We present a new liquid chromatography–tandem mass spectrometry (LC-MS/MS) method for simultaneous determination of sodium lauryl sulfate and sodium laureth sulfate homologues in the range of alkyl chain length C12–C16 with 0–5 ethoxy groups. The method is based on scanning the precursor ions fragmenting to m/z 80 and 97 (Precursor Ion Scanning mode), which makes it specific for species with easily cleavable sulfate groups.
Trang 1journalhomepage:www.elsevier.com/locate/chroma
Katarzyna Pawlaka, Kamil Wojciechowskia,b,∗
a Faculty of Chemistry, Warsaw University of Technology, Noakowskiego 3, Warsaw 00-664, Poland
b SaponLabs Ltd, Noakowskiego 3, Warsaw 00-664, Poland
a r t i c l e i n f o
Article history:
Received 29 March 2021
Revised 16 July 2021
Accepted 16 July 2021
Available online 22 July 2021
a b s t r a c t
Wepresentanewliquidchromatography–tandemmassspectrometry(LC-MS/MS)methodfor simultane-ousdeterminationofsodiumlaurylsulfateandsodiumlaurethsulfatehomologuesintherangeofalkyl chainlengthC12 –C16 with0–5ethoxygroups.Themethodisbasedonscanningtheprecursorions frag-mentingtom/z80and97(PrecursorIonScanningmode),whichmakesitspecificforspecieswitheasily cleavablesulfategroups.Bymonitoringfragmentationofthusdiscoveredquasi-molecularionswewere abletounequivocally identifyallsulfatespecies presentincomplex mixtures ofalkyland alkyl-ether sulfateswithmolecularweightrangingfrom200to600m/z.Becauseoftheintrinsicsulfate-sensitivity, the presented methodcanbe alsoappliedto non-sodiumsalts ofalkyl- and alkyl-ethersulfates (e.g ammonium,mono-ortriethanolamine,etc.),whichareoftenusedbycosmeticmanufacturerstojustify themisleadingSLS-andSLES-freeclaims(whereSLSandSLESrefertosodiumlaurylsulfateandsodium laurethsulfate,respectively).Theuseofreversedphaseliquidchromatography(RPLC)columnwithC4 insteadofC18shortenedsignificantlytheoverallanalysistimeandallowedustouseasemiquantitative method(basedonsinglestandardforQuantitativeAnalysisofMulti-componentSystem,QAMS)to deter-mineseveralSLSandSLEShomologuesinonerunwiththelimitofquantification(LOQ)=0.4μg/mLand
ofdetection(LOD)intherange0.12–0.97μg/mL.Themethodwassuccessfullyappliedto17commercially availablecosmetic/householdproductsallowingverificationoftheirmanufacturers’declarations
© 2021TheAuthor(s).PublishedbyElsevierB.V ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/)
1 Introduction
Anionic surfactants are major constituents ofmost detergents
andcosmeticcleaning/washingproductsbutarealsoubiquitousin
many other formulations, where wetting, dispersing, emulsifying
orfoaming activitiesarerequired[1].Mostofthecurrently
avail-able shampoos,shower gels,liquidsoapsanddishwashingliquids
are based on alkyl andalkyl-ether sulfates produced by
sulfona-tion ofalkyl alcoholsandethoxylated alkylalcohols Inindustrial
practicepurealcoholsareveryrarelyused,typicallytheirmixtures
areemployedinstead(e.g.amixtureobtainedfromhydrolyzedand
hydrogenated coconut and palm oils, as inthe case of the most
popularalkylsulfate– sodiumlaurylsulfate(SLS)).Itsethoxylated
analogue (sodiumlaureth sulfate, SLES) is produced analogously,
withanadditionalintermediate alcoholethoxylationstep[2].The
∗ Corresponding author at: Faculty of Chemistry, Warsaw University of Technol-
ogy, Noakowskiego 3, Warsaw 00-664, Poland
E-mail addresses: kpawlak@ch.pw.edu.pl (K Pawlak),
kamil.wojciechowski@ch.pw.edu.pl (K Wojciechowski)
“lauryl” nameand its derivative“laureth” were introducedto high-lighttheill-definedstructuresoftheresultingmixtures,incontrast
to singlespecies,e.g sodium dodecylsulfate(SDS) The ethoxyla-tionstep introduces anotherheterogeneity ofchemical structures
ofSLES (number ofthe ethoxy units), inaddition tothe variable length of the alkyl chain presentin SLS Consequently,the com-merciallyavailablealkylandalkyl-ethersulfatesshowwide distri-bution ofchemical structures andsurfactantproperties[3,4] De-spitetheirgreatefficacityinloweringsurfacetensionand sustain-ing foams, SLS and SLES may pose some environmental hazards caused by their limitedbiodegradability and persistentfoam for-mation [5] Their high detergent power may also lead to exces-sive lipidand proteinremovalwhen used ona daily basis [6–8] Thispromptssomeconsumerstoavoidproductswithsodium lau-rylandlaurethsulfatesandlookfortheir“milder” alternatives.In response, numerouscosmetic producers replace the sodium salts withthose of ammonium, lithium, ethanolamine,etc., or replace SLSandSLES withtheir homologousmixtures (e.g.“sodium coco sulfate”) to misguide the consumer with a different INCI (Inter-national Nomenclature of Cosmetic Ingredients) name [9] Given
https://doi.org/10.1016/j.chroma.2021.462421
0021-9673/© 2021 The Author(s) Published by Elsevier B.V This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ )
Trang 2the variety of homologous alkyl and alkyl-ether sulfate
surfac-tants,thereisanurgentneedforarobustanduniversalanalytical
methodcapableofdifferentiationbetweennotonlythealkylchain
lengthbutalsotheextentofethoxylationofSLES
One ofthemajorobstacles inquantitative analysisofcomplex
mixtures ofnaturalorsyntheticcompounds isthelackof
analyt-ical standards Incontrast tomanywell-defined compounds with
distinctstructures,noreferencematerialsareavailableformostof
the complex molecules(like proteins, lipids oralkaloids) or
mix-tures ofhomologues(suchasalkylandalkyl-ethersulfatesinthe
presentcase) Thisproblemcan be circumvented using
computa-tionalmethodsbasedonQuantitativeStructureandIonization
In-tensity Relationship(QSIIR).The methodwassuccessfullyapplied
to predictrelative levelsof29organicacids incomplexmatrices,
registered by product ionscanning.It offered accurate results(in
therangeof80–120%)for16organicacidsforwhichtheabsolute
concentrationswere quantified andusedasreference.Forthe
re-mainingorganicacids,suchaccuracycouldnotbeachieveddueto
lackofstandardsortoolowconcentrationoftheacidinthe
sam-ples [10] Suchapproachenablesdevelopmentofmethodsrelying
on a singlestandardforQuantitative esofMulti-component
Sys-tem(QAMS)[11,12]
Currently, most analytical methods for determination of alkyl
and alkyl-ether sulfate are based on spectrophotometric,
electro-chemical andchromatographictechniques[13–15].The officialEU
methodfordeterminationofanionicsurfactantsindetergents
de-scribed in Regulation (EC) No 648/2004 of the European
Parlia-ment and of the CouncilandinISO7875-1(1996)standardisbased
on formation of blue-colored salts of anionic surfactants with
the methylene blue dye, whichare determined
spectrophotomet-rically afterextractiontochloroform (MethyleneBlue Active
Sub-stancetest,MBAS).Althoughthemethodiscontinuouslyimproved
[16,17],theion-pairformationreactiondoesnotprovidesufficient
selectivitytoenabledistinctionnotonlybetweenthenatureofthe
anionicgroup(phosphate,carboxylic,sulfateorsulfonate)butalso
thealkylchainstructure
Another groupof non-chain-length-selectivemethods isbased
on formation ofionic associates betweenanionic surfactants and
cationic speciesusing potentiometric sensors [18] orsuppression
ofionicconductivityusingionchromatography[19].Usinga
sim-ilar approach,Levine et al.were capable ofdetermining
concen-tration of ammonium lauryl sulfate, sodium laureth sulfate and
sodium alkyl (C10SO4−– C16SO4−) ether sulfates present in
com-mercially available detergents (dynamic linear ranges: 1.0–500,
2.5–550 and 3.0–630 mg/L, respectively) [20] After
chromato-graphicseparation,amixtureoffourlinearalkylbenzenesulfonates
(C10SO4−- C13SO4−) was detected by UV spectroscopy[21] anda
mixtureofanionicandnonionicsurfactants(includingsodium
lau-rylsulfateandα-olefinsulfonate) usinganevaporativelight
scat-teringdetector(ELSD)[22].Nevertheless,selectivityisoftenwhen
suchdetectorsareemployed[21,23,24]
Regardless ofthedetectionmethod,anionic surfactantscanbe
separatedusingionexchangechromatography[19].However,most
commonlytheseparation isachievedthankstodifferencesinthe
affinity ofaliphaticchains to ahydrophobic stationaryphase For
this purpose, deprotonated anionic surfactants (pH > 7) can be
separated as neutral ion pairs formed with quaternary
ammo-nium cations (ion pair liquid chromatography, IPLC [25]) orin a
non-dissociatedform (pH< 7), usingreversed-phase liquid
chro-matography, RPLC The former method is, however, not suitable
for mass spectrometry because of significant signal suppression
[26] The advantage oflowering pHof the mobile phase is a
re-duction ofhydrophilicinteractions interferingwiththe
chromato-graphicprocess.Ontheotherhand,reducedpHmaylower
ioniza-tion efficiencyinmassspectrometry[26].Matthusandcolleagues
[27] separated the alkylbenzene sulfonate (LAS), alcohol
ethoxy-lates (AE), and alcohol ethoxylated sulfates (AES) by reversed-phasechromatographyusingaC18columnandammoniumacetate
tostabilizepH ofthemobile phase.The partiallyseparated com-pounds were detected using a fluorescence detection (FLD) and
a thermospraymassspectrometer workinginthe scanning mode (m/zrange200–800).Thisprovidedmoredetailedstructural infor-mationattheexpenseofsensitivity.Thepotentialforquantitative analywasexemplifiedbydeterminationofthetotalconcentration
ofactivesubstancesinsewageinrelationtotheAESandLAS com-mercial mixtures Dufour et al.employed ultra-high-performance liquid chromatography with high-resolution mass spectrometry (UPLC-HR-MS) to separate four homologues of alkylbenzene sul-fonate(4-dodecylbenzenesulfonicacid,DBSA).Thehomologues dif-fered in their alkyl chain lengths (decylbenzenesulfonate, unde-cylbenzenesulfonate,dodecylbenzenesulfonate, tridecylbenzenesul-fonate) and could be separated in the RPLC mode using C4, C18 andC30columns[28].Levineetal.employed RPLCcoupledwith electrospray ionization quadrupole ion trap mass spectrometry (ESI-Q-IT-MS)to simultaneouslydetermine threecommon surfac-tants:anamphotericcocoamphoacetate,anonionicalcohol ethoxy-lateandananionicSLES(dynamiclinearrange1.5–40mg/Lfor to-talamount ofSLES normalizedagainst commercialmixture)[29] Fouralkyl sulfateand2 ethoxymers ofalkyl-ether sulfate homo-logues were determined in the SPE-preconcentrated wastewater samplesusing a liquidchromatography–tandemmass spectrome-try(LC–MS/MS) withelectrosprayionization (ESI)innegativeion mode Based on the commercial mixture producer’s declarations, the limits of quantification(LOQ) inthe range0.3–0.4 μg/Land 0.5–1.5μg/LwerereportedfortheSLSandSLEShomologues, re-spectively [30] Another interesting approach to separate surfac-tantspresent incommercial mixtures of SLS andSLES employed theionmobilitymass spectrometry[31] Theionized compounds were distinguished based on their drift time inan electric field, whichdependedontheirmolecular weight.Theauthorsobtained six peaks for the SLS mixture and twelve for the SLES one, al-thoughtheiridentitywasnotdeterminedduetolackofstandards Nevertheless,thedevelopedmethodcouldbeappliedtodetermine thetotalSLSandSLEScontent(withrespecttocommercially avail-ableSLSandSLESmixtures)adsorbedbydifferentdishsurfaces
In this contribution we further extend the analytical capa-bilities of LC-MS/MS technique by employing for the first time
an MS/MS scanning of the precursor ions to follow the sulfate-bearing species We propose a novel method for determination
of individual alkyl and alkyl-ether sulfates in mixtures of cos-metic/householdingredientsandproductsusingasingleandeasily availablestandard(SDS)forquantitativeanalyofmulti-component system(QAMS).Weshowhowtheexperimentallyobserved depen-dencyof signal intensityon theretention, numberof fragmenta-tionionsobtainedinacollisioncellandrecoveryfromastationary phasecan beaccountedforinQAMSmethodsto obtaingood ac-curacy.SDScouldbeemployedasauniversalandsolestandard,as
itwaspresentineverysample.Thenewmethodallowedusto de-tect5non-ethoxylatedand20ethoxylatedsulfatesincommercial SLESproducts.Toshowtheunprecedentedapplicationpotentialof thenewmethodforreal-lifesamplesanalyweverifiedthe manu-facturers’declarationsaboutthepresence/absenceofSLSandSLES ingredientsin17cosmetic/householdcommercialproducts
2 Experimental
Separationofthealkylandalkyl ethersulfateswascarriedout using a 1220 Infinity II LC Systems (Agilent Technologies, USA), whereastheir identificationandquantitationwasachievedwitha
6460Triple Quadtandem mass spectrometric detectorwithaJet Streamionsource(AgilentTechnologies,USA).Analyteswere sep-aratedby anAerisWidepore C4column(2.1× 150mm,3.6μm,
Trang 3Fig 1 Product ion mass spectra registered for dodecylsulfate anion (DS − , C 12 SO 4 − , m/z 265) and a selected anion present in SLES with their fragmentation ions
300 ˚A,Phenomenex).Thechromatographicseparationemployeda
gradient elution from (40%acetonitrile: 60% water) to (95%
ace-tonitrile: 5% water),both containing0.15% (v/v) formicacid, over
8 minataflow rateof0.2mL/min Boththe columnandmobile
phasewerethermostatedat45°C
Themassspectrometer wasoperatedinnegativeionmode
es-tablishedby ionizationvoltageof3000Vandnozzlevoltage0V
Heated(300°C)nebulizationgasflowof8L/minappliedat30psi
wasselectedasthemostappropriatetoenhancetheionization of
low molecularweightcompounds.The PrecursorIon(PrecI)
scan-ningwasappliedforthediscoveryofSLSandSLEShomologues.A
tandem massspectrometer operatinginthe PrecImode findsthe
parentions(thefirstquadrupole,Q1,operatesinthem/z200–800
scanning mode)forthem/z valuesofcharacteristicfragmentions
(the secondquadrupole,Q3,operatesintheselectedion
monitor-ing mode) The negative fragmentation ionsSO3− at m/z 80 and
HSO4− at m/ 97 were selected as characteristic of the
sulfate-bearing compounds (Fig 1, Table SM1) Structurally similar
sul-fonate ions would produce the m/z 81 signals of HSO3− ions in
addition to m/z 80 The m/z values of the parent ions were
se-lectively discoveredby themass spectrometer Theanalyses were
controlled and processedby a MassHunter Workstation software
(Agilent Technologies, USA) The employed chromatographic and
MSconditionsarecollectedinTableSM1
The commercial shampoos, hair conditioner and liquid soap
were purchasedinalocalcosmeticstoreinWarsaw(Poland).The
referenceshampoowithoutSLSandSLESwasprovidedby
Sapon-labsLtd.(Poland).ThreeSLESandfourSLS-typemixtureswiththe
specified amountof active substances(27–70%w/w) and
ethoxy-lation degree were obtainedfroman industrialsupplier All SLES
were declared assodiumsaltsofethoxylated sulfatesof
predom-inantly C12C14 alcohols with different ethoxylation degree and
average molecular weights (Mav ∼ 340 g/mol, ethoxylation
de-gree 1–2.5: SLES-340; Mav ∼ 384 g/mol, ethoxylation degree 1–
2.5:SLES-384;Mav ∼432g/mol,ethoxylationdegree>2.5:
SLES-432) The following SLS-type products were used: sodium
lau-ryl sulfate (SLS), ammonium lauryl sulfate(ALS), triethanolamine
lauryl sulfate (TEALS), monoethanolamine laurylsulfate (MEALS)
Sodiumdodecylsulfate,SDS(purissACSreagent,≥ 99.0%)was
pur-chased from Sigma Aldrich (Poland) Acetonitrile (LC/MS purity) from POCH (Gliwice, Poland); formic acid (LC/MS purity), from FisherScientific (FairLawn,NJ,USA).Demineralizedwaterfroma Milli-QsystemModelMilliporeElix3(Molsheim,France)was em-ployed
StandardsolutionofSDS(1.0mg/mL)waspreparedinMilli Q-water.ThesurfactantSLESmixtures(0.2–0.5mg/mL)and commer-cialcosmeticformulations(0.7–1.4mgofcosmetic/mL)were pre-pared by dissolving them inMilli Q-water For semi-quantitative analyzesofalkylandalkyl-ethoxysulfates,toassurethatthe con-centrationofallsulfatesiswithinthelinearresponserangeof cal-ibrationcurve,all sampleswere dilutedattwo ormorelevels.All solutionswerefilteredthrough0.45μmsyringefilterspriorRPLC analy
3 Results and discussion
3.1 Optimization of the detection conditions
Inordertosimultaneouslyquantifyallalkylandalkyl-ether sul-fate homologues, the separation method should allow for selec-tionofallspeciesyieldinghydrogensulfateion(HSO4−,m/z=97) and/or radical sulfate anion (•SO3 −, m/z = 80) upon fragmenta-tion.Thesespecies,resultingfromdissociationofthe C-O-Sbond
inthesulfategroup,canbeconvenientlyselectedbyscanningions withm/zbeingreducedto80and97inthePrecursorIonScanning (PrecI)Mode Althoughthelatteroffers additionalintrinsic speci-ficityin MS analy, untilnow the PrecI modehas been employed almost solely in lipidomic and proteomics analyses [32] and its potentialinalkylsulfateanalyhasbeenlargelyunexploited There-fore,inthefirstpartofthestudy,thedetectionconditionsfor elec-trospraytandem massspectrometer (ESI-MS/MS) were optimized using sodium dodecylsulfate (SDS) and a cosmetic/household in-gredient SodiumLaureth Sulfatewithdeclared averagemolecular weightof384g/molandaverageethoxylationdegree1–2.5 (SLES-384) To this aim, an SDS solution of 1 μg/ml was introduced (5 μL) into the mobile phase stream (acetonitrile:water, 50:50 (v/v)) at a flow rateof 0.2ml/min (FIA-ESI-MS) Using SDS, the ionization voltage, gas flow rateand temperature, aswell asion
Trang 4transmission voltage (fragmentor voltage) providing the highest
signalsobservedatm/z265(correspondingtododecylsulfate
an-ion, CH3-(CH2)11-SO4−,(“C12SO4−” or “DS−“),were optimized (see
Table SM1 in Supplementary Materials) As alkyl and alkyl-ether
sulfates tendtoformstableanionsalsoinacidicsolutions dueto
high dissociationconstant ofsulfate group (pKa= ~2.4) the
mo-bile phasewasacidifiedwithformicacid(0.15%(v/v)) to
sensitiv-itybyeliminationofadductsformationandtoreducethenoiseof
themassspectrum[26].Analogousanalyemployingtheoptimized
conditionswasperformedforSLES-384,wherethehighestsignals
were observed atm/z 265,309,381 and441.They corresponded
to the C12 anion (CH3-(CH2)11-SO4−) previously found in SDS)
and to different laureth sulfate anions predominant in the
mix-ture:CH3-(CH2)11-(OCH2)1-SO4−,CH3-(CH2)13-(OCH2)2-SO4−,CH3
(CH2)11-(OCH2)4-SO4−,respectively(C12EO1SO4−,C14EO2SO4− and
C12EO4SO4−,respectively).Theselecteddodecylandlaurethsulfate
ionsweresubjectedtofragmentation(ProductIonScanning)using
the10,20,30,40eVcollisionenergy.Allresultingspectrafeatured
the m/z97signal,corresponding toHSO4−anion,confirmingthat
allselectedionsindeedcontainedthesulfategroup.Insomecases,
additionalsignalatm/z80,correspondingtoaradicalanionSO3−,
was alsoobserved Other signals, ifpresent, correspondedto the
breakdownoftheC OorC Cbondswithintheethoxylatedpartof
SLES molecules(Fig.1).Thehighestsignalswereobtainedforthe
collision energyof20and40 eV,andtheseconditionswere
em-ployedinsubsequentoptimizationofseparationconditions
3.2 Optimization of the separation conditions by RPLC
Having selected the optimum conditions for detecting the
speciesreleasingSO3− andHSO4−(m/z80and97)inPrecImode,
we optimizedthechromatographicmethod,startingwitha
selec-tion of mobile and stationaryphases Formic acid allowed
sepa-ration of SLS and SLES homologues with a selectivity
compara-ble to that achievedin the presence of trifluoroacetic acid(TFA)
in the mobile phase andbetter than that forammonium acetate
[23,24] Thepresenceofformicacidinthemobilephase helpsto
reduce the hydrophilicandenhance the hydrophobic interactions
ofseparatedcompoundswiththealkylatedsilicastationaryphase
It should be notedthat formic acid, incontrastto TFA, isnot an
ion-pairingagent,henceitspresencedoesnotdeterioratethe
sen-sitivityofMS-basedmethod.Theabilityofthemobilephase(0.15%
aqueoussolutionofformicacidwitha linearlyincreasingamount
of acetonitrile from0 to98% in20 min) to recoverSDS and the
SLES-384componentsfromthesurfaceofahydrophobicstationary
phase wastestedusingtwoPhenomenex150× 2,1mm columns
One was loaded with fully porous silica particles modified with
C18 aliphatic chains (Luna) andthe other withcore-shell silica
particles ofwide pores modifiedwithC4 aliphaticchains (Aeris)
Chromatographic performance duringgradient elution wasbetter
fortheC4-bedcore-shellcolumnwhichproducedsignificantly
bet-ter shape ofthe peaks(intensity 3to 10times higherandlower
width, see Fig SM1) The shape of the peak can be by: (1) too
stronginteractionsofthecompoundwiththestationaryphasenot
balancedby thecompositionofthemobilephasesolution,(2)the
rateofmovementofthecompounds inthecolumndependingon
the temperature, (3)particle orporesize inthe stationaryphase
[33–35].Consideringthat thecolumnsize,particlesizeand
gradi-entelutionmethodwerethesameinbothcases,theobserved
re-ductionofthepeakwidthmustberelatedtothesizeofthe
parti-cles’pores.Duetobetterchromatographicefficiency(peaks’shape)
withoutselectivityincomparisontothefullyporousC18column,
the core-shell C4 column was chosen for all subsequent
experi-ments Then,optimizationofthechromatographicseparation
pro-cess wascarriedout byexamining the gradient elutionprogram,
temperatureofthemobilephaseandcolumn,samplevolumeand sampledilution(Fig.SM2)
In order to validate selectivity of the MS detection in PrecI modeunderoptimizedconditions,twoother SLESmixtures
(SLES-340withdeclaredaveragemolecularweightof340g/moland av-erageethoxylation degree 1–2.5, and SLES-432with declared av-erage molecular weight of 432 g/mol and average ethoxylation degree > 2.5) were analyzed using the method developed with SLES-384.Thechemicalidentity oftheanionsdiscoveredby PrecI wasassignedusingtheProductIonScanningmode.Fragmentation ions m/z 80and 97, confirmedthe presence ofthe sulfate group
in all cases The difference between theoretical and experimen-tally established monoisotopicmass M parameter(defined as
M =|Mtheoretical-Mexperimental|/Mtheoretical·106) varied between 71 and 639 ppm (Table SM2), which is typical for low-molecular-weight-compoundsanalyzedwithquadrupoleanalyzers,duetothe intrinsiclowresolutionofthelatter[36]
AlltestedSLESmixturesshowsimilarchromatogramswiththe proportionofmorelipophilicderivativesincreasinginorder:
SLES-340< SLES-384 < SLES-432.A more detailedanalyshowed that retentiontime withthenumberofboth alkyl andalkoxylgroups (Fig.SM3),whichwasusedtoadditionallyconfirmidentityofthe homologues This wasespecially usefulwhen the compound sig-nalwastoosmalltoobtainarichProductIonspectrum.Alltested SLES sampleswere abundantin ethoxylatedspeciesspanningthe alkylchainlengthsfrom12to16carbonatoms,andthenumberof ethoxy groupsfrom1to5 (Fig.2, TableSM2).Surprisingly, how-ever, all samples featured also signals that could be assigned to non-ethoxylated sulfates(C12SO4−- C16SO4− typicalfor SLS-type products
3.3 LC-MS calibration using SDS
Inordertoallowforasemi-quantitativeanalyofthealkyland alkyl-ether sulfates in cosmetic/household ingredients and prod-ucts, the LC-MS system was calibrated using SDS as a standard (notethatincontrasttoSLES,SDSisasinglemoleculewithknown anddefinedstructure) As showninFig 3(inset), the calibration curve for the m/z 265 signal in PrecI mode is linear up to at least10μg/mLofSDS,confirmingthat theproposed methodcan
beusedforquantitativedeterminationofindividual alkylsulfates Nevertheless, an analogous quantitative analy of alkyl-ether sul-fates(SLES-type)ismoredemandingfortworeasons:(1)pure ref-erencesubstancesarenoteasilyavailable,(2)thehomologuesmay differnotonlyinthealkyl chainlength(Cn)butalsoin ethoxyla-tiondegree(EOm).Especiallythelatterfactmaycomplicate quanti-tativeanalysincethealkylandethoxychainscontributedifferently
topartitioningandfragmentationbehaviorofSLESmolecules(see Fig.SM3)
Astheconcentrationsofindividualcomponentsoftheanalyzed SLES mixtures were not known, calibration was performed by a series of dilutions to obtain concentrations ranging from 0.1 to
10 μg of raw material/SLES mixture in 1 ml of water SLES-432 wasused forthis purpose,as it contains the highestnumber of different SLS- and SLES-type molecules (Fig 2) The diluted so-lutions were analyzed using the protocol describedabove (PrecI mode) For each compound detected in the mixture,a good cor-relationwas achievedbetweenthe peak area andthe amountof SLES-432insolution(Fig.3,TableSM3).Forcomparison,alsothe curve forSDS isincludedinthesamegraph(m/z265).The slope
oftheresultingcurvesisproportionaltotheretentiontimeforall compounds detected inSLES-432,which isclearlya consequence
ofdifferentcontributions ofthe-CH2 and-OC2H5 groupstothe solubilityandconsequently– tothepartition coefficientof differ-entSLSandSLEShomologues.Theobservedreductionofthe
Trang 5sen-Fig 2 Extracted ion chromatograms (EIC) of SLES-384 (A) and SLES-432 (B) obtained for the precursor ions (indicated in A) discovered for the fragmentation ion m/z 97
(PrecI: ∗∗ → 97 (40 eV)), see Table SM1 for experimental details) The colors visible in the on-line version correspond to the calibration curves in Fig 3
sitivityforthecomponentselutedafterlongertimemaybecaused
byseveralreasons:
• reducedchromatographicrecoveryofanalytes[37]
• lower ion transport efficiency fromthe ionization to the
vac-uumchamberforhighermolecularweightions[38]
• suppressionofionsignalintensitybystrongionparingagents
andlower stabilityofionsin ESIchamber orcollisioncell for
highlyethoxylatedhomologues[26]
Without the use ofanalytical standards, the observed
depen-dence of sensitivity on retention time excludes a direct
quanti-tative determination of multiple components To circumvent this
problem, we propose a simple correction scheme usingthe
cali-bration data forSDS As shownin Fig 2,the latter ispresent in
all tested SLES mixtures, and its amount can be reliably
quanti-fied using the calibrationcurve fromFig 3.Note that in a more
general case, for samples devoid of SDS, it can always be added
to the sample in known amount as an internal standard In the
proposed scheme,each anionsignal isnormalized tothat ofDS−
(dodecylsulfate, C12SO4 − First, acalibration curve forDS− anion
wasobtainedusingsolutionswiththeSDSstandard.Next,changes
ofsignalsforeachiondetectedinthemixtureSLES-482(this
mix-turewasselectedforits highestnumberofdetectedSLSandSLES homologues)were acquiredbyserialdilutions.Thedependenceof signalheightonconcentrationwasthenestablishedusingalinear curvey=ax+b,whereyisthepeakareaandxisthe concentra-tionofthesample(ingredient, cosmeticproduct,etc.)inthe ana-lyzed solution(expressedinμg/mL,seeFig.SM4).Thesensitivity coefficient,fi,foranygivenalkyloralkyl-ethersulfateion(i)inthe sampleiscalculatedby normalizingtheslope (a) forthei-thion
tothatforDS− (Eq.(1)):
f i = a i , sample
Thedependenceoff i onretentiontimeisshowninFig.4andthe numerical values are collected in Table SM4 The retention time significantly with the length of the alkyl chain forboth SLSand SLES.Thesensitivityofthemethodisprimarilydependentonthe numberof-CH2 groupsintheSLShomologuesandethoxygroups presentintheSLEShomologues.Ithasalreadybeenreportedthat thelengthofalkyl chaininfluencesthechromatographicrecovery [37]aswell astheefficiency ofionizationprocess andion trans-missionfromtheelectrospraytovacuumchamber [38].However, forSLES homologues, the decreasein sensitivityismore notable
Trang 6Fig 3 Dilution correlation for SLES-432 obtained by LC-MS/MS method using Precursor Ion Scanning mode Colors of points and trend lines in an electronic version
correspond to the peaks in Fig 2 The inset shows a calibration graph for DS − obtained for SDS
Fig 4 Dependence of sensitivity coefficient on retention time depending on the number of -CH 2 - ( n ) and ethoxy groups ( m ) in different SLS ( C n ) and SLES (C n EO m SO 4 − ) derivatives
Trang 7Table 1
Comparison of the declared and semi-quantitatively (LC-MS/MS) determined amount of active ingredients and the average molecular weight for 7 cos- metic/household products ingredients
Product name Sum of active ingredients Average molecular weight and ethoxylation degree (in brackets)
Declared by manufacturer (%) Determined by LC-MS/MS (%) Declared by manufacturer Determined by LC-MS/MS
Thisismostlikelyrelatedtothenumberoffragmentationproducts
forSLES homologuescontaininghighernumberofethoxy groups,
whichlowerstheimpactofthem/z 97ion(Fig.1).Thechangeof
sensitivitywiththenumberofethoxygroupsislinearforthesame
lengthofalkylchainandconsequentlyalinearregressioncouldbe
applied Thelatter providesa convenientstatisticaldescription of
theagreementbetweenthefitandtheexperimentaldata,allowing
calculationsofstandarddeviationofrelativeconcentrationsofSLS
andSLEShomologues.Theconcentrationofeachsurfactantion(i)
canbethenestimatedusingthesensitivitycoefficient(Eq.(2)):
C i = (peak area)i − b DS , SDS
where:a DS , SDS andb DS , SDS arethecalibrationcurveparametersfor
DS− ionin a standard SDS solution (externalstandard), anddis
thesampledilution
The detectionlimitscanbecalculatedusingthesensitivity
co-efficients,f i ,andcalibrationparametersforDS−(Eq.(3))
LO D i =3.3·S D b,DS,SDS
whereS D b ,DS, SDS isthestandarddeviationofcalibrationcurve
pa-rameterbforDS−ioninastandardSDSsolution
Additionally,standarddeviation(SD)forthepeakareasinblank
samples (the shampoos devoid of SLS andSLES) wasestablished
(SDSh,n= 5) andit wasfound significantly lower thanSDb .The
LOD was in the range 0.12–0.97 mg/L, which is satisfactory for
determination ofSLSand SLEShomologuesin cosmeticproducts
Moreover,thesensitivityofthemethodcouldbefurtherimproved
by switchingtoaMultipleReaction Monitoring(MRM)mode and
monitoringthespecifiedpairsofparent/fragmentationions
3.4 Validation of the semi-quantitative method for alkyl and alkyl
ether sulfates determination
The method describedabove allows to correct forthe
experi-mentally observeddependenceofsensitivityon retentiontimein
mixtures ofpractically unlimited numberofSLS andSLES
homo-logues and enables their semi-quantitative determination In the
absenceofreliableanalyticalstandardsforanyotherthanSDS
ho-mologueofSLSandSLES,theanalyticalvalidityofthemethodwas
criticallyassessedbycomparingthesumofallsemi-quantitatively
determinedcomponentswiththetotalcontentofactivesubstance
(manufacturerdeclaration)insevencosmetic/householdingredient
products.ThreeSLESmixtures(SLES-340,SLES-384,SLES-432)and
four SLS-typederivatives:sodium laurylsulfate(SLS), ammonium
laurylsulfate(ALS),triethanolaminelaurylsulfate(TEALS)and
mo-noethanolamine lauryl sulfate (MEALS) were employed for this
purpose.Figs.5andSM5collecttheextractedionchromatograms
(EIC) showingthe signalintensityfromthenon-ethoxylated alkyl
sulfate anions (SLS-type, Fig.5A) and alkyl ethoxy sulfate anions
(SLES-type, Fig.5B) species(allselected in PrecImode) Thefirst
striking observation is that while the SLS-type ingredients (SLS,
ALS, TEALS andMEALS) aredevoid ofanyethoxylated impurities (Fig.5B)asexpected, theSLES-typeingredientscontainsignificant amountsofSLS-typederivatives(Fig.5A),asalreadynotedduring the qualitative analy(see Fig.2) Their presencecan be probably explainedbyincompleteethoxylationoftherawmaterialsusedfor theirproduction.Whensuch amixtureissubjectedtosulfonation reaction, the corresponding mixture of alkyl andalkyl-ether sul-fatesisobtained
The concentration of each alkyl sulfate and alkyl ethoxy sul-fateanionwascalculatedusingthesemi-quantitativemethod de-scribed in the preceding section and their sum in each of the products was compared with the active component content de-clared by the manufacturer (Table 1) The agreement between the declared and determined sums is satisfactory,especially tak-ingintoaccountthesemi-quantitativenatureofourmethod.Only forTEALSandMEALS thedifference ismore significant,probably dueto thelower dissociationdegree oftriethanolamine and mo-noethanolamine saltsof thealkyl sulfates which mayaffect both thechromatographicseparationandsubsequentMSdetection.The averageethoxylationdegree(seeSupportingMaterials)ofthethree SLESmixtureswithincreasingtheir averagemolecularweightand agreeswiththemanufacturer’sdeclaration.Theaveragemolecular weight of the alkyl and alkyl ether sulfates was also established
onthe basis oftheir composition(Table1) Also, inthiscasethe agreementwiththemanufacturer declarationsis satisfactory, fur-thervalidatingoursemi-quantitativemethod
Thepresentmethodhasbeendevelopedandvalidatedfor de-tection ofsulfate-based surfactantsbutcan be extended toother ionic surfactants It offers good sensitivity and allows the deter-mination ofSLS andSLES homologues usinga cheap and widely available standard substance– sodiumdodecyl sulfate(SDS) The employed PrecI mode allows even for detection of (ethoxylated) alkyl sulfates not yet described inthe literature In addition, the methoddoesnotrequireanya prioriknowledgeofm/zvalues spe-cificforeachparent andfragmentationion,which isan essential requirement of the MRM method Although the latter offers the highestsensitivity[39],thePrecImodestilloffers goodsensitivity andfacilitates interpretationofthedataascomparedtothemass spectraandchromatogramsobtainedinMSscanningmode
3.5 Alkyl (SLS-type) and alkyl ether sulfate (SLES-type) anions determination in cosmetic/household products
Havingestablishedandvalidatedthesemi-quantitativemethod for determination of SLS and SLES homologues, we analyzed 17 cosmetic/householdproductsandcomparedtheir contentofalkyl andalkylethoxy sulfates withthedeclarationsprovided by man-ufacturersintheINCI(InternationalNomenclatureofCosmetic In-gredients) lists The products were chosen in a wayto represent formulationswithdeclaredpresenceandabsenceofSLS-and SLES-typederivatives
Trang 8Fig 5 Extracted multi-ion chromatograms (EIC) of: A (left panel) - alkyl (non-ethoxylated, SLS-type) and B (right panel) - alkyl ethoxy (ethoxylated, SLES-type) sulfates
detected in cosmetic/household ingredients using Precursor ion mode (PrecI: ∗∗ → 97 (40 eV))
According to declarations, some of the selectedproducts
con-tain other anionic surfactants similar to SLS andSLES: sulfonate,
sulfosuccinate,sulfoacetate,taurateorisethionate,wherethesulfur
atom intheheadgroup isbound directlytoC-atom(see Table2)
BecauseofmuchhigherenergyrequiredtodissociateanS Cbond,
underthepresentlyemployedfragmentationconditionanddueto
lack of the forth oxygen presentin sulfate group, thesemoieties
wouldnotbe abletoproduceboth m/z 80and97signals There-fore, they could not interfere withdetermination of the sulfate-basedsurfactantsusingourmethodbasedonPrecImode.Itshould
be notedthat thefragmentationion observedatm/z 97can also
be obtainedfor H2PO4− anionproduced during fragmentationof phospholipids.Nevertheless,suchsignalscouldbeexcludedonthe basisofdifferentretentiontimesduetolongeraliphaticchainsin
Trang 9Table 2
SLS and SLES homologues content determined semi-quantitatively in cosmetic and household products compared with the producer’s declarations in the INCI lists (names in italics indicate alkyl (SLS-type) and alkyl-ether (SLES-type) sulfates)
No Product description
Sulfur-containing surfactants declared in
Alkyl ethoxy sulfates [%]
Compliance with INCI (SLS/SLES)
1 Moisturizing shampoo for dry hair Sodium Coco-Sulfate 14.69 ± 1.31 < LOD + / +
2 Nutrifying shampoo for dry hair Sodium Laureth Sulfate , Disodium Laureth
Sulfosuccinate, Sodium Lauryl Sulfoacetate
11.59 ± 0.53 11.25 ± 0.50 -/ +
4 Moisturizing shampoo for dry hair Sodium Methyl Cocoyl Taurate, Sodium
C14-16 Olefin Sulfonate 1.51 ± 0.79 < LOD -/ +
5 Shampoo for dry and damaged hair Sodium Lauryl Sulfate, Sodium Laureth
Sulfate , Sodium Xylenesulfonate
28.75 ± 1.39 4.54 ± 0.91 + / +
7 Moisturizing shampoo for normal and
dry hair
Sodium Coco-Sulfate 22.88 ± 1.03 < LOD + / +
8 Strengthening shampoo for greasy and
falling out hair
Ammonium Lauryl Sulfate, Sodium Laureth Sulfate
12.14 ± 0.52 0.62 ± 0.14 + / +
11 Anti-dandruff shampoo Sodium Laureth Sulfate, Sodium Lauryl
Sulfate , Sodium Xylenesulfonate, TEA-Dodecylbenzenesulfonate
9.22 ± 1.04 1.77 ± 0.42 + / +
12 Hypoallergenic shampoo for fair, dyed
and bleached hair
Sodium Laureth Sulfate , PEG-2
Dimeadowfoamamido-ethylmonium methosulfate
5.10 ± 1.06 4.31 ± 0.49 -/ +
13 Hair-repairing shampoo Sodium Coco-Sulfate 42.10 ± 1.87 < LOD + / +
15 Anti-dandruff shampoo Sodium Cocoyl Isethionate < LOD < LOD + / +
16 Shampoo for delicate and damaged
17 Bath and shower gel Sodium Laureth Sulfate 5.42 ± 0.32 5.97 ± 0.62 -/ +
phospholipids Anadditional verificationcould be done by
moni-toringthePO3−signal(m/z79)[40,41]
Thetestedcosmetic/householdproductsweredilutedwith
wa-ter andto ensurethat theresulting concentrationlies withinthe
linearrange,eachsamplewasadditionallydilutedtwicemore.The
total SLSand SLES homologues concentrations determined
semi-quantitatively are collected in Table 2 and the respective
chro-matograms are shown in Fig SM6 The majority of tested
prod-ucts indeed conforms to declarations and only in one case both
the alkyl and alkyl ethoxy sulfate anions were detected in
sig-nificant amounts ina shampoodeclared as devoidof any
sulfur-based surfactants (shampoo 6) Two other products declared as
free fromSLS-orSLES-typeingredients (ahairconditioneranda
shampoopreparedinourlabwithnoSLSorSLESadded)did
con-form to their INCI declarations It is worth stressing that one of
theproducts(shampoo12)accordingtotheproducer’sdeclaration
contained PEG-2 dimeadowfoamamidoethylmonium methosulfate
(Meadowquat),whereasulfategroup ispresentinthecounterion
(methosulfate) The latter would fragment producing the m/z 80
andm/z97ions.However,ourmethodisinsensitivetofalse
posi-tive resultsofthistype,since scanningforthePrecI-selectedions
isperformedonlyforionswithm/z above200.Thus,theamount
ofalkylsulfatesdeterminedusingourmethodisnotbiasedbythe
presenceofMeadowquat,andtheproductmostlikelyindeed
con-tainsnon-ethoxylatedSLS-typesurfactantsnotdeclaredintheINCI
list
In some cases (e.g shampoono.4) thelow amount of
unde-clared alkyl sulfates could be detected, suggesting their
uninten-tional use – forexample asan impurity presentin other
ingre-dients Given the abundance of alkyl sulfates in SLES-type
cos-metic/household ingredients depicted in Fig 5, the undeclared
quantities of SLShomologuescould have beeneven
unintention-ally introduced into final formulations (see e.g shampoos no 2,
12,14,17)
4 Conclusions
Inthiscontributionwe havedevelopeda newMS/MSmethod for selectivedetection of alkyl and alkyl ether sulfates based on PrecursorIonScanning(PrecI)Modefollowingtheirseparationon
aC4reversed-phaseLCcolumn.Forthispurpose,onlyions releas-ing the SO3− and HSO4− moieties (m/z 80 and 97, respectively) areselectedandfurthermonitored,providingselectivitytothe or-ganic sulfatespeciescharacteristic to homologuesofsodium lau-rylsulfate(SLS)andsodiumlaureth sulfate(SLES)withm/zinthe range 200–600 Other sulfur-bearing species commonly found in SLS-andSLES-freeproducts(sulfonate,sulfosuccinate,sulfoacetate, taurate or isethionate), where the sulfur atom in the headgroup
is bound directly to C-atom, are excluded based on the absence
ofHSO4− signal (m/z 97) Furtherwe havedevelopeda quantita-tive methodfordeterminationofdodecylsulfateionsanda semi-quantitativemethodfordeterminationofanyethoxylatedor non-ethoxylated alkyl sulfates The latter method is based on series
of dilutions providing the sensitivity coefficient (f i ) for each sig-nal,which enablessubsequentestimation ofconcentration ofthe givenspeciesin thesample(allin onerun using singlestandard – easilyaccessiblesodiumdodecylsulfate(SDS)).Theproposed LC-MS/MSmethodallowssimultaneousdeterminationofSLSandSLES homologues,confirmation oftheir identity,determinationof aver-agemolecularweightofsurfactantsanddegreeofethoxylation,as wellasdiscoveryofnewsulfatecompounds.Conventionally,these tasksrequiretheuseofthreedifferentmethods,whichcannowbe replaced bya singleone developedwithin thiswork The overall accuracyofthemethodstronglydependsonthequality of corre-lationbetweenf i andretentiontimeforeachgroupofhomologues established underthe samedetection conditions.Forthisreason,
we recommend to calibrate the method using mixtures with as manySLSandSLEShomologuesaspossible.Extensionofthe pre-sentedmethodtotheenvironmentalorfoodanalyisalsopossible, althoughbecause ofhigherrequiredsensitivity, thePrecursorIon
Trang 10Scanning(PrecI) shouldpreferentiallybe replacedbytheMultiple
ReactionMonitoring(MRM)mode
Using thenewlydevelopedmethodwehaveassayedfour
SLS-type andthree SLES-type cosmetic/householdingredients forthe
presence of CnEOmSO4− species in the range of n (alkyl chain
length) 12–16 and m (number of ethoxy groups) 0–5.While the
SLS-typeingredientsareindeeddevoidofanyethoxylatedspecies,
the opposite is not true for the SLES-type ingredients which
are contaminated with non-ethoxylated derivatives Finally, the
method was applied to 17commercial cosmetic/household
prod-ucts to verifytheir consistency with the manufacturers’
declara-tionsintermsofpresenceofalkyl(SLS-type)andalkylether
(SLES-type)sulfates.WhiletheSLES-likesurfactantscontentwasusually
consistent with the declarations, several formulations contained
undeclared SLS-like ingredients, most likely originatingfrom
im-puritiesinSLES-typeingredients
Declaration of Competing Interest
Theauthorsdeclarethattheyhavenoknowncompeting
finan-cialinterestsorpersonalrelationshipsthatcouldhaveappearedto
influencetheworkreportedinthispaper
CRediT authorship contribution statement
Katarzyna Pawlak:Conceptualization,Methodology,Validation,
Datacuration,Visualization.Kamil Wojciechowski:
Conceptualiza-tion,Writing– review&editing
Acknowledgment
Thiswork wasfinancially supportedbythe WarsawUniversity
of Technology, Poland Ms Aleksandra Chybicka is acknowledged
fortechnicalassistance
Supplementary materials
Supplementary material associated with this article can be
found,intheonlineversion,atdoi:10.1016/j.chroma.2021.462421
References
[1] M.J Rosen, J.T Kunjappu, Surfactants and Interfacial Phenomena, 4th ed., John
Wiley and Sons, 2012, doi: 10.1002/9781118228920
[2] D Myers, Surfactant Science and Technology, 3rd ed., John Wiley and Sons,
2005, doi: 10.1002/047174607X
[3] J Chen, X Hu, Y Fang, G Jin, Y Xia, What dominates the interfacial properties
of extended surfactants: amphipathicity or surfactant shape? J Colloid Inter-
face Sci 547 (2019) 190–198, doi: 10.1016/j.jcis.2019.04.002
[4] X Liu, Y Zhao, Q Li, T Jiao, J Niu, Adsorption behavior of fatty alcohol ether
sulfonate at different interfaces, J Surfactants Deterg 20 (2) (2017) 401–409,
doi: 10.1007/s11743- 016- 1918- 4
[5] S.A Mousavi, F Khodadoost, Effects of detergents on natural ecosystems and
wastewater treatment processes: a review, Environ Sci Pollut Res 26 (2019)
26439–26448, doi: 10.1007/s11356- 019- 05802- x
[6] J Bandier, B.C Carlsen, M.A Rasmussen, L.J Petersen, J.D Johansen, Skin re-
action and regeneration after single sodium lauryl sulfate exposure stratified
by filaggrin genotype and atopic dermatitis phenotype, Br J Dermatol 172 (6)
(2015) 1519–1529, doi: 10.1111/bjd.13651
[7] Jurek I., Góral I., Mierzy ´nska Z., Moniuszko-Szajwaj B., Wojciechowski K., Effect
of synthetic surfactants and soapwort (Saponaria Officinalis L.) extract on skin-
mimetic model lipid monolayers, Biochim Biophys Acta Biomembr 1861 (3)
(2019) 556–564, doi: 10.1016/j.bbamem.2018.12.005
[8] T Bujak, M Zagórska-Dziok, Z Nizioł-Łukaszewska, Complexes of ectoine with
the anionic surfactants as active ingredients of cleansing cosmetics with re-
duced irritating potential, Molecules 25 (6) (2020) 1433–1446, doi: 10.3390/
molecules25061433
[9] V.C Robinson, W.F Bergfeld, D.V Belsito, R.A Hill, C.D Klaassen, J.G Marks,
R.C Shank, T.J Slaga, P.W Snyder, F.A Andersen, Final report of the amended
safety assessment of sodium laureth sulfate and related salts of sulfated
ethoxylated alcohols, Int J Toxicol 29 (4) (2010) 151S–161S, doi: 10.1177/
1091581810373151
[10] L Wu, Y Wu, H Shen, P Gong, L Cao, G Wang, H Hao, Quantitative structure- ion intensity relationship strategy to the prediction of absolute levels without authentic standards, Anal Chim Acta 794 (2013) 67–75, doi: 10.1016/j.aca.2013 07.034
[11] A Stavrianidi, E Stekolshchikova, A Porotova, I Rodin, O Shpigun, Combina- tion of HPLC–MS and QAMS as a new analytical approach for determination
of saponins in ginseng containing products, J Pharm Biomed Anal 132 (2017) 87–92, doi: 10.1016/j.jpba.2016.09.041
[12] S.P Li, C.F Qiao, Y.W Chen, J Zhao, X.M Cui, Q.W Zhang, X.M Liu, D.J Hu,
A novel strategy with standardized reference extract qualification and single compound quantitative evaluation for quality control of panax notoginseng used as a functional food, J Chromatogr A 1313 (2013) 302–307, doi: 10.1016/j chroma.2013.07.025
[13] E Olkowska, Z Polkowska, J Namie ´snik, Analytical procedures for the deter- mination of surfactants in environmental samples, Talanta 88 (2012) 1–13, doi: 10.1016/j.talanta.2011.10.034
[14] E Olkowska, Z Polkowska, J Namie ´snik, Analytics of surfactants in the envi- ronment: problems and challenges, Chem Rev 111 (2011) 5667–5700, doi: 10 1021/cr100107g
[15] Y.R Bazel, I.P Antal, V.M Lavra, Z.A Kormosh, Methods for the determina- tion of anionic surfactants, J Anal Chem 69 (2014) 211–236, doi: 10.1134/ S1061934814010043
[16] B Wyrwas, A Zgoła-Grze ´skowiak, Continuous flow methylene blue active sub- stances method for the determination of anionic surfactants in river water and biodegradation test samples, J Surfactants Deterg 17 (1) (2014) 191–198, doi: 10.1007/s11743- 013- 1469- x
[17] K Sini, M Idouhar, A C Ahmia, A Ferradj, A Tazerouti, Spectrophotometric de- termination of anionic surfactants: optimization by response surface method- ology and application to algiers bay wastewater, Environ Monit Assess 189 (12) (2017) 1–12, doi: 10.1007/s10661- 017- 6359- 7
[18] N.M Makarova, E.G Kulapina, New potentiometric sensors based on ionic as- sociates of sodium dodecylsulfate and cationic complexes of copper(II) with some organic reagents, Electroanalysis 27 (3) (2015) 621–628, doi: 10.1002/ elan.201400491
[19] M.C Bruzzoniti, R.M De Carlo, C Sarzanini, Determination of sulfonic acids and alkylsulfates by ion chromatography in water, Talanta 75 (3) (2008) 734–
739, doi: 10.1016/j.talanta.2007.12.026 [20] L.H Levine, J.E Judkins, J.L Garland, Determination of anionic surfactants dur- ing wastewater recycling process by ion pair chromatography with suppressed conductivity detection, J Chromatogr A 874 (2) (20 0 0) 207–215, doi: 10.1016/ S0 021-9673(0 0)0 0155-2
[21] S Wangkarn, P Soisungnoen, M Rayanakorn, K Grudpan, Determination of linear alkylbenzene sulfonates in water samples by liquid chromatography-UV detection and confirmation by liquid chromatography-mass spectrometry, Ta- lanta 67 (4) (2005) 686–695, doi: 10.1016/j.talanta.2005.03.011
[22] H.S Park, C.K Rhee, Simultaneous determination of nonionic and anionic industrial surfactants by liquid chromatography combined with evaporative light-scattering detection, J Chromatogr A 1046 (1–2) (2004) 289–291, doi: 10 1016/j.chroma.2004.06.061
[23] S.H Im, Y.H Jeong, J.J Ryoo, Simultaneous analysis of anionic, amphoteric, nonionic and cationic surfactant mixtures in shampoo and hair conditioner by RP-HPLC/ELSD and LC/MS, Anal Chim Acta 619 (1) (2008) 129–136, doi: 10 1016/j.aca.2008.03.058
[24] S.H Im, J.J Ryoo, Characterization of sodium laureth sulfate by reversed-phase liquid chromatography with evaporative light scattering detection and 1 H nu- clear magnetic resonance spectroscopy, J Chromatogr A 1216 (12) (2009) 2339–2344, doi: 10.1016/j.chroma.2009.01.005
[25] M.Y Ye, R.G Walkup, K.D Hill, Determination of surfactant sodium lauryl ether sulfate by ion pairing chromatography with suppressed conductivity detection,
J Liq Chromatogr 17 (19) (1994) 4087–4097, doi: 10.1080/10826079408013602 [26] M Hol ˇcapek, K Volná, P Jandera, L Koláˇrová, K Lemr, M Exner, A Církva, Effects of ion-pairing reagents on the electrospray signal suppression of sulphonated dyes and intermediates, J Mass Spectrom 39 (1) (2004) 43–50, doi: 10.1002/jms.551
[27] E Matthus, M.S Holt, A Kiewiet, G.B.J Rijs, Environmental monitoring for linear alkylbenzene sulfonate, alcohol ethoxylate, alcohol ethoxy sulfate, al- cohol sulfate, and soap, Environ Toxicol Chem 18 (11) (1999) 2634–2644, doi: 10.1002/etc.5620181133
[28] A Dufour, D Thiébaut, L Ligiero, M Loriau, J Vial, Chromatographic behavior and characterization of polydisperse surfactants using ultra-high-performance liquid chromatography hyphenated to high-resolution mass spectrometry, J Chromatogr A 1614 (2020) 460731–460742, doi: 10.1016/j.chroma.2019.460731 [29] L.H Levine, J.L Garland, J.V Johnson, Simultaneous quantification of poly- dispersed anionic, amphoteric and nonionic surfactants in simulated wastew- ater samples using C 18 high-performance liquid chromatography-quadrupole ion-trap mass spectrometry, J Chromatogr A 1062 (2) (2005) 217–225, doi: 10 1016/j.chroma.2004.11.038
[30] C Fernández-Ramos, O Ballesteros, A Zafra-Gómez, R Blanc, A Navalón, J.L Vílchez, Determination of alcohol sulfates and alcohol ethoxysulfates in wastewater samples by liquid chromatography tandem mass spectrometry, Mi- crochem J 106 (2013) 180–185, doi: 10.1016/j.microc.2012.06.007
[31] Y Valadbeigi, M Tabrizchi, Application of Ion mobility spectrometry in study
of surfactants adsorbed on different dish surfaces, Int J Ion Mobil Spectrom
17 (1) (2014) 35–41, doi: 10.1007/s12127- 013- 0142- 4 [32] I.H.K Dias, R Ferreira, F Gruber, R Vitorino, A Rivas-Urbina, J.L Sanchez- Quesada, J Vieira Silva, M Fardilha, V De Freitas, A Reis, Sulfate-based lipids: