Retention of Antibacterial Activity in Geranium Plasma Polymer Thin Films

23 Pages·2017·3.93 MB·English

nanomaterials
Article
Retention of Antibacterial Activity in Geranium
Plasma Polymer Thin Films
AhmedAl-Jumaili1,KaterynaBazaka1,2andMohanV.Jacob1,* ID
1 ElectronicsMaterialsLab,CollegeofScienceandEngineering,JamesCookUniversity,Townsville,
QLD4811,Australia;[email protected](A.A.-J.);[email protected](K.B.)
2 SchoolofChemistry,Physics,MechanicalEngineering,QueenslandUniversityofTechnology,
Brisbane,QLD4000,Australia
* Correspondence:[email protected];Tel.:[phone]
Received:3August2017;Accepted:5September2017;Published:13September2017
Abstract: Bacterial colonisation of biomedical devices demands novel antibacterial coatings.
Plasma-enabledtreatmentisanestablishedtechniqueforselectivemodificationofphysicochemical
characteristicsofthesurfaceanddepositionofpolymerthinfilms. Weinvestigatedtheretentionof
inherentantibacterialactivityingeraniumbasedplasmapolymerthinfilms. Attachmentandbiofilm
formation by Staphylococcus aureus, Pseudomonas aeruginosa, and Escherichia coli was significantly
reducedonthesurfacesofsamplesfabricatedat10Wradiofrequency(RF)power,comparedtothat
ofcontrolorfilmsfabricatedathigherinputpower. Thiswasattributedtolowercontactangleand
retentionoforiginalchemicalfunctionalityinthepolymerfilmsfabricatedunderlowinputpower
conditions. The topography of all surfaces was uniform and smooth, with surface roughness of
0.18and0.69nmforfilmsfabricatedat10Wand100W,respectively. Hardnessandelasticmodules
offilmsincreasedwithinputpower. Independentofinputpower,filmswereopticallytransparent
withinthevisiblewavelengthrange,withthemainabsorptionat~290nmandopticalbandgapof
~3.6eV.Theseresultssuggestthatgeraniumextract-derivedpolymersmaypotentiallybeusedas
antibacterialcoatingsforcontactlenses.
Keywords: antibacterial coatings; essential oils; geranium oil-derived polymer; plasma
polymerisation
1. Introduction
Medical devices are a critical part of the current healthcare system. However, their usage
has led to the emergence of device-associated infections. Contamination of medical devices by
microorganismsisassociatedwithsubstantialmorbidity,aswellassubstantialhealthcarecosts[1,2].
Abiotic surfaces are vulnerable to microbial attachment and growth, and eventually, biofilm
formation; as such, these surfaces may act as a reservoir of chronic infection [3]. Indeed, 64% of
hospital-acquired infections worldwide are attributed to attachment of viable bacteria to medical
devices and implants [4], and a reported 80% of the global impact of surgical site infections
involve microbial biofilms [5]. When protected by the biofilm, bacterial cells are significantly less
susceptible to antibiotics and host immune responses than planktonic bacteria of the same strain,
and as such, biofilms are more difficult to clear [6]. Emergence of microbial strains resistant to
one or more of the available synthetic antibiotics creates further challenges for the treatment of
implant-associatedinfections. Significantexamplesincludemethicillin-resistantStaphylococcusaureus,
NDM-1producingKlebsiellapneumoniae,vancomycin-resistantEnterococcus,andmultidrug-resistant
Mycobacteriumtuberculosis[7].
Tosubstantiallyalleviatepathogenicinfections, thedevelopmentofeffectiveself-disinfecting
surfacecoatingsformedicaldevicesisvital. Severalsurfaceshavebeenadaptedtoinhibitand/or
Nanomaterials2017,7,270;doi:10.3390/nano7090270 www.mdpi.com/journal/nanomaterials
Nanomaterials2017,7,270 2of23
reducemicrobialadhesionandproliferationviaantibiofoulingand/orbactericidalactivity,depending
on the effect the surface exerts on the microorganism [8]. Antibiofouling materials may resist the
initial attachment of microorganisms, due to the existence of unfavourable surface micro-features
and/orsurfacechemistry. Ontheotherhand,bactericidalsurfacesmayimpedecolonisationbykilling
themicroorganismsoncontact,viasurface-immobilisedantimicrobialmacromolecules,orbyreleasing
biocidalmolecules,e.g.,polymericbiocides[9]. However,self-disinfectingsurfacesoftensufferfrom
significantdrawbacks,suchasuncontrolledmaterialdegradation,prematuremechanicalfailure,and
limitedbiocompatibility[10].
Inrecentyears,therehasbeenanincreasinginterestintheuseofplant-derivedcompoundsas
naturalantimicrobials[11,12]. Essentialoilsandplantextractsarerichsourcesofbiologically-active
compoundswithmechanismsthataredistinctfromthoseofcurrentlyusedsyntheticantibiotics,which
shouldlimittheemergenceofbacterialresistance[13]. Amongthesenaturalantimicrobialagents,
geranium(Pelargoniumgraveolens)essentialoilexhibitsstrongactivityagainstabroadspectrumof
bacterialstrains[14,15]. Whenusedinsolutionorasanaerosol,geraniumoilwaseffectiveagainst
clinically-significanthumanpathogens,suchasgram-positiveS.aureusandEnterococcusfaecalis,and
gram-negativeP.aeruginosa,Proteusmirabilis,andEscherichiacoli,andthefungiCandidaalbicans[16,17].
However, until recently, essential oils have rarely been used for fabrication of biologically-active
coatings,duetoacomplex,multi-componentnatureofessentialoilsandextracts,whichmayvary
with season and geographic location [18]. Recent advances in cold plasma polymerisation have
enabledtheconversionofessentialoilsandtheirderivativesintothin,highly-adherentanddefect-free
optically-transparentcoatings,whosebiologicalactivityanddegradationkineticscanbecontrolledby
controllingthechemicalstructureoftheprecursorandtheplasmaprocessingconditions[19–21].
Thispaperinvestigatesthesynthesis,materialproperties,andantibacterialactivityofgeranium
oil-basedpolymersfabricatedusingplasmapolymerisation. Toourknowledge,thisisthefirstreport
ontheuseofpolymersderivedfromgeraniumoilasanantibacterialcoating.
2. Methods
2.1. Materials
Aprecursormaterialwasselectedbecauseofitsstrongantibacterialactivityagainstgram-negative
and gram-positive bacteria, and compatibility with the plasma polymerisation process [14,22].
The geranium essential oil was procured from Australian Botanical Products (ABP, Victoria,
Australia), and was used without further modification. According to the manufacturer, the main
compounds include citronellol (32%), geraniol (15%), linalool (6%), isomenthone (6%), geranyl
formate (2.5%), tiglate (2%), citronellyl formate (6%), guaia-6,9-diene, and 10-epi-γ eudesmol
(5%). Citronellol(C H O)andgeraniol(C H )arearomaticacyclicmonoterpenealcohols,and
10 20 10 18
are considered to be very potent bactericides responsible for the biological action of geranium
oil[23–25]. Fromaprocessingpointofview,geraniumoiliscompatiblewithplasmapolymerisation
(plasma-enhancedchemicalvapourdeposition),asthisoilishighlyvolatileatroomtemperature,and
sonoexternalheatingnorcarriergasisrequiredtodelivertheprecursorunitstothesiteofdeposition.
Geraniumoilhasadensityof1.044g/mLat25◦C,aboilingpointof250–258◦C,andrefractiveindex
of1.53at20◦C.
2.2. PolymerSynthesis
Microscope glass slides (76 mm × 26 mm) and cover glass No.5 (19 mm) were cleaned with
commercialdecon,washedextensivelywithdistilledwater,thensonicatedindistilledwaterfor20min.
Thereafter,substrateswererinsedwithacetone,subjectedtopropan-2-olbathfor15min,andfinally
driedbyair. Plasmasynthesiswascarriedoutinacustom-madecylindricalglasschamber(l: 90cm,
d:5cm). Thereactorchamberwasevacuatedtopressureof0.2mbarusingadoublestagerotarypump
(JVAC–DD150,Victoria,Australia). Radiofrequency(RF)generatormodelACG-3B(MKSInstruments,
Nanomaterials2017,7,270 3of23
Andover,MA,USA)wasruninthecontinuousmodeat13.56MHztoprovideinputpowerthrough
a matching network. Two external parallel copper rings were utilised as electrodes, separated by
adistanceof9cm.Substrateswereplacedatequaldistancefrombothelectrodes.Thedistancebetween
thesubstrateandthemonomerinletwas20cm. Depositiontimesrangedfrom30minto120min,
yieldingfilmswiththicknessesrangingfrom~450nmto1500nm. Ineachexperiment,themonomer
vialwasloadedwith0.5g(~12drops)ofgeraniumoil. Themonomerflowrate(F)wascalculated
usingtherelation1,derivedfromtheidealgasequation[26]:
dp V
F = ×16172 (1)
dt T
wherepisthepressureinsidethechamber(mbar),andtistime(s),Visthevolumeofthechamber(L),
andTistheambienttemperature(K).
Initially, the chamber was evacuated to 0.2 mbar, then the precursor gas was released into
thechamberuntilthepressurereachedastablevalue,atwhichpointtheoutletvalvewasclosed,and
thepressurewasmeasuredevery5sfor1min. Itwasestimatedthatthegeraniumflowrateduring
thepolymerisationprocesswas16.22cm3/min.
3. PolymerCharacterisation
3.1. ChemicalProperties
PerkinElmerSpectrum100Fouriertransforminfrared(FTIR)spectrometer(PerkinelmerInc.,
Boston, MA,USA)wasemployedinthetransmissionmodetoidentifycomponentsandchemical
properties of fabricated films. Films were deposited on potassium bromide pellets (KBr), and
thespectrawereacquiredfrom4000to500cm−1withresolutionof4cm−1averagedacross32scans.
3.2. OpticalProperties
Optical constants and film thickness were identified using Variable Angle Spectroscopic
Ellipsometry (JA Woollam-M2000 D, Lincoln, NE, USA). Ellipsometry measures a change in
polarisation as light is reflected from the surface of a film. The polarisation change is represented
as an amplitude ratio (Ψ), and the phase difference (∆). The ellipsometric parameters Ψ and ∆
were acquired at three different angles (55◦, 60◦ and 65◦, in addition to transmission data) to
measuretherefractiveindex(n),extinctionco-efficient(k),andfilmthicknessatwavelengthrangeof
200–1000nm. Thesoftwarepackage(WVASE32,Lincoln,NE,USA)wasusedformodelling. First,to
estimate film thickness, a three-layer model consisting of a previously modelled substrate layer,
aCauchylayer(torepresentthefilm),andasurfaceroughnesslayerwereappliedtothedatawithin
the400–1000nmregionwherethefilmisopticallytransparent. Afterpoint-by-pointapproximation,
optical constants were obtained by converting the Cauchy layer to GenOsc layer and applying
Gaussianoscillatortoobtainthebestfittothedata. UV–visdatawerecollectedusinganAvantes
spectrophotometer(Avaspec-2048,Apeldoorn,TheNetherlands)fittedwithanAvalight-DHclight
source. Taucequationwasappliedtocalculatetheopticalbandgap.
3.3. SurfaceTopographyandMechanicalProperties
Surfacemorphologywasexaminedusingalow-noisescanningandhigh-resolutionatomicforce
microscopeAFM(NT-MDTNTEGRA,Moscow,Russian)withascanningareaof10µm×10µmand
3µm×3µm. AFMwasoperatedinthetappingmode,wherethecantileveroscillateddirectlyabove
thesurfacetoacquiredata. Thismodewaspreferable,asitdecreasedtheinelasticdeformationsof
theinvestigatedsurface,aswellasreducedtheeffectiveforcesappliedtothesample.Allmeasurements
weredoneunderambientconditions. SoftwareNova(Version1.0.26,Moscow,Russian)wasusedto
analysethedatawiththefittingcorrectionvalue(polynomialorderof4).
Nanomaterials2017,7,270 4of23
ABerkovichTriboscopeindenter(Hysitron,Minneapolis,MN,USA)wasinterfacedwiththeAFM
Triboheadforthedeterminationofmechanicalproperties. Berkovichindenterhasthegeometryof
athree-sidedpyramid(70.3◦ equivalentsemi-openingangle). Toobtainaccuratedata,thecantilever
Nsaennomsiattievriiatlys 2w01a7s, 7c, a2l7i0b r atedpriortomeasurementsusingfusedsilica,then,driftcorrectionwasap4 pofl i2e2d .
Typicalloadsrangedfrom300to1000µNwithafixedloadingtimeof3s,holdingtimeof3s,and
holding time of 3 s, and unloading time of 5 s. Figure 1 shows loading (unloading) versus indenter
unloadingtimeof5s. Figure1showsloading(unloading)versusindenterdisplacement,illustrating
displacement, illustrating the elastic/plastic response of geranium oil-derived films.
theelastic/plasticresponseofgeraniumoil-derivedfilms.
Maximum force
Holding
Unloading nt
e
m
)
N e
c
µ a
( pl
d Loading s
a di
o m
L S= dP/dh
u
m
xi
a
M

Final depth
Displacement (nm)

Figure1.Schematicofthenano-indentationtestforhardnessandthemodulusmeasurement.
Figure 1. Schematic of the nano-indentation test for hardness and the modulus measurement.
3.4. ContactAngle,SurfaceTension,andSolubility
3.4. Contact Angle, Surface Tension, and Solubility
Contactanglemeasurementswereusedtodeterminethehydrophobicityofgeraniumoil-derived
Contact angle measurements were used to determine the hydrophobicity of geranium oil-
thin films deposited on glass. Sessile drop contact angle was measured using goniometer
derived thin films deposited on glass. Sessile drop contact angle was measured using goniometer
(KSVCAM101,Helsinki,Finland)andthreeliquids,namelydistilledwater,diiodomethane(DIM)
(KSV CAM 101, Helsinki, Finland) and three liquids, namely distilled water, diiodomethane (DIM)
andglycerol. Young–Laplacefittingwasperformedtoestimatethecontactangle,asaminimumof
and glycerol. Young–Laplace fitting was performed to estimate the contact angle, as a minimum of
fivemeasurementspersample.
five measurements per sample.
Thecalculatedvaluesofthecontactanglewereusedtoestimatesurfacetension(alsoknownas
The calculated values of the contact angle were used to estimate surface tension (also known as
surfaceenergy)parametersandsolubilityofgeraniumoil-derivedpolymerfilmsemployingvanOss,
surface energy) parameters and solubility of geranium oil-derived polymer films employing van Oss,
Chaudhury,andGood(VCG)method[27]:
Chaudhury, and Good (VCG) method [27]:
(cid:113) (cid:113) (cid:113)
(cid:4666)1+(cid:1855)(cid:1867)((cid:1871)1(cid:2016)+(cid:4667)γc(cid:3013)o=sθ)2γ (cid:3495)Lγ=(cid:3020)(cid:3013)(cid:3024)2+γγ(cid:3013)(cid:3013)SL(cid:3024)W++(cid:3495)γγLL(cid:3020)(cid:2878)W++γ(cid:3013)(cid:2879)+γ+S(cid:3495)γ+(cid:3020)_γ+−Lγ+(cid:3013)(cid:2878) γ _S+γ+L (2) (2)
where θ is the contact angle, γL the surface tension of the liquid in contact with the solid (mJ/m2), γ(cid:3013)(cid:3013)(cid:3024)
thweh aepreolθari scothmepcoonnetnact toaf nthgele s,uγrLfatchee tseunrsfiaocne otfe nthseio lniqoufidth (emlJi/qmu2i)d, iγn(cid:3013)(cid:3024)co tnhtea catpwoliathr cthomespoolniden(mt oJf/ tmhe2 ),
suγrLLfWacteh eenaeproglya rofc othmep soonliedn (tmofJ/tmhe2),s uγr(cid:2878)f atcheet eelnescitornono-facthceepltiqoru ipdar(mamJ/(cid:3020)emter2 )o,fγ tSLhWe tshoelidap (omlaJ/rmco2)m, γp_o ntheen t
eolefctthroens-udrofancoer epnaerragmyeotfetrh oefs othlied s(moliJ(cid:3020)d/ m(m2)J,/γm+S2)t, hγe(cid:2878)e ltehcetr oenle-catcrcoenp-taocrceppatroarm peaterramofetthere soofl itdhe(m li(cid:3020)Jq/umid2 ),
(mγ_SJ/tmhe2),e laencdtr oγn(cid:2879)- dtohneo erlepcatrraomn-edteornoofr tphaersaomliedte(mr oJ/f mth(cid:3013)2e) ,liγq+Luitdh e(melJe/cmtr2o).n -acceptorparameteroftheliquid
(mJ/m2),and(cid:3013)γ− theelectron-donorparameteroftheliquid(mJ/m2).
The VCG mLethod requires a minimum of three liquids, an apolar liquid such as DIM (where γ+
The VCG method requires a minimum of three liquids, an apolar liquid such as DIM (where
= γ− = 0), as well as two polar liquids, like water and glycerol. The surface tension parameters of the
γ+=γ− =0),aswellastwopolarliquids,likewaterandglycerol. Thesurfacetensionparametersof
used liquids are commonly presented in the literature [28], and are summarised in Table 1.
theusedliquidsarecommonlypresentedintheliterature[28],andaresummarisedinTable1.
Table 1. Surface tension parameters for water, diiodomethane (DIM), and glycerol used in
experiments.
Surface Tension Parameters, mJ/m2
Solvent
γ γLW γAB γ+ γ−
Water 72.8 21.8 51.0 25.5 25.5
DIM 50.8 50.8 0.0 0.0 0.0
Glycerol 64.0 34.0 30.0 3.9 57.4
Once the surface tension values for the solid and liquid are determined, the interfacial tension
γ can be calculated from the individual surface tension parameters as:
(cid:3020)(cid:3013)
Nanomaterials2017,7,270 5of23
Table1.Surfacetensionparametersforwater,diiodomethane(DIM),andglycerolusedinexperiments.
SurfaceTensionParameters,mJ/m2
Solvent
γ γLW γAB γ+ γ−
Water 72.8 21.8 51.0 25.5 25.5
DIM 50.8 50.8 0.0 0.0 0.0
Glycerol 64.0 34.0 30.0 3.9 57.4
Oncethesurfacetensionvaluesforthesolidandliquidaredetermined,theinterfacialtension
γ canbecalculatedfromtheindividualsurfacetensionparametersas:
SL
(cid:18)(cid:113) (cid:113) (cid:19)2 (cid:18)(cid:113) (cid:113) (cid:113) (cid:113) (cid:19)
γ = γLW − γLW +2 γ+γ−+ γ+γ−− γ+γ−− γ−γ+ (3)
SL S L S S L L S L S L
Thesolubilityofthegeraniumoil-basedpolymerfilms(∆G)indifferentsolventsisthenestimated
fromthevaluesoftheinterfacialtensionas:
∆G = −2γ (4)
121 SL
3.5. BacterialStudies
3.5.1. CellCultures
Amongclinically-significantpathogenicbacteria,infectionsduetotheStaphylococci,particularly
gram-positiveS.aureus,andgram-negativeP.aeruginosaandE.coli,aremostfrequentlyassociatedwith
theuseofimplants[8,29]. Giventheirsignificance,inthisstudy,S.aureusCIP65.8,P.aeruginosaATCC
9025,andE.coliK12wereacquiredfromtheAmericanTypeCultureCollection(ATCC,Manassas,VA,
USA)andCultureCollectionoftheInstitutPasteur(CIP,Paris,France). Foreachexperiment,afresh
suspensionwaspreparedbyfirstrefreshingthefrozenstockcultureonnutrientagar(Oxoid),then
growingthemovernightin100mLofnutrientbrothat37◦C,whileshakingat100rpm. Cellswere
harvested at the logarithmic stage of growth, and their density adjusted to OD = 0.3 to ensure
600
uniformstartingculture. Ahaemocytometerwasusedtoquantifycellnumbersinsuspensionpriorto
seedingontopolymersurfaces.
3.5.2. Incubation
Sterilepolymer-coatedglassslidesanduncoatedglassslides(ascontrol)wereplacedinto24well
plates,andanaliquotof100µLofbacterialsuspensionwascarefullyplacedonthesamplesurface.
Sampleswerethenallowedtoincubateat37◦Cand5%CO for18h. Theexperimentwasrunin
2
triplicate. Afterincubation,themediawasaspirated,andtheunattachedcellswereremovedbygently
rinsingthesurfacesofthesampleswithcopiousamountsofdouble-distilledwater. Sampleswere
thenallowedtodryat22◦Cfor30minat55%humiditytomaintaincellsinasemi-hydratedstate
formicroscopy.
3.5.3. Visualisation
Theretainedbacteriawerevisualisedbyscanningelectronmicroscopy. Priortoimaging,samples
werecoatedwithathinlayerofgoldusingSputterCoater(LeicaEM-SCD005,Wetzlar,Germany).
High-resolutionimagesoftheattachedcellswereobtainedusingtheScanningElectronMicroscope
(JEOL7001F,Tokyo,Japan)at1000,5000,and20,000×magnifications.
Todifferentiateviablecellsamongtheattachedbacteriaandquantifytheamountofextracellular
polysaccharides produced by the attached cells, SYTO® 17 Red (Molecular Probes™/Invitrogen,
ThermoFisher,Carlsbad,MA,USA)andAlexaFluor® 488(MolecularProbes™/Invitrogen,Thermo
FisherScientific,Carlsbad,MA,USA)stainswereusedtostainthebacterialcellredandextracellular
polymericsubstances(EPS)green,respectively. Thedyeswereappliedfollowingtheprotocoloutlined
Nanomaterials2017,7,270 6of23
in[30]. Theimageswereobtainedusingaconfocalscanninglasermicroscope(CSLM)(NikonA1R
ConfocalMicroscope,NewYork,NY,USA).TheCSLMimageswereprocessedtoconstruct3Dimages,
andtoquantitativelydescribetheimagesintermsoftheoverallvolumeofthebiofilmperunitareaof
substrate(termed“biovolume”,andinclusiveofbothcellsandEPS),andtheaveragebiofilmthickness.
4. ResultsandDiscussion
4.1. PolymerSynthesis
The polymer thin films were successfully fabricated on substrates, including microscope
glass slides, cover glass, and KBr disks, at input powers of 10, 25, 50, 75, and 100 W.
Spectroscopicellipsometrywasusedtoestimatethepolymerthicknesses. Thethicknessofthepolymer
increasedlinearlywithtime,yielding~10.6nm/minat10W.Also,thethicknessofthepolymerwas
foundtobeincreasingwithanincreaseintheinputpower.
4.2. ChemicalProperties
Geranium essential oil is a complex, multi-component mixture of monoterpenic alcohols
(geraniol, citronellol, linalool, etc.), esters (citronellyl tiglate, citronellyl formate, geranyl tiglate,
etc.), monoterpenic ketones (isomenthone and menthone), monoterpenic aldehydes (geranial and
neral),sesquiterpenichydrocarbons(guaia-6,9-diene,etc.),sesquiterpenicalcohol(10-epi-γ-eudesmol),
and several aromatic and oxide components [31]. Citronellol and geraniol are considered the key
components,whilelinaloolispresentinamuchsmallerportion(formoredetailonoilcomposition
seeSection2: Materials).
Principally, the organic molecules are held together via covalent bonds, which are relatively
strong, while creating a solid material via fairly weak van der Waals forces [32]. During plasma
polymerisation,plasma-generatedelectronsgainsufficientenergytobreakchainsoftheprecursor
molecules,creatingarichassortmentofhighlyreactivechemicalspecies. Thedegreeoffragmentation
isdirectlyrelatedtotheamountofappliedenergy. Thus-createdprecursorfragmentsthenundergo
recombinationinthegasphaseandonthesurfaceofthesubstrate,givingrisetoahighlycrosslinked
polymer,withastructurethatisirregularandunlikethatofaconventionalpolymer.
Figure2showstheFTIRspectraobtainedforgeraniumoil(precursor)andgeraniumoil-based
polymer thin films. Table 2 summarises the key bands and corresponding bond vibrations.
In the precursor spectrum, a very broad and strong peak of around 3367 cm−1 was related to
the stretch vibrations of (O–H) bonds of the alcohol. In the film spectrum, this band reduced in
intensityandappearedat3436cm−1. Theverystrongpeaksformedat2962and2872cm−1 canbe
attributedtotheasymmetricstretchingvibrationmodes(CH andCH ,respectively)ofthemethyl
2 3
group. Inthepolymer,thesepeaksalsodecreasedinintensityandoccurredat2961and2875cm−1,
respectively. AverystrongstretchvibrationoftheCH bondappeared at2926cm−1, andrelated
3
to the methylene group, shifted slightly in the polymer spectrum towards a higher wave number
at about 2933 cm−1. A very weak peak at 2728 cm−1, related to C=O bond stretching, vanished
completely, and was not observed in the polymer spectrum. Strong peaks at 1730 and 1713 cm−1,
possiblylinkedtothesymmetricstarchingvibrationsofC=Obonds,andamediumpeakat1671cm-1,
probablyrelatedtothevibrationof(C=C)wereobservedintheprecursorspectrum. Thesebands
merged, decreased in strength, and appeared as a broad band in the polymer spectrum at about
1708cm−1. Sharppeaksat1452and1377cm−1 wereattributedtotheasymmetricandsymmetric
bending vibrations of C–H bonds, respectively. These vibrations were retained in reduced form
andobservedinthepolymerat1453and1376cm−1,respectively. Furthermore,peaksobservedin
thefingerprintregionoftheprecursorat1267,1174,1058,and1008cm−1,combinedandradically
reducedinthepolymerspectrum,andweredetectedasveryweakandunitedbroadbands.
Nanomaterials 2017, 7, 270 7 of 22
Table 2. FTIR spectra assignments for geranium essential oil and geranium oil-based polymer.
Group Frequency, cm−1
Assignment
Precursor Polymer
Stretching (OH) 3367 3436
Asymmetric stretching, methyl (–CH2) 2962 2961
Symmetric stretching, methylene (–CH3) 2926 2933
Asymmetric Stretching methyl (–CH3) 2872 2875
Stretching aldehyde (C=O) 2728 -
Stretching (C=O), aldehyde 1730 1708
Stretching (C=O) carbonyl 1713 -
Alkenyl (C=C) 1671 1625
Asymmetric bend methyl (C–H) 1452 1453
Symmetric bend methyl (C–H) 1377 1376
In-plane bending (O–H) 1267
Merged in broad
Skeletal (C=C) 1174
band
Stretching, alcohol (C–O) 1058 and 1008
Comparison of the resulting films at different deposition powers (10 W and 100 W) revealed that
with increasing input power, the crosslinking of the fabricated polymers increases. This is possibly
owing to the higher fragmentation rate that occurs in higher density plasmas, due to an increase in
inelastic collisions among high energetic electrons and precursor species, which could cause the
Nanomaterials2017,7,270 7of23
overlapping of electronic orbitals [34].
Precursor
C=O
C=C
C=O O-H
u.)
a. O-H
n ( C-H3 C=O
o C=O
ssi C-H2 C-H C-H C=C
mi
s C-H3
n
a 10 W
Tr
100 W
4000 3500 3000 2500 2000 1500 1000
Wavenumber (cm-1)
FFiigguurree 22.. FFTTIRIRs psepcetrcatroaf goef ragneiruamniuesmse netsisaelnotiila(lp roeicl u(rpsorer)cuanrsdogr)e raanndiu mgeorialn-biuasmed opiol-lbyamseedr fapborliycamteedr
faatb1r0icWateadn adt 11000 WW .and 100W.
4.3. OptiTcaalb Plero2p.eFrTtIiRes spectraassignmentsforgeraniumessentialoilandgeraniumoil-basedpolymer.
Determination of material optical properties, including transparency, the spectral dependence
GroupFrequency,cm−1
of the refractive index aAnssdig enxmtienncttion coefficient, and optical band gap energy, are essential in optics-
Precursor Polymer
related industries [35,36]. Independent of applied power, geranium oil-derived polymer films were
Stretching(OH) 3367 3436
revealed to be optically transparent within the visible wavelength range.
Asymmetricstretching,methyl(–CH2) 2962 2961
RefracStiyvmem inetdriecxst r(entc),h ian gm,meaetshuyrleen eth(–aCt Hd3e)scribes how 2li9g2h6t propagates throu2g9h3 3the medium, was
found to be Ansoytm smigentriifcicStarnettclhyi ndgempeetnhydle(–nCt Ho3n) the RF powe2r8, 7w2ith all curves chara2c8t7e5rised by a similar
Stretchingaldehyde(C=O) 2728 -
shape (Figure 3). At a short wavelength of 200 nm, the variation in the refractive index between
Stretching(C=O),aldehyde 1730 1708
polymers fabricaStteredtc haint ga(Cp=pOli)ecdar bponoywlers of 10 W and1 711300 W was approxim- ately ~0.0097. At
Alkenyl(C=C) 1671 1625
wavelengths above 900 nm, the variation in the refractive index was ~0.0069. This result agrees with
Asymmetricbendmethyl(C–H) 1452 1453
Symmetricbendmethyl(C–H) 1377 1376
In-planebending(O–H) 1267
Skeletal(C=C) 1174 Mergedinbroadband
Stretching,alcohol(C–O) 1058and1008
Thereductioninthepeakintensitiesanddisappearanceofsomepeaksinthepolymercanbe
interpretedastheprecursormoleculesbeingpartiallydissociatedasaresultofbeingsubjectedto
the RF plasma field [33]. However, the absorption band of C=O asymmetric stretching observed
inthe1800−1600cm−1 regionwaspossiblyduetothepostoxidationofthetrappedfreeradicals,
confined during film formation [34]. The emergence of a strong methyl peak at 1452 cm−1, and
relativelyweakermethylenebandat1377cm−1,confirmsthatgeraniumoil-derivedfilmscomprise
alargequantityofshortunsystematically-branchedchains,ratherthanlonglinearbackbonestructures,
astructuretypicallyassociatedwithplasmapolymer[26].
Comparisonoftheresultingfilmsatdifferentdepositionpowers(10Wand100W)revealedthat
withincreasinginputpower,thecrosslinkingofthefabricatedpolymersincreases. Thisispossibly
owing to the higher fragmentation rate that occurs in higher density plasmas, due to an increase
in inelastic collisions among high energetic electrons and precursor species, which could cause
theoverlappingofelectronicorbitals[34].
4.3. OpticalProperties
Determinationofmaterialopticalproperties,includingtransparency,thespectraldependence
of the refractive index and extinction coefficient, and optical band gap energy, are essential in
Nanomaterials 2017, 7, 270 8 of 22
previous studies that showed a variation of less than 1% for thin films derived from other essential
oil precursors, e.g., γ-terpinene and linalyl acetate [37,38]. In contrast, Cho et al. found that the
refractive index of ethyl-cyclohexane films increased when the RF power was increased [39].
The extinction coefficient (k), a measure of how a medium absorbs light at a specified
wavelength, also showed very little dependence on the applied power, especially in the high
wavelength region (above 900 nm), as the variation was 0.0013. At a short wavelength of 200 nm, the
variation in the refractive index for polymers fabricated at applied powers of 10 W and 100 W was
~0.0268. As the films have optical constants similar to glass and good transparency in the visible
region, it is suggested they are suitable candidates for integration into protective coatings in optical
and biomedical devices, such as lenses [38,40].
NanomFatiegruiarlse2 041 7s,h7o,2w70s UV–vis spectra of geranium oil-derived polymer thin films deposited on 8golaf2s3s
with increasing RF powers, from 10 to 100 W. The maximum absorption peak was observed at
approximately 290 nm, which possibly relates to π–π* transitions [41]. The absorption peak width
optics-relatedindustries[35,36]. Independentofappliedpower,geraniumoil-derivedpolymerfilms
increased slightly with an increase in the RF power. This could be associated with an increase in the
wererevealedtobeopticallytransparentwithinthevisiblewavelengthrange.
length of the conjugated π-system, which usually shifts the absorption maximum peak to a longer
Refractiveindex(n),ameasurethatdescribeshowlightpropagatesthroughthemedium,was
wavelength [42]. However, no significant shift in location or variation in magnitude of the absorbance
foundtobenotsignificantlydependentontheRFpower,withallcurvescharacterisedbyasimilar
peak was detected with respect to RF power. This was also observed in films fabricated from linalyl
shape (Figure 3). At a short wavelength of 200 nm, the variation in the refractive index between
acetate by plasma polymerisation [38]. The findings show that the absorption spectra are repeatable
polymersfabricatedatappliedpowersof10Wand100Wwasapproximately~0.0097. Atwavelengths
under similar deposition conditions (RF power, pressure, temperature, etc.). The optical absorbance
above 900 nm, the variation in the refractive index was ~0.0069. This result agrees with previous
results strongly suggest that geranium films can be used as encapsulating (protective) layers for
studiesthatshowedavariationoflessthan1%forthinfilmsderivedfromotheressentialoilprecursors,
organic electronics to extend the lifetime and preserve efficiency of oxygen- and water-sensitive
e.g.,γ-terpineneandlinalylacetate[37,38]. Incontrast,Choetal. foundthattherefractiveindexof
organic materials [40].
ethyl-cyclohexanefilmsincreasedwhentheRFpowerwasincreased[39].
1.70
10 W
1.68
1.5732 25 W
1.66 50 W
1.5675 75 W
x
e 1.64 100 W
d
e in 1.62 1.5618
activ 1.60 1.5561
fr
e
R 1.58 400 450 500
1.56
1.54
1.52
200 400 600 800 1000
Wavelength(nm)
(a)
0.14
0.020 10 W
0.12 25 W
0.018 50 W
ent 0.10 0.016 7150 0W W
ci
effi 0.08 0.014
o
c 0.012
ction 0.06 0.010400 450 500
n
xti 0.04
E
0.02
0.00
200 400 600 800 1000
Wavelength(nm)
(b)
Figure 3. Optical constants of geranium oil-derived polymer films fabricated at various deposition
Figure3. Opticalconstantsofgeraniumoil-derivedpolymerfilmsfabricatedatvariousdeposition
powers; (a) Refractive index; (b) Extinction coefficient.
powers;(a)Refractiveindex;(b)Extinctioncoefficient.
Theextinctioncoefficient(k),ameasureofhowamediumabsorbslightataspecifiedwavelength,
alsoshowedverylittledependenceontheappliedpower,especiallyinthehighwavelengthregion
(above 900 nm), as the variation was 0.0013. At a short wavelength of 200 nm, the variation in
the refractive index for polymers fabricated at applied powers of 10 W and 100 W was ~0.0268.
Asthefilmshaveopticalconstantssimilartoglassandgoodtransparencyinthevisibleregion,itis
suggestedtheyaresuitablecandidatesforintegrationintoprotectivecoatingsinopticalandbiomedical
devices,suchaslenses[38,40].
Figure4showsUV–visspectraofgeraniumoil-derivedpolymerthinfilmsdepositedonglass
with increasing RF powers, from 10 to 100 W. The maximum absorption peak was observed at
approximately290nm,whichpossiblyrelatestoπ–π*transitions[41]. Theabsorptionpeakwidth
Nanomaterials 2017, 7, 270 9 of 22
The energy gap of geranium oil-derived polymer films was determined using the optical
absorption coefficient data acquired from UV–vis spectroscopy measurements. The optical energy
gap was calculated using Tauc relation αE = A(E − Eg)1/n, where E is the photon energy, A is a constant,
and n(1/r) is related to the density-of-states distribution in the transport gap, and equal to 1/2, 3/2, 2,
or 3 for direct or indirect transitions [43]. The MATLAB program was used to convert UV–vis
spectroscopy data to a Tauc plot, determine the linear portion of the high energy region, and
extrapolate the line until it intersected with x-axis to estimate the optical band gap (Figure 4). The
best fit was observed at n = 1/2, which is related to directly allowed transitions.
NanomAat evriealrsy2 0s1m7,a7l,l2 d70ecrease of optical band gap with an increase in the input power was observ9edof i2n3
the films. Samples fabricated at 10, 25, 50, 75, and 100 W had Eg ≈ 3.67, 3.65, 3.60, 3.61, and 3.60 eV,
respectively. The optical properties of plasma polymerisation films significantly relied on the
increased slightly with an increase in the RF power. This could be associated with an increase in
structure of the p-conjugated chains in both the ground and the excited levels, and on the inter-chain
thelengthoftheconjugatedπ-system,whichusuallyshiftstheabsorptionmaximumpeaktoalonger
woraievnelteantigotnh [[4442]].. HThoew nevarerro,wnoinsgig noiffi Ecga nist sphoifstsiibnllyo coawtieodn otor vdaarniagtliionng ibnomndags ncirtueadteedo fitnh ethaeb spoorlbyamnceer
structure during the fabrication. At low RF power, there was low concentration of dangling bonds
peakwasdetectedwithrespecttoRFpower. Thiswasalsoobservedinfilmsfabricatedfromlinalyl
because of their saturation with hydrogen atoms, while higher RF power increased the fragmentation
acetatebyplasmapolymerisation[38]. Thefindingsshowthattheabsorptionspectraarerepeatable
rate in the plasma field that accelerated deposition of chains with unsaturated bonds [45]. The
undersimilardepositionconditions(RFpower,pressure,temperature,etc.). Theopticalabsorbance
unsaturated bonds were expected to be reason for the foundation of structural defects and/or created
results strongly suggest that geranium films can be used as encapsulating (protective) layers for
some intermediate energy levels due to structural reorganisations enhancing the density of localised
organic electronics to extend the lifetime and preserve efficiency of oxygen- and water-sensitive
ostragtaens iicnm thaete braianlds [s4t0ru].cture that end in low values of Eg.
0.35 10W
25W
0.30 50W
75W
0.25 100W
)
.
U
.
A 0.20
(
e
nc 0.15
a
rb
o
s 0.10
b
A
0.05
0.00
200 300 400 500 600 700
Wavelength (nm)
(a) (b)
Figure 4. (a) UV–vis absorption spectrum of geranium oil-derived films; (b) The optical energy gap
Figure4.(a)UV–visabsorptionspectrumofgeraniumoil-derivedfilms;(b)Theopticalenergygapof
of geranium oil-derived films fabricated at various radio frequency (RF) powers.
geraniumoil-derivedfilmsfabricatedatvariousradiofrequency(RF)powers.
4.4. STuhrfeaceen Teorpgoygrgapahpy of geranium oil-derived polymer films was determined using the optical
absorptioncoefficientdataacquiredfromUV–visspectroscopymeasurements. Theopticalenergygap
Topological and biochemical characteristics of a surface in part determine the rate of microbial
wascalculatedusingTaucrelationαE=A(E−E )1/n,whereEisthephotonenergy,Aisaconstant,and
adhesion. Bacterial growth and proliferation agre highly associated with the size of the micro/nano-
n(1/r)isrelatedtothedensity-of-statesdistributioninthetransportgap,andequalto1/2,3/2,2,or3
features of the surface. There is an ongoing debate among scientists regarding preferential
fordirectorindirecttransitions[43]. TheMATLABprogramwasusedtoconvertUV–visspectroscopy
attachment of bacteria to rougher surfaces. Proponents of the theory attribute increased cell adhesion
datatoaTaucplot,determinethelinearportionofthehighenergyregion,andextrapolatetheline
to three main factors, namely higher surface area available for attachment, protection from shear
untilitintersectedwithx-axistoestimatetheopticalbandgap(Figure4). Thebestfitwasobservedat
forces, and chemical changes in cells that cause preferential physicochemical interactions [46]. At the
n=1/2,whichisrelatedtodirectlyallowedtransitions.
microscale, where the cell is much smaller than surface features, the roughness would indeed
Averysmalldecreaseofopticalbandgapwithanincreaseintheinputpowerwasobserved
provide “hiding places”. While at nanoscale, where the features are smaller than the size of the cell,
in the films. Samples fabricated at 10, 25, 50, 75, and 100 W had E ≈ 3.67, 3.65, 3.60, 3.61, and
rougher surfaces would provide less points of attachment than gsmooth surfaces. Also, the
3.60 eV, respectively. The optical properties of plasma polymerisation films significantly relied
distribution of the features on the surface is important. It has been reported that Pseudomonas
on the structure of the p-conjugated chains in both the ground and the excited levels, and on
aeruginosa and S. aureus cells preferentially attached to surfaces constructed of regularly spaced pits
the inter-chain orientation [44]. The narrowing of E is possibly owed to dangling bonds created
of 1 µm and 2 µm in size, but not those constructed ogf irregularly spaced pits of 0.2 µm and 0.5 µm
inthepolymerstructureduringthefabrication. AtlowRFpower, therewaslowconcentrationof
[47].
danglingbondsbecauseoftheirsaturationwithhydrogenatoms,whilehigherRFpowerincreased
In this study, AFM images were obtained to determine information regarding the surface
the fragmentation rate in the plasma field that accelerated deposition of chains with unsaturated
properties. Several parameters were used to describe the surfaces, including maximum height (Smax),
bonds [45]. The unsaturated bonds were expected to be reason for the foundation of structural

defectsand/orcreatedsomeintermediateenergylevelsduetostructuralreorganisationsenhancing
thedensityoflocalisedstatesinthebandstructurethatendinlowvaluesofE .
g
4.4. SurfaceTopography
Topological and biochemical characteristics of a surface in part determine the rate of
microbial adhesion. Bacterial growth and proliferation are highly associated with the size of
the micro/nano-features of the surface. There is an ongoing debate among scientists regarding
Nanomaterials 2017, 7, 270 10 of 22
average roughness (Sa), root mean square (Sq), skewness (Ssk), and kurtosis (Ska). Smax, Sa, and Sq were
used to estimate the topographical characteristics of geranium oil-derived polymer films, while Ssk
and Ska were referred to when describing the surface regularity. Representative images of the surfaces
of geranium oil-derived films are shown in Figure 5, and roughness parameters for the films are
summarised in Table 3.
The maximum peak height of 8.30 nm was observed on the surface of the sample fabricated at
the highest deposition power (100 W) with a scanning area of 3 µm × 3 µm. Sa and Sq increased slightly
as a result of an increase in the input RF power. At powers of 10, 25, and 50 W, Sa values were 0.18,
0.21, and 0.29 nm, respectively, with higher Sa values reported for the samples deposited at 75 and
100 W, at 0.63 and 0.69 nm, respectively. The increase in the roughness may be related to more
energetic ions at higher RF power causing more surface bombardment and etching [48]. Irrespective
of fabrication power, the average roughness values remained below 0.7 nm, confirming the smooth
and uniform nature of plasma polymers from geranium oil.
Surface skewness measures the symmetry of the deviations of a surface profile about the mean
line. Its value can be positive or negative, and it is sensitive to the irregularity of deep valleys or high
peaks [49]. The skewness parameter can be employed to distinguish between surfaces with similar
root mean square roughness or arithmetic average height values, but different shapes [50]. The
mNaxanimomuatmeri avlsa2l0u1e7 ,o7f, 2s7k0ewness for geranium oil-derived plasma polymer films was found to be 100.7o5f 23
nm on the sample deposited at 75 W (10 µm × 10 µm). Another parameter frequently used to describe
surfaces, surface kurtosis (Ska), describes the distribution of the protrusions with respect to the mean
preferentialattachmentofbacteriatoroughersurfaces. Proponentsofthetheoryattributeincreased
line. For a mesokurtic distribution that is similar to or identical to normal distribution, the kurtosis is
celladhesiontothreemainfactors,namelyhighersurfaceareaavailableforattachment,protectionfrom
zero. Distributions with positive kurtosis are leptokurtic, and are characterised by high peaks,
shearforces,andchemicalchangesincellsthatcausepreferentialphysicochemicalinteractions[46].
whereas platykurtic distributions have negative kurtosis, and are characterised by flat-topped
Atthemicroscale,wherethecellismuchsmallerthansurfacefeatures,theroughnesswouldindeed
curves [51]. All investigated samples exhibited Ssk and Ska values falling fairly close to 0 and 3,
provide“hidingplaces”. Whileatnanoscale,wherethefeaturesaresmallerthanthesizeofthecell,
respectively, suggesting surfaces with a symmetrical distribution of peaks and valleys [29]. It has
roughersurfaceswouldprovidelesspointsofattachmentthansmoothsurfaces. Also,thedistribution
been observed that skewness and kurtosis parameters of geranium oil-derived films are independent
of the features on the surface is important. It has been reported that Pseudomonas aeruginosa and
of the RF deposition power.
S.aureuscellspreferentiallyattachedtosurfacesconstructedofregularlyspacedpitsof1µmand2µm
Independent of deposition power, the topographical features appeared to be uniform, smooth,
insize,butnotthoseconstructedofirregularlyspacedpitsof0.2µmand0.5µm[47].
and pinhole free. The uniformity indicates that polymerisation reactions occurred essentially on the
In this study, AFM images were obtained to determine information regarding the surface
surface of the substrate, instead of in the gas phase. The variances in the sample profiles, and both
properties. Severalparameterswereusedtodescribethesurfaces,includingmaximumheight(S ),
max
scanning areas, were not statistically significant. Besides possessing considerable smooth surfaces,
averageroughness(S ),rootmeansquare(S ),skewness(S ),andkurtosis(S ).S ,S ,andS were
a q sk ka max a q
films fabricated by the plasma polymerisation process reveal high spatial uniformity and good
usedtoestimatethetopographicalcharacteristicsofgeraniumoil-derivedpolymerfilms,whileS
sk
adhesion to the substrate [52]. AFM results also revealed that the films’ entropy, or the level of
andS werereferredtowhendescribingthesurfaceregularity. Representativeimagesofthesurfaces
ka
disorder or randomness in a system, increased as a result of fabrication power. The entropy increases
of geranium oil-derived films are shown in Figure 5, and roughness parameters for the films are
are possibly related to the surface flatness decreases [53].
summarisedinTable3.

Figure 5. Typical three-dimensional atomic force microscope images of 3 µm × 3 µm scanning area of
Figure5.Typicalthree-dimensionalatomicforcemicroscopeimagesof3µm×3µmscanningareaof
geranium oil-derived film surfaces fabricated at various RF power.
geraniumoil-derivedfilmsurfacesfabricatedatvariousRFpower.

Table3. Surfaceprofilesof3µm×3µmand10µm×10µmofgeraniumoil-derivedfilmsurfaces
fabricatedatvariousRFpower.
Sample 10W 25W 50W 75W 100W
Scanningarea(µm) 3×3 10×10 3×3 10×10 3×3 10×10 3×3 10×10 3×3 10×10
Max,Smax(nm) 1.93 3.75 2.39 4.21 3.07 3.52 6.59 6.02 8.07 8.30
Averageroughness,Sa(nm) 0.18 0.23 0.21 0.23 0.29 0.30 0.63 0.58 0.69 0.60
Rootmeansquare,Sq(nm) 0.23 0.30 0.27 0.30 0.36 0.38 0.81 0.74 0.89 0.77
Surfaceskewness,Ssk 0.02 0.08 0.17 0.03 0.04 0.04 0.56 0.75 0.59 0.67
Coefficientofkurtosis,Ska 0.06 0.55 0.05 0.81 0.01 0.03 0.54 1.37 0.69 1.07
Entropy 3.09 3.45 3.28 3.44 3.72 3.80 4.84 4.70 4.98 4.75
Themaximumpeakheightof8.30nmwasobservedonthesurfaceofthesamplefabricatedatthe
highestdepositionpower(100W)withascanningareaof3µm×3µm. S andS increasedslightlyas
a q
aresultofanincreaseintheinputRFpower. Atpowersof10,25,and50W,S valueswere0.18,0.21,
a
and0.29nm,respectively,withhigherS valuesreportedforthesamplesdepositedat75and100W,at
a
0.63and0.69nm,respectively. Theincreaseintheroughnessmayberelatedtomoreenergeticions
athigherRFpowercausingmoresurfacebombardmentandetching[48]. Irrespectiveoffabrication