ReviewAdvancesandchallenges for the useofengineerednanoparticlesinfoodcontactmaterialsJosephC.Hannona,JosephKerryb,MalcoCruz-Romerob,MichaelMorriscandEndaCumminsa,*aBiosystemsEngineering,SchoolofAgriculture,FoodScienceandVeterinaryMedicine,AgricultureandFoodScienceCentre,UniversityCollegeDublin,Beleld,Dublin4,Ireland(Tel.: D35317167476;e-mail:enda.cummins@ucd.ie)bSchoolofFood&NutritionalSciences,FoodPackagingGroup,UniversityCollegeCork,Cork,IrelandcDepartmentofChemistry,UniversityCollegeCork,Cork,IrelandTheuseofnanotechnologyinthefoodindustryhasgreatpo-tential,particularlyintheareaoffoodpackaging.Thispaperlooksat recent advancesandindustrychallengesinrelationtotheuseof metal andnon-metal engineerednanoparticles(ENPs)infoodpackagingtograntactiveandintelligentprop-erties. A particular focus will be placed on risk assessment stra-tegies andpolicy developments associatedwiththe use ofnanotechnology in food contact materials (FCMs). Theabsenceof aregulatoryframeworkfor NPFCMs has beenhighlighted as a drawback for the development of nanoparticleFCMs. To aid the understanding of nanotechnology in the areaof FCMs, a NP specic exposure framework providing promptriskassessment couldbeinvaluabletoindustry, consumersandregulatorybodies.IntroductionCurrently the worlds population stands at 6.47 billion,however this is expected to increase to 9.08 billion by2050 (WPO, 2008). This creates a number of complexproblems, particularly the issue of an adequate food supply.Theworldsfoodresourcesareunevenlydistributedglob-ally,resultinginthedifculttaskofpreservingfoodstuffstoallowfor transportationtoawider geographical area.Foodpackagingisacommonmethodof preservingfoodstuffs,combinedwithpreservatives,temperatureandpres-sure treatments. The advent of materials containingNPsinthe sizerange1e100nm, granting improvedproperties,has provenadvantageous ina vast number of industriessuchasthecosmetics,foodandbeverage,textile,medical,electronicsandcomputing,appliancesandcookingutensilindustries(Maynard&Michelson, 2014). Emergingfoodpackagingmaterials containingENPs that possess activeand intelligent properties have the potential to alleviatesome of the global food supply issues. These materialsmay increase the shelf life of food products, improvefoodsafetyandreducetheamount of foodwasteduetospoilage. However, the uptake of novel food packaging ma-terials containing ENPs has been met with concerns in rela-tion to the risk posed to humans from consumption of ENPswhich may migrate fromNPfood packaging into food(Kanmani &Rhim, 2014b). This issueisexacerbatedbytheimmenseuncertaintywhichsurroundstheeldofNPhumanoral toxicity. Advancementsintheareaofinvivomousetoxicity(Park, Bae, et al., 2010; Park, Marsh, &Dawson, 2010) and in vitro human cell studies (Loh,Saunders, &Lim, 2012)presentingorgandamageandin-ammatoryresponses inmice, andextensive damage tointracellularorganellesincellshavebeenchallengedbyarecent invivohumantoxicitystudyshowingnoclinicallysignicant effects of engineered silver nanoparticles(EAgNPs) under acute oral dose conditions (Mungeret al., 2014). EAgNPs are silver NPs which exist as a resultof some size reducing process, whether intentional or unin-tentional.Additionally,contradictionsexistconcerningthetoxicity of NPs, mainly surrounding the argument that * Correspondingauthor.http://dx.doi.org/10.1016/j.tifs.2015.01.0080924-2244/2015ElsevierLtd.Allrightsreserved.TrendsinFoodScience&Technology43(2015)43e62humansare, andhavebeenexposedtoquantitiesofnatu-rally occurring nanoparticles (NONPs) in food and theenvironment that wouldbeconsideredharmful under thepresent conservative regulations (Sk, Jaiswal, Paul,Ghosh,&Chattopadhyay,2012).Anoteworthydistinctionis the disparity between ENPs which are intentionally man-ufactured to possess enhanced properties and NONPswhich arenaturally occurringand unintentional. Currently,in the literature studies focussing on the presence ofNONPsinfoodhavebeenlimitedtoaselect number offoods.AstudybySketal.(2012)conrmedthepresenceofcarbonNONPsinfoodproductssuchasbread,jaggery,corn akes and biscuits. In their conclusions they noted thatNPs existed innaturelong beforeanalytical techniquesfordetection of NPs were developed. Similarly, a study byYangetal. (2014)foundthat foodgradetitaniumdioxide(E171) contained between 17 and 35% nanosized particles.Another food additive which has been found to contain ag-gregates withparticles S.aureus. ContradictorystudiescanbefoundwhichpresentmildinhibitionofS. aureusincomparisontothesatisfac-toryinhibitionofE. coliusingEAgNPsofmeandiameter13.5nminanaqueoussolution(Kimet al., 2007). WhencomparingantimicrobialstudiesitisimportanttoconsiderthemediainwhichtheENPsarerestrained.Manyantimi-crobial studies deal with EAgNPs mobilized in aqueous so-lutions. The antimicrobial activityof ENPs containedinsuchastatecouldbeconsideredapoor representationofENPsimmobilizedinfoodpackagingastheionisationpo-tentialofENPsisincreasedinaliquidmedium.Thusitisimportant that antimicrobial studies beavailablefor NPsincorporatedinfoodpackaging.Despitetheconsiderableresearchfocusonnanocompo-sitescontainingEAgNPstoprovideantimicrobialactivity,there has been instances of non-metal, metal-oxide andmetal-hybrids being used to provide antimicrobial function.In particular, nanoclays such as MMTwhich would beconventionallyexploitedfor barrier properties havebeenincorporatedinfoodpackagingasanantimicrobial. Rhimet al. (2006) observed greatly improved antimicrobialactivityfor four different chitosan-basednanocomposites50 J.C.Hannonetal./TrendsinFoodScience&Technology43(2015)43e62containingunmodiedMMT, organicallymodiedMMT,Ag and Ag-zeolite ENPs. In later work, particular attentionwas paid to two chitosan/organoclay nanocomposites(Cloisite30Band Cloisite20A) which increased gasbarrier properties in linear low density polyethylene(LLDPE)andgaveantimicrobial propertiesagainst gram-positive bacterium(Hong &Rhim, 2008, 2012). OtherENPs which have been found to be effective antimicrobialsagainst food pathogens include ZnO (Akbar & Anal, 2014)andTiO2(Bodaghietal.,2013).Examples of materials containingnatural antibacterialproperties can also be found, which include protectivenanostructured coatings on animals and insects. It hasbeenshownthatinsectscoatthemselveswithantibacterialsubstances to forma protective coating frompredatorsandtheenvironment. Thesurfaceof Cicadainsect wingspossess natural bactericidal characteristics attributed totheir nanopillar surface coating (Hasan, Crawford, &Ivanova, 2013). Anumber of naturallyoccurringantioxi-dants have been studied for use in food packaging applica-tionssuchasa-tocepherol, plantextractsandessential oilextracts fromherbs and spices (Woranuch &Yoksan,2013).For NPs to be adopted in industry for use in packaging toprovideantimicrobialpropertiesthebenetsmustbeclearandsubstantial.Asignicantlyincreasedantimicrobialac-tivity must be observed for materials containing NPs whichcan compete with alternative antimicrobial materials whichare already in use such as chitosan. Chitosan has beenshowntobeaffectiveagainst awiderangeofbacteria, isbiodegradableanddoesnothavethesameregulatorybar-riers as nanocomposites due to its non-toxicity (Aider,2010).Themanufactureofnanocompositessuchasbiode-gradable food packaging (e.g. chitosan) containingEAgNPs(Kanmani &Rhim, 2014a)providingsynergisticimprovements in terms of mechanical and antimicrobial ac-tivitycouldprovide analternativetoconventional pack-aging. More attention is required for applications wereNPcombinationsinpackagingcanpresentsynergisticim-provements andthus tackle some of theworlds currentpackagingproblems(Busoloetal.,2010).RiskassessmentstrategiesfornanoparticlesinfoodpackagingENPshavethepotential tocauseharmtohumansandtheenvironment throughincreasedtoxicity, mobilityandbioaccumulation. For nanotechnology to be accepted byconsumers, all associated risks should be clearly communi-catedinsucha waythatconsumerscanmakeaninformeddecision. Furthermore, thelevel of riskposedtohumansshouldbeinvestigatedundertheworst caseconditionsofexposure. If anunacceptablelevel of riskispresented, ariskmanagementstrategyshouldbedevelopedtomitigatethe risk. Risk assessment is a methodology commonlyusedtoassesstheriskposedtohumansandtheenviron-ment fromexposure to a substance or process. Whenapplied to ENP food packaging, the level of exposure to hu-mans fromingestionof NPs is determinedviamigrationstudiesandinvivotoxicitystudies. Ifanacceptablelevelofriskisobserveditisthentheresponsibilityofthegov-erning authority to allow or disallow the use of the product.A recent success of this process was the acceptance on TiNENPsbytheEuropeanFoodSafetyAuthority(EFSA)foruse in PET bottles in concentrations up to 20 mg/kg(EFSA, 2012). Thefunctionof TiNENPs is toimprovethe oxygen barrier properties of the walls of the PETcontainer that they are incorporated into. Excellent contain-ment of the TiN NPs within the walls of the PET containersmay be a unique aspect of the migration mechanism whichis not sharedbyother NPs whichrequiresomelevel ofmigration to carry out their function. NPs such as EAgNPsmust migrate in the form of Ag ions to allow for their anti-microbial function, while complying with the migrationlimitssetoutbytheEuropeanCommission(EFSA, 2008;European Commission, 2011). Therefore a compromisemust bemadebetweenthelevelofmigrationandantimi-crobialactivity.ExposureassessmentmodelsFrequently, substances that are consideredharmful tohumansmaynot exist inhighenoughdosestoposeanyreal risk to humans. Mathematical exposure models provideamethodfor quantifyingtheriskposedtohumans fromNPs. Usingthe results fromNPmigrationstudies as aninput toanexposuremodel, theassociatedriskfromNPscanbepredictedbasedonthescenariosurroundingtheiruse. Two common scenarios are often identied for humanexposuretoENPs, theworst casescenario(wcs) andthemost likelyscenario(mls) (Cushen, Kerry, Morris, Cruz-Romero, & Cummins, 2013). The mls is an exposure valuebasedonthemostprobableintakeofasubstanceobtainedfrommigrationstudiesandsurveydata. ThewcsinvolvesthegreatestexposuretoENPspossible,basedonexagger-ated migration and consumption data. To produce a humanexposuremodelsuitabletoxicitystudies,migrationstudiesandconsumer datamust beavailable. Giventhelackofin vivo toxicity studies for exposure to ENPs, it is necessarytoadaptnon-nanorodentoraltoxicitystudiesandapplyasafety factor. Only four human exposure models arecurrentlyavailableintheliteraturethat quantiestheriskposedtohumansfromoral exposuretoENPswhichhavemigrated fromfood packaging (Bachler, von Goetz, &Hungerbuhler, 2013; Cushen et al., 2013, 2014a; vonGoetzet al., 2013; Smirnovaet al., 2012). Ineachstudyamathematicalmodelisgeneratedtopredictthepotentialmigrationandresultingmigrant amountsarecomparedtoactual migrationresults. Theresultingmigrant quantitiesarethencoupledwithconsumerdatatogenerateamodelthatpredictstheriskposedtohumansfromoralexposure.von Goetz et al. (2013) observed worst case acute exposureof 4.2 mg EAgNPs caused by storage of 100 ml of food sim-ulantinanEAgNPfoodcontainer.Althoughthiscouldbe51 J.C.Hannonetal./TrendsinFoodScience&Technology43(2015)43e62consideredalargequantityof migratingparticles it wasnotedthat other potential sourcesof EAgNPsarepresentinnaturewhichcancontaincomparablylargequantities,suchasdrinkingwater(Akaigheetal.,2011).Weaknessesinthemodelwereattributedtotheuncertaintiessurround-ing the toxicological effects of EAgNPs and the possibilityof a Trojan horse mechanism (Kreuter, 2004). In themodel nolinkwas made tofoodconsumptiondata andalternativelyit was statedthat a givenamount of liquidfoodwouldcauseworst caseacuteexposurefor humans.Aparticular strength of the exposure model formulatedbyCushenetal. (2013)wastheuseofchickenconsump-tiondatafromanIrishsurveytopredict theexposuretoanindividual consumingtheaveragequantityof chickenper day. In addition, when determining the toxicity ofENPs to humans the surface area of the dosage was consid-eredalongsidethe weightof thedose whencalculatingtheProvisionalIngestionLimit(PIL)whichwasadaptedfromOBrienandCummins(2010).Itwasfoundthattheworstcaseconditions wouldcausemigrationof 8.85mg/kgofEAgNPs, considerably lower than the conservative60mg/kgoverallmigrationlimitallowedbytheEuropeanUnion(EuropeanCommission(EU),2002).Thisvaluefarexceedsthespecicmigrationlimitof0.01mg/kgforun-authorizedsubstances outlinedinDirective10/2011/EEC(European Commission (EU), 2011). Migration limitshavebeenset bytheEuropeanCommissionfor productsthatareusedinapplicationsinvolvingparticularlysuscep-tible persons, such as infants (Commission Regulation(EU), 2009). Giventhenumber of applicationsfor nano-compositesintheinfantfoodstorageindustryitissurpris-ingthatfewstudieshavespecializedinthearea.InarecentstudybyBachleretal.(2013)aphysiologi-cally based pharmacokinetic model was generated for ionicandNPsilver for veexposurescenarios. Oral exposurewas quantied for EAgNPs fromtwo sources; dietaryintakeandfromfoodwhichhasbeenstoredinENPfoodstorageboxes. Thepharmacokineticmodel wasvalidatedbycomparingsimulatedorganconcentrationstothoseob-tainedfrominvivoexperimental studies. It was demon-stratedthat for EAgNPssizeandcoatingdidnot showasignicant effect onbiodistribution. Furthermore, invivostudies suggested that EAgNPs are more likely to be storedas insoluble salt particles than dissolve into silver ions.Interestingly, in all exposure scenarios the Ag levels inmost organswerebeloworaroundthebackgroundlevelsofdietaryintakeandlowerthanlevelswhichwouldcauseadverse effectsin vitro. The resultsindicatethatoutsideofanoccupational setting, the level of risktoadults fromexposuretonanosilverconsumerproductsislow.OBrienandCummins(2011) presentedariskassess-mentframeworkonthreenanomaterials;EAgNPs, ceriumoxide NPs and TiO2 NPs which have the potential to accu-mulate in surface and waste water in the environment(OBrien&Cummins, 2011). Theframeworkutilizedun-certainty and variability principles, alongside qualitativerisk assessment principles to generate a ranking systemfor metallic NPconcentration, transport and persistencein aquatic environments. Nanomaterial characteristics aswell as aquatic environmental characteristics werecompiledtorankriskof exposureunder threescenarios.The study highlights were data critical to NPexposureare lacking and suggests research needs in order to populatethequalitativeframeworkwithquantitativeexposuredata.Although, the risk assessment of human exposure toENPs was not featuredinthestudy, thestructureof theframeworkwasveryapplicabletofoodpackagingriskas-sessments. SuchaqualitativeriskrankingframeworkforhumanexposuretonanoFCMswouldgiveapreliminaryindicationofhumanexposureandhelpprioritisequantita-tiveexposureassessmentstopopulateananofoodpack-agingexposuremodel.TheNanoReleaseFoodAdditivesExpertGroup(Nano-ReleaseFoodAdditivesExpert Group, 2015)hasfocusedon the uptake of ENPs in the alimentary tract (Alger,Momcilovic, Carlander, & Duncan,2014),characterisationmethodsandrelatedriskmanagement aspects. Theexpertgrouphavepublishedanumber of papers, particularlyintheareaof characterisationof NPs releasedfromFCMs(Noonan, Whelton, Carlander, &Duncan, 2014) andtheuseofgastrointestinalmodelstoassessthedigestionandabsorptionof ENPs releasefromFCMs (Lefebvreet al.,2014). Aparticular strengthof thereviewcarriedout byLefebvre et al. (2014) is the presence of an exampleapproachfor theassessment oftheuptakeofNMsinthehumangastrointestinal tract (GIT). Themodel accountsforInvivoanimal models, Exvivotissuemodels, Invitrocellculturemodels, Invitronon-cellularuidmodelsandIn silico computational models. Each of these elementscanbeusedtostrengthenanoverall methodologyfortheassessment ofNPreleasefromFCMsandsubsequent hu-manexposure.Acomprehensivehumanexposure frameworkfor a-vours, additives and FCMs has recently been published un-der the FACET project (European Commission (EU),2012). TheframeworkcombinesEuropeanconsumersur-veyswithtoxicitystudiestoallowfortheriskassessmentof existingandemergingmaterials. Amajor downfall ofthe framework is the exclusion of nanomaterials in thelist of contaminants. However, the project has the potentialto be used as a methodology for the human risk assessmentof nanomaterials. There is a growing need for a frameworkdealingwithhumanexposuretoNPsinFCMs.Barrierstosuch a framework being established include; gaps in knowl-edge related to ENP migration, ENP physicochemical prop-erties,humantoxicityandthefateofENPsintheGIT.FateofnanoparticlesintheGITIn the human body there are three main routes of exposurefrom ENPs, these are dermal contact, inhalation and inges-tion. Other uncommon routes of exposure which haverecently become applicable, due to emerging ENP medicines52 J.C.Hannonetal./TrendsinFoodScience&Technology43(2015)43e62and hygiene products, are through rectal administration,through the female genital tract and by direct administrationintothebloodbyinjection(Chen&Schluesener, 2008).There are numerous scenarios in which humans can beexposedtoNPsthroughany ofaforementionedroutes.Forexample, astudyonthelevels ofvinyl chloridefromPVCinadomesticwater supplyfoundthat thevinyl chloridecouldbeingestedandalsoinhaledfromshowerwaterdueto the formation of aerosols (Lee et al., 2002). In this reviewthe principal focus is on ENP exposure via the oral route ofexposure. It should be noted that although the possible risk tohumansfromoralexposuretoENPsaregreat,studiesthatfocus on oral exposure and the fate of ENPs in the GIT arelimited (Silvestre et al., 2011). At present there are noin vivo studies related to the toxicity of ENPs to the humanbody through the ingestion route. Therefore, the toxicity ofENPs in the GIT has been investigated by means of in vivostudies of rodents (Kimet al., 2008; Park, Bae, et al.,2010; Park, Marsh, et al., 2010), in vitro studies on represen-tative human GIT cells (Aueviriyavit, Phummiratch, &Maniratanachote,2014)and invitrostudiesofENPswhenexposed to a synthetic human stomach (Rogers et al.,2012). Each independent study has the potential to contributeto abroaderinvestigation intothe toxicity of ENPsfor hu-mans. However, few studies have linked the numerous frag-mented studies to build a general human exposure model forENPs in food packaging materials.TheambiguousnatureofENPfateintheGIThasnotaidedtheacceptanceofENPsinfoodpackagingapplica-tions. Recent studies presenting data on important GITmechanisms have shown similarities concerning ENPbehaviour.Rogersetal.(2012)carriedoutinvitrostudiesontheexposureofEAgNPstoasynthetichumanstomachandtheeffects of synthetichumanstomachuidontheagglomerationof NPs(Rogerset al., 2012). Followinga1hexposureperiodit wasnotedthat EAgNPsagglomer-atedandreactedwiththesyntheticstomachuidtoformsilver chloride(AgCl). It was notedthat theresultsmaynot havebeenrepresentativeof ahumanstomachsexpo-sure to NPs due to theeffectsof some of the coating com-pounds used during the preparation of the EAgNPs.Similarly, Mwilu et al.(2013)carriedoutan in vitro studyonthe inuence of synthetic stomachuidonEAgNPswithdifferentsizesandcappingagents.Signicantaggre-gationwasnoticed, particularlyinrelationtothesmallerENPs (