salmonellosis. In this study the S. subterranea(Shebolina et al., 2004) was used as an innocuous target substitute forS. enteritidis. The inoculated egg samples (12 mL volume) were placedin polystyrene Petri dishes (60.3 cm2) in the central part of the UV-Cchamber. They were treated in batch under continuous stirring ata fluence of 3.94 mW cm−2. The corresponding surface doseswere 2.1 J cm−2, 4.2 J cm−2, 6.3 J cm−2, 8.4 J cm−2, 10.5 J cm−2 and15.5 J cm−2 for S. subterranea and 1.3 J cm−2, 2.5 J cm−2, 5.2 J cm−2,7.6 J cm−2 and 15.4 J cm−2 for S. enteritidis.

2.4.2. UV-C inactivation using the UVivatec® Lab reactorA series of trialswas conducted to evaluate the effect of UV-C dose on

the inactivation of non-pathogenic strains of pathogenic food spoilagemicroorganisms on liquid egg products. Inoculated egg samples wererecycled through the reactor at flow rates of 14.5 L h−1 for egg white,9.5 L h−1 for whole egg and 2.6 L h−1 for egg yolk, which lead to aUV-C output per volume per cycle of 0.468 kJ L−1, 0.725 kJ L−1 and2.626 kJ L−1, respectively (Table 2). Samples of egg white were takenafter 0 to 21 passes through the reactor; whole egg samples weretaken after 0 to 25 cycles and egg yolk samples were collected after 0to 24 passes. The interval of pass sampling was done according to need.

2.4.3. UV-C inactivation using the MRI2010 reactorA series of trials was conducted to evaluate the effect of UV-C dose

on the inactivation of S. subterranea DSM 16208, Escherichia coli DH5αand L. innocua WS 2258 populations on liquid whole egg at a higherflow rate of 20 L h−1. Dose per cycle amounts to 3.6 kJ L−1 (Table 2).Samples of egg white were taken after 0, 1, 2, 3, 4 and 5 passes.

2.5. Shelf life studies

Three replicate shelf life studies were performed to evaluate theeffect of ultraviolet processing and pasteurization on microbial growthduring refrigerated storage, taking untreated liquid egg products asreference for the initial counts after the treatment. Before breaking,the shells of the eggs were cleaned with Bacillol® AF (Bode, Hamburg,Germany).

All fractions were UV-C treated at room temperature using the dosenecessary for a 5 Log reduction on the inoculated bacteria kinetic study

-8

-6

-4

-2

0

0 2000 4000 6000 8000 10000

Dose (J L-1)

Log

(N

N0-1

)

E. coliL. innocuaS. subterranea

Fig. 3.UV-C inactivation of Escherichia coliDH5α, Listeria innocuaWS 2258 and Salmonellasubterranea DSM 16208 on egg white using the UVivatec® Lab reactor, in function of thevolumetric dose (Dv).

-8

-6

-4

-2

0

0 4000 8000 12000 16000 20000

E. coli

L. innocua

S. subterranea

Log

(N

N0-1

)

Dose (J L-1)

Fig. 4.UV-C inactivation of Escherichia coliDH5α, Listeria innocuaWS 2258 and Salmonellasubterranea DSM 16208 on whole egg using the UVivatec® Lab reactor, in function of thevolumetric dose (Dv).

93P.M. de Souza et al. / Innovative Food Science and Emerging Technologies 21 (2014) 90–98

which was 9.8 kJ L−1 for egg white, 63.0 kJ L−1 for egg yolk and18.0 kJ L−1 for whole egg.

For the conventional pasteurization, 15 mL tubes containing 10 mLof samplewere treated using a thermostatic bath (Unitronic OR, Selecta,Spain) set to 56.6 °C, 60 °C and 61.1 °C, with a deviation of ±0.5 °C,respectively for egg white, whole egg and egg yolk. The conditions forpasteurization were chosen in conformity with the requirements ofthe USDA (USDA ARS 74-48, 1969). The holding time used for thethree fractions was 3.5 min when the coldest point of the sampleattained the pasteurization temperature, and after this period the sampleswere cooled within 15 s until 4 °C using a water/ice bath.

Immediately after processing, the samples were aseptically trans-ferred in sterile flasks and aerobically stored at 4 °C. On a weekly basissampleswere asepticallywithdrawn from the stored samples formicro-bial enumeration. The shelf life studies were conducted over a storageperiod of 8 weeks to ensure that 5 weeks of shelf life currently appliedfor heat pasteurized liquid whole egg is satisfied. The European Regula-tion 2073/2005 only limits the number of Enterobacteriaceae (targetafter production of 1 Log CFU mL−1 and tolerance after production of2 Log CFU mL−1), and Salmonella spp. (absent in 25 g). In this studyadditional microbial guidelines were used following the recommenda-tions of Gonzalez et al. (2009) for whole egg: 6 Log CFU mL−1 fortotal aerobic psychrotrophic counts, 7 Log CFU mL−1 for lactic acidbacteria and 5 Log CFU mL−1 for yeasts and molds.

2.6. Statistical analysis

One-way analysis of variance (ANOVA) was performed with the soft-ware XLSTAT-Pro (Win) 7.5.3 (Addinsoft, NY). Comparisons betweentreatments were evaluated with the Tukey test. Statistical analysis wasrun with a confidence level of 95%.

3. Results and discussion

3.1. Comparing results on S. enteritidis ATCC 13076 and the non-pathogenicstrain S. subterranea DSM 16208 in liquid egg products

Fig. 2 shows the response to the ultraviolet light of S. subterraneaDSM 16208 in liquid egg products compared to the literature datapresented by Souza and Fernandez (2011) for S. enteritidis ATCC13076. It can be observed that the behavior of both Salmonella strains issimilar in all egg fractions, being the S. subterraneaWS 2258 more resis-tant than the S. enteritidis ATCC 13076, even if the data do not statisticallydiffer. For eggwhite a 4.3 and 4.2 Log reductionwas achieved at a surfacedose of 7.6 and8.4 J cm−2 for S. enteritidis and S. subterranea, respectively,with a tailing of both dose response curves at higher doses. In liquidwhole egg both Salmonella strains show a 3.0 Log reduction at a dose of15.4 and 15.5 J cm−2, respectively. However in egg yolk, only a 2.7 Logreduction was achieved for S. enteritidis at a dose of 15.4 J cm−2

and S. subterranea shows a 2.4 Log reduction at 15.5 J cm−2. Eggshave been themost common food source linked to S. enteritidis infections.S. subterraneawasdescribed by Shebolina et al. (2004). ForUV-C inactiva-tion studies in egg products, the S. subterraneaDSM 16208 was thereforeused as a target microorganism replacing the pathogenic S. enteritidisATCC 13076. But further studies must be carried out to confirm this mi-croorganism as surrogate in other kinds of processes.

3.2. UV-C inactivation kinetics of spoilage microorganisms in liquidegg products

Semi-logarithmic survivor curves of S. subterranea DSM 16208,Escherichia coli DH5α and L. innocua WS 2258 reached using theUVivatec® Lab reactor are represented in Figs. 3, 4 and 5 for eggwhite, whole egg and egg yolk, respectively, relating the viablemicroor-ganism population (Log (NN0

−1)) to the UV-C dose (Dv). An effective re-duction of 5 Log (NN0

−1) for the investigated microorganisms was

obtained for egg white around 9.8 kJ L−1 and around 18.0 kJ L−1 forwhole egg, and 4.8 Log (NN0

−1) was obtained for egg yolk at a dose of63.0 kJ L−1. Tailing phenomena were evident after 4.0 kJ L−1 for eggwhite, 6.0 kJ L−1 for whole egg and 30.0 kJ L−1 for egg yolk. Differentparameters explain the microbial response and the tailing phenomena.The specific response of microorganisms to physical inactivationmethods is driven by physiological factors. A survival curve is consid-ered the cumulative distribution of lethal events for all individuals of apopulation. Therefore sensitive microorganisms can be expected to beinactivated at a faster rate, leaving behind a fraction of more resistantsurvivors (Peleg, 2000). Moreover, the characteristic physical parametersof the egg fractions, in particular the differences in optical density andturbidity (Table 1), are conditioning the degree of inactivation achievedbeing probably the egg low light transmittance more important thanthe suspended particles (Geveke, Boyd, & Zhang, 2011). The opticaldensities at 254 nm of egg white (42.03), whole egg (630.67) andegg yolk (1266.67) explain the effectiveness of the treatments (eggwhite N whole egg N egg yolk) in the UVivatec® Lab reactor as well asin the UV-C chamber (Figs. 2, 3, 4, 5). In general, the efficacy of UV-C in-activation of microorganisms decreases with increasing optical density(Koutchma, Keller, Chirtel, & Parisi, 2004;Müller et al., 2011;Murakami,Jackson, Madsen, & Schickedanz, 2006).

Fig. 6 presents the inactivation kinetics for the threemicroorganismsinoculated inwhole egg and treated in theMRI2010 reactor. The volumeflow through the device was 20 L h−1, which doubles the flow used onthe UVivatec® Lab reactor. The kinetic parameters are comparable tothe kinetics obtained for the same microorganisms in the UVivatec®Lab reactor. Around 6.0 and 8.0 kJ L−1 the tailing phenomena startedto be observed and 5 Log reductions were attained around 16.0 and18.0 kJ L−1. Other authors working with eggs did not reach doseshigher than 1 kJ L−1, limiting the studies on the linear decay assays(Geveke et al., 2011; Ünlütürk et al., 2007, 2010). Based on the energy

-8

-6

-4

-2

0

0 15000 30000 45000 60000 75000

E. coli

L. innocua

S. subterranea

Log

(N

N0-1

)

Dose (J L-1)

Fig. 5. UV-C inactivation of Escherichia coliDH5α, Listeria innocuaWS 2258 and Salmonellasubterranea DSM 16208 on egg yolk using the UVivatec® Lab reactor, in function of thevolumetric dose (Dv).

94 P.M. de Souza et al. / Innovative Food Science and Emerging Technologies 21 (2014) 90–98

input, less energy is needed to reach the same dose and the same inac-tivation rate on the MRI reactor in comparison with the UVivatec® Labreactor (Table 2). The difference in efficacy is probably caused by thecharacteristics of the lamps and the geometric dimensions of the twodevices. Therefore reactor design could be further optimized.

3.3. Shelf life of UV-C treated liquid egg

To confirm the extension of the shelf-life in liquid egg products, theanalysis of the impact of UV-C processing on the microbial contamina-tion was carried out at specific intervals with samples treated in theUVivatec® Lab reactor. Samples were preserved under refrigeration at4 °C for 8 weeks. The evolution of the spoilage-related microorganismswas compared with heat pasteurized samples. Conditions evaluatedwere defined during previous experiments on the microbial quality ofliquid egg products.

3.3.1. Initial countsThe prevalence of microorganisms in eggs is mainly due to eggshell

contamination during housing and handling. Main factors are relatedto the deposition of fecal material on the shell, the ovarium/oviductgut flora, breaks in the egg shell, or lack of hygienic conditions duringmanipulation or storage. Eggs are prone to be contaminated byGram+ and Gram− microorganisms, being microbial growth consid-erably decelerated due to the presence of natural antimicrobials, in par-ticular lysozyme and conalbumin, in the egg white fraction. In thisstudy, the initial counts of all fresh egg fractions were similar and addup to 4.9 ± 0.3 Log CFU mL−1 for total aerobic mesophilic count(TAM), 4.9 ± 0.3 Log CFU mL−1 for total anaerobic mesophilic count(TAnM), 5.0 ± 0.4 Log CFU mL−1 for total aerobic psychrophilic count

-8

-6

-4

-2

0

0 4000 8000 12000 16000 20000

E. coli

L. innocua

S. subterranea

Log

(N

N0-1

)

Dose (J L-1)

Fig. 6. UV-C inactivation of Escherichia coliDH5α, Listeria innocuaWS 2258 and SalmonellasubterraneaDSM 16208 onwhole egg treated byUV-C using theMRI2010 reactor, in func-tion of the volumetric dose (Dv).

(TAP) and 4.7 ± 0.3 Log CFU mL−1 for total anaerobic psychrophiliccount (TAnP) (Fig. 7a, b and c). More specifically, 4.7 CFU mL−1 wasidentified as lactic acid bacteria (LAB), 2.8 ± 0.3 Log CFU mL−1 yeastsand molds (YM), 4.1 ± 0.4 Log CFU mL−1 Pseudomonadaceae (PS),4.8 ± 0.4 Log CFU mL−1 Enterobacteriaceae (EN) and 1.9 ± 0.2 LogCFU mL−1 Salmonella spp. (SA). Spores (TAS and TAnS) were not de-tected in the fresh egg fractions.

The initial counts were in the range presented by Gonzalez et al.(2009) in liquid whole egg. But in their work, Salmonella was belowthe detection limit. Campylobacter spp., L. innocua and Salmonella iso-lates have been detected by Jones, Anderson, and Guard (2012) duringrandom sampling of conventional cage and free-range egg productions.Prevalence of Salmonella spp. is irregular and highly dependent on thehousing system and the processing conditions (Martelli & Davies,2012), however internal egg contamination is usually associated tothe serovar Enteritidis. Commercial practices are expected to reducethe microbial load of egg shell surfaces (Musgrove, Jones, Northcutt,Harrison, & Cox, 2005) and decrease the risks associated to microbialcontamination.

At the conditions established above, UV-C treatments in theUVivatec® Lab reactor resulted in lower microbial loads than in heatpasteurized samples (Fig. 7a, b, and c). Heat pasteurized samplesshow typically a large number of TAM and TAP microorganisms, butthose groups are not always able to reactivate during refrigerated storage(Garcia-Gonzalez et al., 2009). In egg white, initial counts were reducedup to the detection limit by UV-C treatment. In pasteurized egg white3.2 ± 0.1 Log CFU mL−1 of TAM and 2.9 ± 0.1 Log CFU mL−1 of TAPremained. After UV-C treatment on whole egg the mesophilic andpsychrophilic counts were around 1 Log CFU mL−1 and all the othermicrobial groups were under the detection limit. However, in wholeegg 3.4 ± 0.2 CFU mL−1 of TAM and 3.1 ± 0.2 CFU mL−1 of TAPwere detected after pasteurization. In egg yolk, the countswere approx-imately 1.5 Log CFU mL−1 after UV-C processing, except for spores andSalmonella which remained below the detection limit. For TAM, TAnMand TAP a higher count at around 2–3 Log CFU mL−1 was detected inpasteurized egg yolk. The assays showed absence of Salmonella spp. in25 g of sample after UV-C treatment and pasteurization.

The resistance of microorganism to UV-C is determined principallyby its ability to repair the DNA damage, in general they are ordered interms of complexity, being Gram-negatives b Gram-positives b yeast bbacterial spores b molds b virus (Almazara, 2007). The inactivationattained for the naturalmicroflora is similar in almost allmicrobial groupsstudied, only yeasts and molds were more resistant than prokaryotes.

3.3.2. StorageDuring 8 weeks of storage of egg white, no significant growth was

observed for the TAnM after UV-C or thermal pasteurization (Fig. 8aand b). Anaerobic mesophiles are generally not reactivated afterpasteurizationwith high pressure carbon dioxide or heat pasteurization(Garcia-Gonzalez, Geeraerd, Elst, Van Ginneken, & Van Impe, 2009).Contrary to this, the TAM showed a small and constant increase afterUV-C that did not attain 2 Log CFU mL−1 during storage, whereas thecounts of TAM in pasteurized egg white increase from week 3 to week6 up to 7.5 ± 0.4 Log CFU mL−1 (Fig. 9a and b). The TAnP countsstarted to increase on the 4th and 7th weeks of shelf life after pasteuri-zation and UV treatment, respectively (Fig. 11a and b). The TAP afterUV-C treatment presents a logarithmic growth on the fourth week ofstorage, attaining 2.2 ± 0.2 Log CFU mL−1 on the last day of analyses,whereas the counts after pasteurization showed an increase to7.6 Log CFU mL−1 attained in week 6 (Fig. 10a and b). For the LABcounts, samples treated with UV-C remained with steady values(Fig. 12a). However, pasteurized egg white showed an increase of lacticacid bacteria after the 5th week (Fig. 12b).

Similar to results shown by Garcia-Gonzalez et al. (2009), no reacti-vation was observed in the UV-C treated samples for the TAnM and LAB(Figs. 8a and 12a). The TAM showed a small and constant increase that

Fig. 7. Inactivation of naturally occurring microorganisms in liquid egg products by UV-C and heat treatment; (a) eggwhite, (b) whole egg and (c) egg yolk. The dashed line represents thedetection limit.

95P.M. de Souza et al. / Innovative Food Science and Emerging Technologies 21 (2014) 90–98

Fig. 8. Changes in naturally occurring microorganisms of egg white (white bars), wholeegg (gray bars) and egg yolk (black bars) UV-C treated during storage at 4 °C. (a) TAnMUV process, (b) TAnM pasteurization. The dashed line represents the detection limit.

96 P.M. de Souza et al. / Innovative Food Science and Emerging Technologies 21 (2014) 90–98

did not attain 2 Log CFU mL−1 during 8 weeks of storage in UV treatedwhole egg (Fig. 8a). The TAnP counts started to increase on the 6th weekof shelf life (Fig. 11a). AfterUV treatment, the TAPpresented a logarithmic

b

0

1

2

3

4

5

6

7

8

0 7 14 21 28 35 42 49 56

Storage time (days)

a

0

1

2

3

4

5

6

7

8

0 7 14 21 28 35 42 49 56

Storage time (days)

Lo

g (

CF

U m

L-1

)L

og

(C

FU

mL

-1)

Fig. 9. Changes in naturally occurring microorganisms of egg white (white bars), wholeegg (gray bars) and egg yolk (black bars) UV-C treated during storage at 4 °C. (a) TAMUV process, (b) TAM pasteurization. The dashed line represents the detection limit.

growth on the fourthweek of storage, attaining 1.8 ± 0.2 Log CFU mL−1

on the last day of analyses (Fig. 10a). In pasteurizedwhole egg, the countsof TAM, TAP, TAnP and LAB increased after week 4 (Figs. 8b, 10b, 11b and12b). However, no growth was detected for TAnM counts after pasteuri-zation during 8 weeks of storage (Fig. 9b).

In UV-C treated egg yolk, even if the initial counts after treatmentwere higher than in the other fractions, no growth was observed for theTAnM and LAB during 8 weeks of storage (Figs. 8a and 12a). The TAMshowed a small and constant increase in UV-C treated egg yolk and theTAnP counts started to increase after week 6 (Figs. 8a and 11a). Sincethefirstweek after theUV-C exposure the TAP presented a linear growth,attaining 2.7 ± 0.1 Log CFU mL−1 in the last day of analyses (Fig. 10a).In pasteurized egg yolk samples the counts of TAM, TAP and TAnP in-creased after the 4th week (Figs. 8b, 10b and 11b). However, the countsof TAnM and LAB showed a small but constant increase during storage(Figs. 8b and 12b).

In all pasteurized and UV-C treated egg fractions, no increase of YM,Pseudomonas spp., Enterobacteriaceae and Salmonella spp. TAS andTAnS was detected during 8 weeks of storage (data not shown).

Remarkably, the critical counts for the determination of the endof shelf life were not reached during 8 weeks of storage at 4 °C forthe UV-C treated samples. Gonzalez et al. (2009) reported an in-crease of 5 weeks in shelf-life in liquid whole egg at 4 °C with a treat-ment with high pressure carbon dioxide. In the case of the thermallytreated eggs, values were critical after week 5, in accordance to theexpected shelf-life of heat pasteurized liquid egg. The TAP countswere 5.3 ± 0.3 Log CFU mL−1 for egg white, 6.1 ± 0.2 Log CFU mL−1

for whole egg and 6.0 ± 0.1 Log CFU mL−1 for egg yolk. In week6, the TAP counts were superior to the established safety criteriaof 6 Log CFU mL−1 and the experiments for heat treated process wereinterrupted at this point.

Fig. 10. Changes in naturally occurring microorganisms of egg white (white bars), wholeegg (gray bars) and egg yolk (black bars) UV-C treated during storage at 4 °C. (a) TAPUV process, (b) TAP pasteurization. The dashed line represents the detection limit.

Fig. 11. Changes in naturally occurring microorganisms of egg white (white bars), wholeegg (gray bars) and egg yolk (black bars) UV-C treated during storage at 4 °C. (a) TAnPUV process, (b) TAnP pasteurization. The dashed line represents the detection limit.

Fig. 12. Changes in naturally occurring microorganisms of egg white (white bars), wholeegg (gray bars) and egg yolk (black bars) UV-C treated during storage at 4 °C. (a) LABUV process, (b) LAB pasteurization. The dashed line represents the detection limit.

97P.M. de Souza et al. / Innovative Food Science and Emerging Technologies 21 (2014) 90–98

4. Conclusion

S. subterraneaDSM16208 showed similar UV-C sensitivity in compar-ison to S. enteritidis ATCC 13076. UV-C light in the UVivatec® Lab reactoratflow rates of about 14.5 L h−1with a Renumber of 432.3 for eggwhite,9.5 L h−1 with a Re number of 129.6 for whole egg and 2.6 L h−1 with aRe number of 2.6 for egg yolk was effective to inactivate S. subterraneaDSM 16208, Escherichia coli DH5α and L. innocuaWS 2258, and 5 Log re-ductions of all inoculated microorganisms were attained at doses of8.0 kJ L−1, 16.0 kJ L−1, and 63.0 kJ L−1 respectively. Comparable doseswere required at 20 L h−1 with a Re number of 242.3 at the MRI re-actor. The long term microbial stability of the liquid egg productswas positively influenced by UV-C treatments and shelf-life was upto 8 weeks of refrigerated storage. Under the chosen treatment con-ditions, microbial growth during refrigerated storage was smallerthan in thermally pasteurized egg products. Therefore, the resultsobtained confirm UV-C treatment as a promising technology tolengthen the commercial shelf-life of liquid egg products. Furtherprocess optimization with equipment able to increase the Reynoldsnumbers, or the combination with additional hurdles might increasethe possibilities of UV-C as a non-thermal alternative to heat pasteuriza-tion for liquid egg products.

Acknowledgments

P.M. de Souza thanks the Generalitat Valenciana for grant numberBFPI/2008/230 and grant number BEFPI/2011/027. Authors are indebtedto theMax Rubner-Institut, Department of Safety andQuality of Fruit andVegetables for the support during the performance of the experimentsand the Department of Food Technology and Bioprocess Engineering,namely Dr. C. Franz, Dr. M. Huch and M. Knörr. Authors also thankDr. H. Brod, Dr. M. Poggel, Ms. M. Wübben, and Mr. A. Pastor fromBayer Technology Services for their kind support and supply of theUVivatec® Lab reactor.

References

Almazara, S. M. (2007). La radiación ultravioleta: una alternative “no térmica” para lapasteurización de jugos. Ingeniería Alimentaria, 103–108.

Bintsis, T., Litopoulou-Tzanetaki, E., & Robinson, R. K. (2000). Existing and potential appli-cations of ultraviolet light in the food industry — A critical review. Journal of theScience of Food and Agriculture, 80(6), 637–645.

Bolton, J. R., & Linden, K. G. (2003). Standardization of methods for fluence (UV dose)determination in bench-scale UV experiments. Journal of EnvironmentalEngineering—ASCE, 129(3), 209–215.

Choudhary, R., Bandla, D., Watson, D.G., Haddok, J., Abughazaleh, A., & Bhattacharya, B.(2011). Performance of coiled tube ultraviolet reactors to inactivate Escherichia coliW1485 and Bacillus cereus endospores in raw cow milk and commercially processedskimmed cow milk. Journal of Food Engineering, 107(1), 14–20.

Franz, C. M.A. P., Specht, I., Cho, G., Graef, V., & Stahl, M. R. (2009). UV-C-inactivation ofmicroorganisms in naturally cloudy apple juice using novel inactivation equipmentbased on Dean vortex technology. Food Control, 20, 1103–1107.

Garcia-Gonzalez, L., Geeraerd, A. H., Elst, K., Van Ginneken, L., & Van Impe, J. F. (2009). In-activation of naturally occurring microorganisms in liquidwhole egg using high pres-sure carbon dioxide processing as an alternative to heat pasteurization. Journal ofSupercritical Fluids, 51(1), 74–82.

Geveke, D. J. (2008). UV inactivation of E. coli in liquid egg white. Food BioprocessTechnology, 1, 201–206.

Geveke, D. J., Boyd, G., & Zhang, H. Q. (2011). UV penetration depth in liquid egg whiteand liquid whole egg. Journal of Food Processing and Preservation, 35, 754–757.

Geveke, D. J., & Torres, D. (2012). Pasteurization of grapefruit juice using a centrifugal ultra-violet light irradiator. Journal of Food Engineering, 111(2), 241–246.

Geveke, D. J., & Torres, D. (2013). Liquid egg white pasteurization using a centrifugal UVirradiator. International Journal of Food Microbiology, 162(1), 43–47.

Global Industry Analysts, Inc. (2011). Global eggs consumption to reach 1,154 billionunits by 2015. PRWeb Online Visibility from Vocus 05/12/2011 (Available in:http://www.prweb.com/releases/eggs_conventional_cage/free_range_eggs/prweb3547074.htm Accessed on: December 5th, 2011)

Gonzalez, L., Geeraerd, A. H., Elst, K., Van Ginnekena, L., Van Impe, J. F., & Devlieghere, F.(2009). Inactivation of naturally occurring microorganisms in liquid whole eggusing high pressure carbon dioxide processing as an alternative to heat pasteurization.Journal of Supercritical Fluids, 51, 74–82.

Jones, D. R., Anderson, K., & Guard, J. Y. (2012). Microbiological differences between layinghen strains housed in various production systems. Poultry Science, 91, 27.

98 P.M. de Souza et al. / Innovative Food Science and Emerging Technologies 21 (2014) 90–98

Keyser, M., Müller, I. A., Cilleirs, F. P., Nel, W., & Gouws, P. A. (2008). Ultraviolet radiationas a non-thermal treatment for the inactivation of microorganisms in fruit juice.Innovative Food Science and Emerging Technologies, 9, 348–354.

Koutchma, T. (2009). Advances in ultraviolet light technology for non-thermal processingof liquid foods. Food Bioprocess Technology, 2, 138–155.

Koutchma, T., Keller, S., Chirtel, S., & Parisi, B. (2004). Ultraviolet disinfection of juiceproducts in laminar and turbulent flow reactors. Innovative Food Science andEmerging Technologies, 5(2), 179–189.

Martelli, F., & Davies, R. H. (2012). Salmonella serovars isolated from table eggs: An overview.Food Research International, 45(2), 745–754.

Müller, A., Stahl, M. R., Graef, V., Franz, C. M.A. P., & Huch, M. (2011). UV-C treatment ofjuices to inactivate microorganisms using dean vortex technology. Journal of FoodEngineering, 107, 268–275.

Murakami, E. G., Jackson, L., Madsen, K., & Schickedanz, B. (2006). Factors affecting the ul-traviolet inactivation of Escherichia coli K12 in apple juice and a model system. Journalof Food Process Engineering, 29(1), 53–71.

Musgrove, M. T., Jones, D. R., Northcutt, J. K., Harrison, M.A., & Cox, N. A., Jr. (2005). Impactof commercial processing on the microbiological safety and quality of shell eggs.Journal of Food Protection, 68(9), 2367–2375.

Ngadi, M., Smith, J. P., & Cayouette, B. (2003). Kinetics of ultraviolet light inactivation ofEscherichia coli O157:H7 in liquid foods. Journal of the Science of Food and Agriculture,83, 1551–1555.

Oteiza, J. M., Giannuzzi, L., & Zaritzky, N. (2010). Ultraviolet treatment of orange juice to in-activate E. coli O157:H7 as affected by native microflora. Food Bioprocess Technology, 3,603–614.

Peleg, M. (2000). Microbial survival curves. The reality of flat shoulders and absolute ther-mal death times. Food Research International, 33, 531.

Poggel, M., Wübben, D. I., Brod, H., Jenne, M., & Schmidt, S. (2008). UVivatec® — Anovel scalable reactor system for pathogenic virus inactivation by UV-C irradi-ation. Viral safety for biopharmaceuticals conference, Prague, Czech Republic, 23–24 June.

Schmidt, S., & Kauling, J. (2007). Process and laboratory scale UV inactivation of virusesand bacteria using an innovative coiled tube reactor. Chemical Engineering andTechnology, 30(7), 645–950.

Shebolina, E. S., Sullivan, S. A., O'Neill, K. R., Nevin, K. P., & Lovley, D. R. (2004). Isolation,characterization, and U(VI)-reduction potential of a facultatively, anaerobic,acid-resistant bacterium from low-pH, nitrate- and U(VI)-contaminated subsurface sed-iment and description of Salmonella subterranea sp. nov. Applied and EnvironmentalMicrobiology, 70, 5.

Souza, P.M., & Fernandez, A. (2011). Effects of UV-Conphysicochemical quality attributes andSalmonella enteritidis inactivation in liquid egg products. Food Control, 22, 1385–1392.

Souza, P.M., & Fernandez, A. (2012). Consumer acceptance of UV-C treated liquid eggproducts and preparations with UV-C treated eggs. Innovative Food Science andEmerging Technologies, 14, 107–114.

Ünlütürk, S., Atilgan,M. R., Baysal,H., & Tari, C. (2007). Use of UV-C radiation as non-thermalprocess for liquid egg products (LEP). Journal of Food Engineering, 85, 561–568.

Ünlütürk, S., Atılgan, M. R., Baysal, A. H., & Ünlütürk, M. S. (2010). Modeling inactivationkinetics of liquid egg white exposed to UV-C irradiation. International Journal of FoodMicrobiology, 341–347.

USDA U.S. Department of Agriculture (1969). Egg pasteurization manual. ARS 74-48.Washington, D.C. Agricultural Research Service.

Wright, J. R., Summer, S. S., Hackey, C. R., Pierson, M.D., & Zoecklein, B. W. (2000). Efficacyof ultraviolet light for reducing Escherichia coli O157:H7 in unpasteurized apple cider.Journal of Food Protection, 63(5), 563–567.