Ifps

In our quest tfor 2015. Theprove both timsupports only

Parting withtightly is stre“blanky” orstuffed animtrue that wwhen our haso it is time friend, ICS-A(and incorrdemonstratbeen prope It’s hard to the CERCLSuperfund Superfund pLab Prographrase “pdocumente One thing tof the moreICP-OES tedesigned evaluating was born (interferentstarget anaperhaps be200.7 and decipher) wcorrection vadopted bycountry - evCLP testing. ICS-AB accinterference

For ICP:

to get back onere’s a lot of hmely and helpa single e-ma

h somethingessful. Remer well-worn, mal from onewe cannot rands are ful for us all toAB, that chrectly) usede that all sp

erly addresse

believe thaLA program

program, program spam (CLP), w

producing ed quality”.

the CLP proe current anechnology i

and articspectral inte

the concs only) andalytes) sampecause proto 6010 werewas the CLverification y the vast mven those th

ceptance ces

Move On Fr

n track with phelpful detail pful. Enjoy! Bail contact pe

g we haveember havin

matted, ane of your chilreach for sol of yesterye

o part ways herished vehd by labs npectral interfed.

at it’s been m, better k

was enaawned the Ewhich served

data of

gram got rignd approvencluded, wulated me

erference coept of an ICS-AB (inples. So ocols in mee virtually LP’s spectraprotocol tha

majority of IChat have ne

criteria signi

rom ICS-AB

VolumPublish

publishing TWregarding ICPBut please remer lab.

e clung to g to retire th

nd malodoroldren? It’s aomething neear’s junk. A

with our truhicle so oftnationwide ferences ha

35 years sinnown as t

acted. TEPA’s Contrad to coin t

known a

ght, that nod methods

was a cleaechanism orrection. Thnalyzing ICSterferents ppopular (a

ethods such impossible

al interferenat it has beCP labs in tver perform

ificantly ma

B – It’s Time

me 26 Numbehed: December 201

WO editions ofP interferencemember to sh

so hat ous lso ew nd sty

ten to

ave

ce the The act the nd

one for

arly for

hus S-A

plus nd as to

ce en

the ed

ask

At thedownflawecriterianalyor modetecppb, ICS-AAs if bmovein a sAB”. 200.7 “If thvalueelemechecacronhas thereftargechecAs Brymoviestateddeadtime AB.

e

er 2 15

f LabNotes eae correction phare this with

e risk of buin, the CLP ed protocoia (+ 20% ytes) allows sore to go uction limits fthat’s a sign

AB was desigbeing flawee on, the apsunset claus Both the c (7.13.6) and

he instrumenes, fortify thents of interek) is just anym, ICS. Athe ability fore no one

et elements k samples! yan Brown’se, “Cocktad, “Coughlin

d, they stink to bury the

ach year, hereplus many moh others in the

ilding it up approach

l. The oof true va

spectral intendetected! or many an

nificant flaw.

gned for pre-d wasn’t enproved met

se for the uscurrently appd 6010D (4.5nt does noe SIC checest…”. SIC (a re-arrang

And every fu to display should be sinto interfe

s character il” so eloqn's Law: buup the joint

e concept o

e is our seconore articles thae lab; our sys

just to crushwas (and riginal accalue for arferences of When routnalytes are b

-1985 instrumnough of a rthod themsese of good proved EPA5) clearly stot display ck solution spectral inteement of

unctioning ICy negativespiking erence

in the uently ry the t.” It’s of ICS-

nd issue at should

stem

h it back still is) a

ceptance ll spiked f 200 ppb tine axial below 10

ments! reason to elves built old “ICS-

A method ate that, negative with the

erference the CLP

CP today e values,

N

… these wcalculatio

The LabCertogether standardizefor BOD, PhosphorusAmmonia Colorimetrybenchsheetprinted offversion andocument they are. AExcel benchresults! Simbenchsheetcalculationsfailures anddata doesnsupersaturaphosphorouabsorbancereporting rethe interachaving an aIn addition and highlibenchsheet

• Improve• Improve

and col• Increase• Enhance• Opportu

recomm The spreadson the LabCbenchsheet

BOD TSS b Amm Amm Tota

New Auto-C

will not onlyons, they wi

rt team has updat

ed benchshe TSS, To, a(both ISE a

y). Thts can f from a Pnd used your results

Alternatively, hsheets can

mply providet will not s, they will d provide hn’t seem qution, the

us samplee of the hequirements ctive benchauditor right to the benighting QCts have the b

ed analyst saed lab/auditlaboration ed efficiencied data colunity to imple

mended prac

sheets and PCert websitets can be do

D benchsheebenchsheet.monia by ISEmonia by Coal Phosphoru

Calculating

y provide acll highlight

put ted, eets otal and and ese be

PDF to

as the be used to c

e the raw donly provi

also notifyhelpful feeduite right. T

absorbane dilution high standawill all be m

hsheet. It’ there in thenefits of au

C failures, benefits of:

atisfaction or communi

es lection ement ctices (built-

PDF versions . Or individu

ownloaded: et. .

E benchsheeolorimetry bes benchshee

g Benchshe

ccurate QC failures

calculate yodata, and thde accura

y you of Qdback if yoThings like Dnce of

over thard, and T

made clear bs almost li

e lab with youto-calculatio

standardize

ication

in)

are availablual

et. enchsheet. et.

eets

s

our he

ate QC our DO

a he TSS by ke

ou! on ed

e

If youEkern6364.

… Hfede

The Lhas eto scoloriover ion electrit is aand laboraapplyThereensurand sMost using This mcoloriin waboth sampappro Here’writtedoes requirpreseimme The ETNT83samp Samp

On thCode

Sam

need any a at david.ek

HACH methoeral and sta

LabCert progencouragedswitch to imetric me

the traditsele

rode methoa better me

will save atory time aying to mae’s just a wrire that labs state preservlabs making

the Hach Tmethod useimetric tech

ater samplesapproved

ple analyses. oved.

s the wrinklen in prope not conformrements wit

erve sampleediately).

PA formatte31 (10205) inples by addin

ple collectioCollect samBest results Preserve theless with at Store at 4 CPreserved sa

he other hae) which is

mple Preser

assistance, pkern@wiscon

od instructiate rules

gram d labs

the ethod tional

ective od as ethod

the and moneyke the switnkle we neeare in com

vation requirg the switchNT plus Amm

es Test N Tuhnology to ds. TNT meth

for wastew Note that T

le: The coer EPA formm to NR 219h respect tes (if they

ed HACH menstructs the ng hydrochlo

n, preservamples in clean p

are obtained we samples by releast 2 mL of h

C (39 F) or lesamples may be

and, ch. N based on

rvation for

lease contansin.gov or 60

ions conflic

y. And now tch pretty ed to smoo

mpliance withrements. h to colorim

monia Methoube (TNT) vdetermine a

hod 830 andwater and TNT 832 is no

mplete TNT mat (Method9 or 40 CFR to the acid y will not

ethod for TNTuser to presoric acid (HC

ation and stoplastic or glass with immediateeducing the pHhydrochloric ass. e stored up to 2

R 219, (Wisn Clean Wa

colorimetr

ct Dave 08-785-

ct with

labs are regularly. th out to h federal

metry are od 10205. vials and ammonia d 831 are

aqueous ot similarly

method d 10205) Part 136 used to

be run

T830 and serve the Cl):

orage bottles.

e analysis. H to 2 or acid (HCl).

28 days.

. Admin. ater Act

ric NH3

requiremensamples be

So…what is In this case, trump the m 40 CFR Part

“Restricapprovedefinedmodify awould respecies digestiomay alspreservaand/or htime reqan apprmethod

Therefore youse (sulfuricH2SO4 to presamples, no

There has Force...as iin terror Yeah…chatime ago,

The ForceNaming C

ts in 40 CF preserved w

the correct the federal

method requ

136.6(b)(3) s

ctions. An anaed analytical

d analyte. In aan approved esult in measu of an analyte

on or total cyao may not moation holding

quirements roved

d.” ou MUST

c acid) eserve ot HCl.

been a gf millions of

and weange can hin this very g

e Awakens Changes Ta

FR Part 136,with sulfuric a

approach? and state ru

uirements.

states,

alyst may not m method for addition, an an method if the

urement of a de (e.g., a chananide distillatioodify any sam

great disturb voices sudd

ere suddehave that egalaxy, we

- Standardake Effect

requires thacid (H2SO4)

HCl or H2SO4

ules (NR 219)

modify an a method-nalyst may noe modificationdifferent formnge to a metaon). An analys

mple

o

bance in tdenly cried only silenceffect. A loalerted you

d Methods

hat .

4?

ot n

m or als st

of

the out ed. ng to

the apprStanand werethat;Cleaon Jchapdesiganal

GoneMethobe hathe EPaccoassoccoursCleanthis methoand MethoThereof theversioappro5210B SeemWrongapprocriticaunapfor yoa secreporor PT not blabs amethoan on

So…hMetho

s

fact that roach to hodard Metho Wastewatee cited in the; the EPA pr

an Water AcJune 1, 201pter NR 2gnated as ytical metho

e is the conods. In factard copy prPA will name

ording to thciated Stande, this chann Water Actchange foods…at leaNR 219 nowods proced

efore, rather e 18th, 19th,

on of metoved methoB- 2001.

ms pretty g! Using oved methal. For examproved me

our PT testingcond PT samrted using anmethod co

be uploadedare requiredods. Conseq

n-site evalua

Do not hmethods

Your SOPreferencenaming c

If you doon your r

how do yods” method

the EPA ow approve

ods for the Exer (i.e., “Se Federal Reromulgated

ct (40 CFR Pa5, the cha

219, the a a repositods for waste

ncept of “et, eventually rintings. Frome Standard heir year odard Metho

nge has onlyt to date; theor approvest not yet. T

w only cite dures by the

than being 20th, 21st ethod “5210od must now

trivial… and rep

hod numbemple, reportinethod referg results will rmple. This in inappropride are not d into our d

d to use—anquently, youation if you:

have a cops you referenPs have noe the new convention,

o not report tresults.

you know d(s) are curr

was chaned proceduxamination oStandard Megister. The the changeart 136) in 20nges took

administrativtory for aewater anal

ditions” of S there may m this point Methods pro

of approval ods commity been made EPA has ned drinkingThe Federalapproved S

eir year of aable to citeditions or th0B” for BOw be refere

right? orting ers is ng an rence require you ts because iate methodacceptabledatabase. nd follow—au will be cite

py of the ance on handt been cha Standard or the method

which “Srently approv

nging its ures from of Waters

Methods”) y did just es to the 012 , and effect in e code

approved ysis.

Standard not even forward, ocedures by the tee. Of

de for the not made g water l Register Standard approval. any one

he online OD, the enced as

to obtain PT results

d number e and will Similarly,

approved ed during

approved d, anged to

Methods

properly

Standard ved?

Online ver

The approvfootnote ocopy editio On-line memost currendate can bline subscri(http://www.st Individual mcost of $6purchase aletter desigprice. For(Ammonia),procedures ammonia.

Standard Mcross-refereWater Act Water Act ( Wastewater A

Drinking Wate

For routine wsummary ofmethod as currently ap

rsion:

ved date isn the methns), as illustra

ethod versiont approvede found on tption site _andardmethod

methods can69.00 each. a method, nations for tr example, , you get th as well as

Methods has nce tables, (wastewatedrinking wat

Approved Sta

er Approved S

wastewater f the officia well as a lispproved vers

s found by hod introducated below:

ns will alwad version. the Standard____________ds.org/Store/Bro

n also be pu Note thayou purchathat methodby purchase ISE and thall the othe

also develo one each

er) and the ter):

ndard Method

Standard Met

analyses, thel approved st of editionsion is printed

checking tction (in ha

ays reflect tThe approv

d methods o_____________owseSM.cfm).

urchased atat when yase all of td for that osing 4500-N

he colorimeter methods

oped a pair for the Cle

Safe Drinki

ds link (PDF)

thods link (PDF

e following isversion of ts in which td.

the ard

the val on-___

t a you the

one H3 tric for

of an ng

F)

s a the the

NoticStandnot gonlinewasteTSS teNote that cannhas bRegistmercy

Appro

We acodeLaborchecdatabnot bprope

e that if yodard Methodgoing to cute edition wewater testsesting, BOD r also that wthe 22nd eot allow me

been officiater. So we y here.

oved Metho

re also proves for PT resulratory of Hyk the methobase retains

be correct. er method c

ou have onds—or an ea it. Only th

will cover a. Even if yorequires the

while Standaredition is alsethods from

ally promulgare essentia

d Codes for

viding a list olts. Note tha

ygiene (SLH) od code ves the code u Be proactiv

code for all P

nly the 20tharlier editione 21st editio

all of the cou only do B 21st editionrd Methods so “approv

m that editioated in theally at the

PTs

of approvedat if you use as your PT ery carefullyused last whve and re-rePT results.

h edition n--, that’s on or the common BOD and n or later. suggests

ved”, we on until it e Federal US EPA’s

d method the State Provider,

y as their hich may eport the

B

The back-stBackgroundcompensatcontributionWhy is the ban issue bemission intechange bchanges insupply/flowconcentratitemperaturemeasure thsample for a means backgroundclose proxim BackgroundFor many ensuring thafor backgroICP is quitecorrection methods 20

Backgroun

tory on backd correctioe for

n to the detebackgroundbecause quensity and based on the plasm, changes inion, ane/humidity. he net signa specific e

of accud intensity amity to the p

d correction:analyses w

at labs are nound response the oppos

is an abso00.7(§2.2) an

nd Correcti

kground corron is a m

variable ermination o variable? Tuantitation

background many facta, changesn sample visnd eve Therefore

nal intensityelement, we

rately detat the time eak of intere

: desired…owe focus onot adjustingse (e.g. blansite. In factolute require

d 6010D (§2

on in ICP-

rection mechanism

backgrouof the analythe variabilityis based

intensities cors, includi in the argscosity or acen roo, to correcy of a giv have to haermining t of analysis est.

or required? our efforts sample resu

nk subtractiot, backgrouement of E2.3)].

OES

to nd es. y is on an ng on cid om ctly ven ave the in

on ults

on). nd PA

BackgspectIt is correinterfecorreintereevaluinterfeadjus Backgintensaffecinterfecorrestemminterfecorreone factoto theadjusinterfesubtraof inteare stargeintenspositivback backg BackgModeto dedecismost correcorreelemeis not need the em Convcorreadjacselect

ground corretral correctioimportant tction is sepaerence corrction is a

element corruated as werence cortment of the

ground corsity unrelateted by) erence coction mademing from aerent elemction historicor more

ors. Becausee emission intments (baerent, a poacted from terest). Whesituated inc

et peak integsity which mve spectral the areground corre

ground correern instrumeeal with bions are ele

vendors ection for yoction is madent and wav “plug and to be evalmission lines

ventional, ction (OPBC

cent to eacted for bac

ection is not on to be cleaarate and drection. In adjusted, trection (IEC)

well, while crrections (IEe backgroun

rrection adjed to (althspectral orrection (SI

e to adjust fa neighborinment. Spcally results inter-elemen

e spectral ontensity, IEC ased on the

rtion of thethe gross int

en backgroucorrectly—negration can

must be coml correctiona removeection point

ection modentation offeackground

ement and establish “dou, the poinde may notvelength youplay”; theseuated base YOU choose

or “off-pC) uses eithech target ankground co

required

t the same a

ar that bacdistinct fromfact, if bac

then any ) factors mu

changes to Cs) do not

nd correction

usts for chaough they

overlap. IC) is a for emissionng target anpectral inte

in the genent correctio

overlap alwafactors are e concentr emission in

tensity for anund correctioear an inte result in a mpensated n factor thd by thet (see Fig. 6A

els rs a numbercorrection.line specificefault” bac

nt (or pointst be suitableu choose to

e default paed on YOUR e.

eak” bacer one or twnalyte. Therrection nee

s

ckground spectral

ckground spectral

ust be re- spectral t require n.

anges in can be Spectral

separate intensity nalyte or

erference eration of on (IEC) ays ADDS negative ration of ntensity is n analyte on points

erferent--- negative for by a

hat adds e errant A below).

r of ways . These c. While ckground s!) where e for the use. ICP

arameters lab and

ckground wo points e point(s) eds to be

at a locatroutine-- or We’ve seconfiguratiocorrection aeither side oICP, howevgenerally nbackgroundfor those sappears at potassium). In additionmanufactursystems baAgilent’s FAMSF systemcorrection. these systevarious locadevelop areflects truepeak to bebackgroundsystems offecorrection. Setting backappropriateThe positionfrom spectrsame chanwhich occumeasured. suggests thlocation (backgroundboth sidesdeterminedadjacent tochanged mathematicimproved backgroundsystems aremethods hamuch. BackgroundFigure 1 is abackgroundstandard suThis is part o

ion that ieven unexp

een manons that eat locationsof the targever, is a caot better thd correctionsituations w an area of

to convenrers now of

ased on proACT and FB

m) to estab At the risk ms analyze ations arounan algorithe backgroue properly ind correctioer a means

kground corely n used shoural interferennge in bacurs at the a Note that

hat correcti(“The positd-intensity ms of the d by the coo the analy

and thcal appliccorrection d. These fittee acceptabave yet to

d correction great examd intensity chperimposed

of an excelle

s not affecected-- ana

y instrumestablish tw equidistantt analyte pese where tw

han one. Th points shou

where an esloping bac

ntional OPBffer a numboprietary alBC technoloblish “fitted” of over-sim backgroun

nd each tarm which nd, allowing

ntegrated. on the FACs of spectra

rrection poin

uld be as frence and shokground int

analyte wav while methon is madetion select

measuremenanalytical mplexity of

yte line”), tehere are cations thby calcula

ed backgrouble for usebe updated

-Examples mple demonshanges. This

d over a nitricnt ICP refere

cted by othalytes.

ment defawo points t from and eak maximuwo points ahe use of twld be reserv

element peckground (e.

BC, instrumeber of uniqgorithms (e

ogies; Varia” backgrouplifying thin

nd intensity rget peak a

(theoreticag the analyIn addition CT and Mal interferen

nts

ee as possibould reflect t

ensity as thelength beihod languae at a singted for tt, on either line, will

the spectruechnology h advanchat provi

ating a fittund correcti

e even if td to reflect

strating hows is a 6.25% Cc acid blankence offered

her

ault of

on um. are wo ed

eak .g.,

ent ue

e.g. n’s nd gs, at nd lly) yte to

MSF ce

ble the hat ng ge gle the or be um has ed de ed ion the as

w Ca k. d

by Inohttp:/opera

Figurebackgpointsboth the backgbias fo Figurepointsshortea slocan overlaanalyensur

organic Ven//www.inorgations-guide

e 2 is an eground corrs used on alocations aranalyte peground coror the peak

e 3 illustratess is more er wavelengping baselinappear as

ap from eleyte. Using twre accurate

tures at anicventure.

example ofrection poina non-slopinre containedeak itself. rrection poin integration.

s a case whappropriate

gths, it’s not ne. Somet a peak cevated leve

wo points in tpeak integra

es.com/icp-

f improperlyts. Not onlyng backgrod within the

Using thnts will resu

ere two bace. In the re

uncommontimes analytcoming off els of an athese situatioation.

Fig

y placed y are two und, but wings of hese as

ult in low

ckground egion of

n to have te peaks of wing

adjacent ons helps

. 3

Figure 4 is abackgroundone is reallwavelengthfor backgrothe sampltitanium (at

Figure 5A seEither side standard vipeaks apbackgrounddecision is mon the longwe have to(and impac

Unless we lowith other however, arespect to picture.

an example d correctionly forced to

h side (pointound correcte containeypical).

eems, at firstlooks clear ew!). While

ppears to d correctionmade to mager waveleno consider wct backgroun

ook at how target ana

a poor choiOPBC. Fig

of a case wn point is apo place it ot A). Selecttion would bed copper

t glance like(from the s

e either sidebe app

n (based onke backgrou

ngth side. Awhat else Cnd correctio

alternate linalytes impacice could bg. 5B show

F

F

where a singpropriate, aon the shortion of point

be a problem (typical)

e a no-brainsingle eleme

e of the targpropriate n this scan),und correcti

And that’s wCOULD appeon).

nes associatct integratiobe made w

ws that bigg

ig. 5A

Fig. 5B

gle nd

rter t B

m if or

ner. ent get for , a ion

why ear

ed on,

with ger

Whileshort high ldue ton ththerepeak,explain meversio205.55since 205.55Cr. Figure2 baaffecspect In Figapprobackgaffecintegrthis ccorreprope

Whenbackgnegathe becacorre

there’s a wavelengthevels, and t

to the traffiche longer wa’s a line for , and Ni is al

ains why 267ethod 6010D

on of metho52 line, but 1994. Ultim52 is not the

es 6A and 6Background ted by a chtral overlap.

g. 6A, there oximately 40ground corrted by an inration (and case, a sigction (IEC)werly identified

n the anaground cotive integratIECs had

ause, as discction points

possibility oh side, Be is nhus this is like

c jam of wavavelength si Fe that virtulmost obscu.716 is the pD. The cu

od 200.7 rect 200.7 has

mately, as illue best choic

B illustrate thcorrection

hallenging m

is Pb prese0 ppb. As threction pointnterference thus a negagnificant p

was requiredd and estab

alyst removrrection potion, but muto subsequussed earlie are made i

of Be impacnot typically ely the bettevelengths apide (Fe, Ni). ually overlapred beneath

preferred waurrently promcommends u

not been ustrated in f

ce of wavele

he effects ofpoints wh

matrix and a

ent in the sahe integratiots shows, on resulting in

ative result fopositive inted (which theblished).

ved the oint, there ch less so. H

uently be er, when bac

n regions as

Fig. 6

Fig

cting the found at er choice ppearing In fact, ps the Cr h Cr. This

avelength mulgated using this updated figure 5B, ength for

f 1 versus hen also adjacent

ample at on using 2

e point is negative

or Pb). In relement

e lab had

affected was still

However, adjusted

ckground ssociated

6A

g. 6B

with IECs, calso require What must bYour auditoon-site eval What a

correctiform of

For OPBpoints fo

When wshow me

Can yoselected

Can yoselected(and if n

Successdoes noany staalthougstandardproblem

Ultimately, demonstratreasonably the vendor overcome ppoints by cror over-comdiscussed fu Figure 7 is aThermo software fosoftware feawith nearly epurchased years or smuch oldesome sequehave this ca The Thermo “Wavelengtfor a givenNi 221.647emission elements. sample cobut Si, withthe right at

changing ths adjusting a

be availabler will ask youuation:

pproach doon (OPBG-1“fitted” corre

BC, can you or _______ (e

were these pe documentou explain d? ou demonsd for OPBC anot, do you hsfully responot necessarilyandard mixe

h analysis ds can ce

ms with back

you simplye that ba based and set up”. Itpoorly placereating an IEmpensated urther in a se

a screen caScientific’s

or ICP-OES. atures are aevery ICP ins

over the so. Unfor

er instrumenential units mapability.

Scientific soth Finder” p

n wavelengt7), shows lines from In this examntains not

h a wavelet 221.667 nm

e backgrouany associat

e for your auu the followin

o you use fo1 pt, OPBG-ection)? show me y.g., As, Cr,Ti,oints determtation? how these

trate that are free fromhave an IEC nding to thy require a

es, or individof high

ertainly helpkground corr

y need to ackground d not simplyt is possible,ed backgrouEC to adjust background

eparate artic

apture of iTEVA

Similar available strument last 10

rtunately nts and may not

oftware‘s program, th (here,

nearby other

mple, the only Ni, ngth on m. Had

und correctied IECs.

uditor? ng during an

or backgrou- 2 pt , som

your correcti Tl)?

mined; can y

e points we

the locatiom interferenc

in place?) hese questiolab to analydul standar

level mixp to highligrection.

be able correction

, “That’s wh, of course, und correcti for the undd. This will cle.

ion

n

nd me

ion

you

ere

ons ce?

ons yze ds, ed

ght

to is

hat to ion er- be

the mpointswouldbackgaffec SummThis aconcreporcorrecestabEPA 6retrievdemoalgori Sadlylost ahas informnationmust sameaffecthe sais modetai

method incs on either sd be easy foground corted by the S

mary ll explains werned with rt will likely ction point

blished or d6010D (4.1.1ve their baconstrate thaithm employ

y, this detail oart. Like the

simply notmation. If wnwide accrbe looking

e critical thit data quaaying goes, ost certainlyl.

cluded bacside of the N

or the analysrrection poSi interferenc

why your aud backgrounindicate thts have n

documented1.2) ]. Lab

ckground coat the poyed) are app

of ICP-OES ise old telepht passed oe are ever treditation stg at the ings that ality. As the devil y in the

ckground cNi peak mast to removeoint that wce.

ditor may seend correctioat, ICP bacnot been . [EPA 200.

bs must be rrection scheint(s)s selecpropriate.

s quickly bechone gameon all theto truly havetandard, all

orrection ximum, it

e the right would be

em overly on. Your ckground

properly 7 (4.1.4), able to eme and cted (or

coming a , training

e salient e a single auditors

… Take steavoid systefailures by the time to the cause anecessary Deal with tnest early tgetting stu

A new edittime to incsince it is unresolved actions taklaboratoriesdata qualifreported satravels, we really well, these ideasrequiremen In NR 149,action requ149.38, sectas follows: 1. Take co

procedumeet lim

2. Identify tproblem;reviewinghad theproblem

3. Documecorrectivan experesults are

4. Monitor tadditionanot resolv

Note: On tthat we preach columwork?’ andto make suelements an

Corrective149 and p

eps to ematic taking consider

and make changes. hat wasp to avoid ng later.

tion of LabNlude an art common problems a

en and docs. We havefiers are repample resulthave seen sand this is

s, along witts.

, our admiuirements cations 1-4. The

orrective acres and wits. he source o; and haveg that the

e desired e is not reoccnt the corre

ve actions/ceditious mae released othe effectiveal correctiveve the probl

the correctivrovide to lamn that ask

d ‘How do yre there is and documen

e Action: Hprevent reo

Notes marks ticle on cor

for us auand insufficiecumented, e noticed tpeatedly ass. Through osystems in plour opport

th reminders

nistrative ruan be founde sections ar

tion for dehen QC sa

of the problee a system

corrective ffect (the

curring). ective actio

changes muanner - beor reported);eness of the e action if them.

ve action labs there ak specificall

you know tha reminder tnt them.

ow to comoccurrence

a convenierrective actiditors to sent correctiwhen we vhat the samssociated wour lab to llace that wounity to shas of the ba

ule, correctid in section re summariz

epartures froamples fail

m, correct t in place action tak

cause of t

ons taken (tst be donefore affect changes; ta

he change d

ogbook sheare headers ly ‘Did the e fix worked

to assess tho

mply with Nes

ent ion ee ive

visit me

with ab ork are sic

ive NR ed

om to

the for

ken the

the in ed

ake did

eet in fix

d?’ ose

Oftennot fa

Unforthave depareal cthey a

You athe onis whrespo

Whet

No Th

ov A

rebe

As

Theseis takeheadthen probledata doculess timbeen progr Makeand athe tiproporeducof reJon Syou ssay so Largelabs tanalytheir Lthe Qtrends An augood(and Depaand when

NR n this seems all into the tra

tunately we a long list o

artures) that ocorrective aare repeatin

are NOT menly time a c

hen a clientonse to an ev

her it is a ma

ot knowing the quantity verwhelming hasty dec

equire corree documentssuming the

e can be haen. If failuresed off at the

there will ems leads to (and camentation ome explainin fixed to

ram is vital to

e sure thoseanalysts) arme to studyose the chce the likele-occurrenceStewart saysmell some

omething”.

er commthat do a

ysis may be LIMS to gen

Quality Manas in the data

uditor recend system in p

potential prartment ran this report

never: the co

easier said ap of excuse

e have noticof non-confooccur - but r

action or chng failures.

eeting NR 14corrective act complainsvaluation fin

atter of:

the approac/ complexit

g cision that t

ctive actionted) problem wil

ndled if the s are not ignoe pass - by p be feweo reduced talls from of problems ng why theseyour audit

o the succes

e on the froe strongly e

y the cause hanges thalihood e. As ys, “If

ething,

mercial lot of using

nerate summager can usea.

ntly visited a lace to quicroblems). Tha weekly represented fontrol limit w

than done es!

ced that mormances (frarely do weange taken

49 requiremction is docs or a PT fading.

ch to take ty of the fai

he failure dn/change (o

ll go away o

time it takeored (or evepreventative

er problemstime spent q

data use going forwe problems tor. An acss of a labora

ont line (tecencouraged of the prob

at will elim

mary informae as a review

lab that hackly identify phe Quality Aeport from thfor each pa

was exceede

– but do

many labs ailures or

e see any n, even if

ments IF… umented ails or in

ilures are

does not or that it

on its own

es to do it en better, e action), s. Fewer qualifying ers), less ward, and

have not ctive QA atory.

chnicians d to take blem and inate or

ation that w to spot

ad a very problems

Assurance heir LIMS, arameter ed once,

or when thtimes in a rolab departrequired. A note on wwide as themany sourcIt’s as simp‘warning’. This approarequiring cofailure, it gowatching fowarning limleading to a A couple of

Whinccaswproexpopropro

Macohege

Whdoco

Colimtha

Sochde

he warning ow. This repotments and

warning limitse control limces, one of w

le as it’s na

ach is greaorrective ac

oes one step or trends by

mit exceedaanother failu

f comments:hen correcticomplete (nause), it’s a liwatting at a f

operly dealinxample, a laboor BOD blanoblem of caoducing pooake sure laborrective actelp supervisoet the informhile a failureoes not eliminorrective actorrective act

mits (e.g. wheat is not corr

ometimes wehronic (kind oescribe pain…

limits wereort is provided corrective

s- warning limits. This is we

which is Standamed, it’s m

at because ctions when further since

y looking at ances that a

re.

: ve actions taot getting tottle bit like cfew wasps inng with the nb that has a nks, and nev

alibration accor BOD resul staff understion is about rs and qualitation they n requires quanate the neetion. tions are noten the SOP inrect). e use terms liof like how a…). Acute w

e exceededed back to te action w

mits are not ell explaineddard Metho

meant to be

it is not othere is a Q

e they are a the repeatappear to

aken are o the real ontinuously

nstead of nest. For problem wi

ver solves thecuracy, is ts. stand that having themty managers

need. alification, thed for

t limited to Qncludes info

ke acute ana doctor maywould be a

d 4 the

was

as d in ds.

e a

nly QC lso ed be

th e

m s

his

QC

nd y

We habelowwhencontr

Co

A. WhBOD resolverraticsubsta0.4 mthe dbeen negathe ca B. WhThereand Bto havariab

one-timewould bthat poireviewefour timemay quehas not

Labs musince theindicatoproblem

ave a few ew – we also an solutions harol problems.

orrective“Som

aboobviou

hat was the pprobe ca

ved (an audc blank reantially eithe

mg/L. This waata generat affected. tive and erralibration.

hat caused the are five opBOD is analyave consistenbles that are

e occurrencbe ongoing (nt to a syste

er notices thaes a standarestion why t been workeust documeney could be

or of a biggem.

examples of cappreciate have been fou.

e action

metimes you cout … and sous that some

was

problem? alibration pdit finding). Tesults whicer more negs the secondted for over

Since the ratic, the first

he problem?perators, twoyzed on multncy in the c

e not control

ce, and chro(repeating) f

emic issue. If at one out ord is failing lohis chronic p

ed on and cont the acute

e the beginner (or chronic

corrective ahearing fromund to quali

n exampcould tell wh

ometimes it weone had los on an endle

- Todd R

problems wThis was indicch were frative or posid repeat find three years blanks art issue to res

? o DO probetiple shifts. It calibration wlled.

onic failures a data f every

ow, they problem orrected.

e failures, ing

c)

actions m labs ity

ple 1 hat it was was quite st it and it ess loop.” Rundgren

were not cated by requently itive than ding and has likely re highly solve was

es/meters is difficult

with these

Did it fix it? How do you

know?

C. What was done to try and fix the problem? The audit finding indicated the importance of consistency with calibration, which could be done by making sure a procedure, placed within easy access, was followed exactly by all staff. In addition, all staff attended a meeting to discuss the importance of a good DO probe calibration. These 10 steps were clearly written and accessible for a water saturated air calibration: 1. Fill a bottle with DI water to the line indicated

(1” from the bottom). 2. Cap and vigorously shake the bottle. 3. Take a Kimwipe and dab (very gently, not

rubbing the membrane) any droplet of water from the probe.

4. Place the probe in the bottle quickly. 5. Let the bottle sit for exactly 30 minutes using

the egg timer on the bench. 6. Make sure the room temperature meets the

requirements of 17-23oC and record. 7. Record barometric pressure from the meter. 8. Press the Calibrate button and record. 9. Compare the calibration value to the oxygen

saturation chart using the barometric pressure and DO meter calibration temperature. If the calibration value is off by more than 0.2, check for problems.

10. If the problem can’t be solved let the supervisor know.

D. Monitor the Corrective Action Each day’s calibration is reviewed to make sure it is within expectations compared to the oxygen saturation point. There was only one time in 20 sets that the blank was just high, 0.3 mg/L. The calibrations have been determined to be done well and in control, the problem has been fixed as seen by the stable blank results (and lack of negative blanks). Eventually the lab went the next step when replacing equipment by purchasing an LDO probe.

Corrective action example 2

“If you do not know how to ask the right question, you discover nothing.”

– W. Edwards Deming.

A. What was the problem? The ICP LCS failed for beryllium (Be), at 80% recovery. The control limits are 85-115%. B. What caused the problem? Not sure why Be failed, all other parameters recovered well (~95-105%). The other standard solutions passed. Re -prepared and reanalyzed the samples for Be. The next LCS passed. …This is an example where the cause isn’t really dealt with and it is chalked up to a random occurrence. The analyst needs to ask more questions…experience helps here too, if all the other parameters passed, it’s even more interesting of a problem that only Be failed. One key to ICP analysis is interferences – method 200.7 lists V and Ce as interferences, but looking at 6010B, titanium is listed as an interference. Taking another swipe at this… Reviewed the peak for beryllium in the LCS. There is a titanium (Ti) peak to the left of Be 313.107 (see image below). The LCS that failed contained both Ti and Be mixed in the same solution (which is a different mix from the calibration and verification standards). Typically titanium is not included in the LCS with Be since it is rarely requested for analysis. But still …how did this not show up in the interference studies? Reviewed the interference study data, there was no data that showed there was an interference for Be from Ti! Reviewing spectra from the interference study there was one background point on the right side of Be peak, and now there are two background correction points, one added to the left. Adding the background correction point to the left caused the problem when there is Ti present, since that intensity is subtracted out …resulting in a low result. C. What was done to try and fix the problem? Removed the left side background correction point. Analyzed a Ti standard at 50 ppm and there was no interference on Be 313.107. Reanalyzed the LCS that had previously failed low, now the recovery of all parameters pass 85-115% and Be recovery was 92%. Removing the added correction point fixed the problem, which is seen with the passing LCS.

Make sure tohttp://dnr.wThere is valutraining thatany lab (also

D. Monitor thfix the problproblem)? After 10 anthere have Note: This pthe instrumewithout maquality ICPtraining thecorrection parametersa problem).stop swattinblock whealways gooand reviewthe instrumeresolve the of the lab c

The Bottom When a

water s(synthet(volatilemust re

Reporting detects in

o visit our webwi.gov/regulatuable informat discusses reao see the det

he Correctivlem and how

nalysis sets obeen no fail

roblem is alsent method aking sure a analysis c analyst on

points alo when in the. Get to theng at the wn trying to

od to go bacw the methoent manufacproblem feeertification p

Line a compliancupply) sam

tic organic e organic coport any co

results for Public Wa

bsite: tions/labcert/tion in guidan

al issues that cailed informa

ve Action (diw do you kno

of an LCS thlures for Be.

so about ma settings are

all the basiccan be me

the pitfalls oong with e same solut root of the asps). If the resolve a ck to the baod along wicturer. If this el free to caprogram con

ce drinking ple is analy

contaminaontaminant) ompound de

r non-“targter Supply

Resources.htmnce and can occur in tion on ICP).

id the changow it fixed th

hat includes

aking sure the not changcs for ensurit (along w

of backgrouwhy certa

tion can cau problem (a

ere is a men problem,

asics of the teith contacti does not he

all or email ontacts.

water (pubyzed for SOant) or VOthe laboratoetected, at

get” analyty samples

ml

ge he

Ti,

hat ed ing

with nd ain use nd

ntal it’s est ng elp

one

blic OCs OCs

ory or

abincocl

ThSaWAdCimruSaWcaotnocorere

If rerethinnu

Pertin NR synthe(2) DNOT LAny contarepormonitThe deteclisted tentaresult quan NR 80organ(2) DELISTEDAny dnot lisdeparequirindicahas b

te

bove the strument waompounds rlient.

his is a requiafe Drin

Water, Wiscodministrativeode, the s

mplementatiule of the fedafe Drin

Water Act. Ifalibration ther QC foron-requesteompounds

eported as eported with

the calibraequested coesult is reporthen reportedicating thumeric or No

nent Adminis

809.207 Cetic organic

DETECTION LISTED IN S.NR

detection aminant not rted to thetoring reporlaboratory

cted synthe in s. NR 80tively ident is reportitative or an

09.247 Compnic contaminETECTION OFD IN S. NR 809detection ofsted in s. NR

artment withred under tate whether

been confirm

analyte LOas calibratedrequested to

irement spenking onsin e state on deral nking f the and

r the d are in cona quantitat the approp

tion and otompounds arted as an es

ed with theat the resu

ot Verified.

strative Rule

Compliance c contaminanOF SYNTHETR 809.20 (1). of a listed in s. N

e departmerts required

shall indicetic organic9.20 (1) hastified, and rted, whethn estimate.

pliance requnants F VOLATILE C9.24. f a volatile oR 809.24 shalh the otherhis section. r any detec

med or tenta

OD, for whd – regardleo be reporte

cified unde

ntrol - the tive result,

priate data q

her QC for re NOT in costimated resue data quault was eith

(Ch. NR 809)

requirements TIC CONTA synthetic

NR 809.20 (1)ent with th under this cate whethc contamins been confwhen a n

her the

uirements fo

CONTAMINA

organic conll be reporter monitoring The laboracted volatileatively identif

hich the ess of the ed by the

r NR 809,

result is and it is

qualifier.

the non-ontrol, the ult, and is alifier ‘0’ her Non-

)

ents for

AMINANTS

organic ) shall be he other section. her any nant not firmed or

numerical result is

or volatile

ANTS NOT

taminant ed to the g reports tory shall

e organic fied, and

when a numresult is qua What aboutTentatively be determinnot requirerecommendTIC at a higthe TIC respublic healt Fine. But hoelectronicaAll publiccomplianceWDNR elecThere are fcompoundscomment fi The instruccompounds At the bott“Add a ResuA screen apStoret Code

If you know the “Enter Sthe “Add thwish to addknow the Stthis Storet” b

merical resuntitative or a

t TICs? Identified Cned when used to be rd that if a lgh concentrult - as it mth.

ow do we relly?

c water e results muctronically tfields availas to be eld for data

ction for s are as follo

tom of the ult” button. ppears givinge to Add” or

the Storet cStoret code this Storet” bud a result ftoret code, ebutton, like t

lt is reportedan estimate.

ompounds (sing GCMS ireported. laboratory dration that itmay have a

eport these

supply drust be submhrough the ble for all dreported a qualifiers.

reporting ows:

list of results

g you two o “Find Storet

ode, you cato add:” fielutton. For efor Tert-butyenter it and his:

d, whether t

(TICs) that cnstruments aHowever, w

does detectt would repan impact

extra analyt

inking wamitted to t LDES systedrinking wa

along with

these ex

s, click on t

options: “Seles”:

an enter thatd and click

example, if yylbenzene aclick the “Ad

the

an are we t a ort on

tes

ter the em.

ter a

tra

the

ect

t in on

you nd dd

If youthe ddescr“Find”the dClick wish tresult nameparam After result be ataddin

Note:requeresult or Nocompdata requebelow

the imme The imThis awaterwater

u don’t knowdescription ription, then” button. Adescription the “selectto report. Fo for Tert-butye, click “Fmeter from t

either option page and tt the bottomng the result

: if the calibrested comp is reported wot Verified”pounds ARE qualifier aested compw is displaye

yellow fieldediately abo

mpact article is presr laboratorier supply

w the Storetin the “En

n click find:”A list of paryou entere” button foror example, ylbenzene, eFind”, andthe list that a

n, you will bethe paramet

m of the list. Y as you did f

ration and opounds are with “0”, me, as the d in control,

as you woupounds. Thd as a dropd

d column ove.

sented to mes are awa

drinking w

t code, ententer part o” field and rameters thaed will be r the param if you wish

enter a portio then sel

appears, like

e returned toter you addeYou can confor the other

other QC for NOT in con

eaning “Nonata qualifieuse the app

uld for anyhe text bodown list ide

in the sc

make sure alare of theswater com

er part of of Storet click the

at match returned.

meter you to add a on of the ect the this:

o the ed will ntinue rs.

the non-ntrol, the

n-numeric er. If the propriate y of the x shown

entified in

creenshot

l drinking e public mpliance

requiremenlaboratory unaware requiremenevaluations complianceaddress the If you harequiremenLaboratory Program. … Don’t fo May is less early to staenvironmenchanges anactions to properly. The State ofinancial sypayment topossible latstub at the bTo save timegoing to httinvoice num2016”) and as it appear Along with tFees Prograyour invoice To ensure qwith the follo

Ensuema

If yoMaychan

PleaDNRyour

2016 Env

ts. If not, thisto becomeof these ts will be ass and can

e mechanisese requirem

ve any qt please c

Certificat

orget to incl

than six mort prepping

ntal fees, espnd requires ensure thei

of Wisconsin stem to pro be process

te fee you mbottom of yoe you can ptp://dnr.wi.gmber (your fthe paymenrs on your pa

these changm is trying to

es.

uality on timowing:

ure that we hail address oour email addy 2016 pleasenges

ase add REnvironmentr contacts lis

ironmenta

s is the oppoe complian

requiremesessed during

result in a ms are notents.

uestions atcontact somtion and

ude the pay

nths away! folks for the

pecially wherecipients tor payment

has implemcess all invoed on time a

must includeour invoice. pay your invov/epay/. facility’s 9 dnt amount mayment stub

ges the DNR Eo go green b

me service pl

have the mosn file dress change inform us o

talFees@wiscst and mark

l Fee Invo

ortunity for tnt - if it wnts. Theg future on-s deficiencyt in place

t all on tmeone in t

Registrati

yment stub

It’s never t annual spri

en the proceo take spec

gets credit

mented a neoices. For yoand to avoide the payme oice online Note that t

digit FID plusmust be exacb.

Environmentby emailing

ease assist

st current

ges before of these

consin.gov tas not SPAM

ices

the was ese site y if

to

this the ion

!

oo ng ess ific ed

ew our d a ent

by the “-

ctly

tal

to M

If youcontaemail

…

The umodeand mlabs smeet prefeGC mapplicgene Po Pe 1,

Althoincorpswitchrequirdownapprothe un Beforeargumlet’s To mthis, makethe Machto reconcwhat betterespe

Selec

u have any qact Jeremy Kl at Jeremy.K

… You gottathis fish

use of GC/e is becominmore prevastruggle to g action limitrred option methods. cations for rated using

olynuclear aentachlorop,4-dioxane

ugh it maporate a hing over tored. Thoughn the road, uoved and pnfortunate re

e ments beback up a

make sensewe need

e a quick triWayB

hine. We nevisit the wh

ept of tunin we do now

er understanect to tuning

cted ion m

questions or cKahl at (608)Kahl@wiscon

a tuna h.

/MS SIM ng more lent. As

generate lowts, GC/MS S to using coSome of t

which wSIM mode a

aromatic hydphenol, and

ay seem full scan m

o SIM mode h that could until there is

promulgatedeality.

the egin, bit.

e of to ip in

Back eed hole ng. Understw (e.g., BFBnd where w prior to perf

monitoring

concerns ple) 261-4922 ornsin.gov

wer LODs reqSIM has beconventional the more ce encount

are:

drocarbons (

counter-intumode tune for SIM ana change soms national cod by the EPA

tanding why, DFTPP) wil

we need to forming GC/

(SIM) GC/

ease r by

quired to come the

detector common ter data

(PAHs),

uitive to e before alysis, it is mewhere onsensus, A, that is

y we do l help us be with /MS SIM.

/MS?

A GC/MS system is initially “auto-tuned” using PFTBA (perfloro-tri-nbutyl amine) as a means of optimizing the instrument for maximum sensitivity while providing mass resolution for specific masses designated in the tune algorithm over the range of masses being analyzed. Once tuned, the analyst then runs a specific analytical method. Two main GC/MS analytical methods were ultimately developed, one for volatiles and another for semivolatiles. Both of these methods were developed as “full scan” GC/MS. Although SIM has been around for 15 years or so, it is just now becoming widely used. Back in the late 1980’s, as part of the CLP (Superfund) program, the EPA established “target” tune criteria for BFB (volatiles) and DFTPP (semivolatiles) as a mechanism to ensure generation of uniform mass spectra among all laboratories generating and reporting environmental sample results for compliance testing. These target tunes were designed to actually de-tune the mass spectrometer to meet specific criteria established for the fragmentation pattern of either BFB or DFTPP. It’s this de-tuning process that sparks discussions when we audit labs performing SIM analysis. We get it. Sensitivity is critical for all target masses (in order to obtained the desired lower LODs), and thus “de-tuning” the instrument is counter-productive. Additionally, with SIM we are focusing on a handful of key masses (actually, m/z) and so other interferent or non-essential masses are filtered out. These are the arguments used to support a position that conventional BFB/DFTPP target tunes have no relevance to SIM applications. We agree that using conventional BFB/DFTPP tuning criteria is inherently problematic, but we lack an acknowledged alternate approach. We’re hoping that we can all agree that some tuning process is required. We posed the question to the EPA’s Methods Information Communication Exchange (MICE), although it was framed in terms of method 8270, as that is the base method typically used for SIM. MICE provided the following response (which we modified to include 8260) – with which we agree and we will hold labs to.

Methods 8260 and 8270 require a tune check be performed prior to analysis, note the use of the word "must" below:

…GC/MS system must be hardware-tuned to meet the criteria … for a 50 ng injection of BFB/DFTPP. Analyses must not begin until the tuning criteria are met. [8260B § 7.3.1, 8270C § 7.3.1]

The GC/MS must be tuned to meet

the recommended BFB [DFTPP] criteria prior to the initial calibration and for each 12-hr period during which analyses are performed. [8260C § 9.2, 8270D § 9.3]

MICE went on to add that,

Because the word “must” is used in the guidance and not “should” or “may”, a tune has to be performed by full scan, even for SIM work. The tune requirement is not just for library checks, but is also designed to demonstrate that the mass spectrometer is in control. Many laboratories have suggested that documenting a successful auto-tune report at the beginning of each shift should work, however the use of phrases such as “must” or “shall” in the methods make the specific procedure unalterable.

In summary, if you do not perform and pass a DFTPP check, you cannot call the method 8270C or 8270D. It would be considered a modification of the method.

There’s no ambivalence in the method language, and, as MICE points out, SW-846 with all its “guidance” disclaimers, states that, “The words "shall," "must," or "require" are used to indicate aspects of the method that are considered essential to its performance, based on sound analytical practices (e.g., an instrument must be calibrated before use).” Consequently, a tune which meets BFB/DFTPP criteria becomes a requirement of the method, whether used to perform full scan or SIM analysis. If a laboratory is not performing a full scan mode tune check before each SIM analysis run, it will result in a deficiency. The rules may not always make sense, but we do not make the rules; as a primacy state our role is

IS

to enforce consistentlythe system Some labs fnecessary, that an sufficient, asuggest somestablish deSIM analyzedifferent apsets of PAH Lab A c Lab B o

PFTBA. Lab C

standard Lab D h

focusingto 228.

Lab E hfocusingmuch b

Are the daDefensible?acceptabledamaging viable solutsome sort osimply not asuch time recognized analysis, weprocedure (e.g., 8270). OK…so we that analyzstandard is demonstrathave propand neutrinterferenceyou use? up just a mneed to BRelephant infactors (IEC

ICP EndgaStandard

the metho. This is whebreaks dow

feel no tuneothers fe

autotune and still otheme variationefensibility oed by fourpproaches tSIM data fro

hooses not topts to per

tunes in fud DFTPP critehas developg on 8-10 m

has developg similar to laroader acce

ata from a? Which dae? Being to

as being tion to the p

of tuning is inan option for

as the EP tune protoe are stuck

associated .

now underszing an “ICS of no valuing that perly idenralized spees. So whatWait, let’s b

minute. WeIEFLY discuss

n the room ).

me –Your (ICS)

ods ere wn. e is eel

is ers in betweenr validity of r different lto tuning?

om five differ

to tune at alrform an au

ull scan meria. ped its own /z ratios ran

ped its own ab D, but heptance crit

ll five labs ata should bo flexible coo rigid. C

problem at n order; not r defensible A develops

ocol when p with requiri with the

tand S-AB” ue in

you tified

ectral t DO back e first s the – intereleme

Interferen

n. How do wPAH results abs with fo Consider firent labs:

l. uto-tune usi

ode to me

tuning critenging from 1

tuning critehas establishteria.

comparablbe consider

can be just haos is nothand. Clea tuning at adata. So, us a genera

performing Sing the tunibase meth

ent correcti

nce Check

we by

our ive

ng

eet

eria 28

eria ed

le? red

as a

arly ll is ntil ally SIM ng od

ion

Let’s requirnothinvirtuainterfesamprequiranalychecthe lespectchasepotenWhy?elemeall eleonly interfe So, twhethIECs, interfedemospect So…wan iroensurregarFirst, tthe m(CLP) the reyou cprotointerfeswitchinterfeeleme Buildi 1. O

Ascrofinsimsowanca

be clear: wred. In facng but proc

ally interfereerences at

ples, then red. BUT….yyzing somethk standard,

evels you entral interferee, analyzing ntial interfer? Becauseents interfereements into

serves to erence pictu

the commoher you useor nothing

erences, yoonstrate thattral interferen

what do we onclad collere that your rding spectthere is no s

major flaw in [ICSA, ICS-

esolution of ychoose, an

ocol. Whileerence fromhing to werence free ent “Z” may

ng the perfe

One size doess discussereating a cf all theterferents inmply not olution. This

why both mend 6010 spalibration mi

we are not sct we havecessed drinence free.

the analyteno interferyou DO havhing (let’s c or ICS) to d

ncounter, yoence. And a single starents is sime many (if e with one a one standa

further muure.

on denomie an algorithg at all to ou must stit your data nce.

require? Hection of ICS data can tral interferesingle solutio the EPA’s C-AB] approayour instrumed your bac element “

m element “Ywavelength from eleme become an

ect ICS syste

s not fit all. ed abovechef’s salade possiblen one mix is

a viables is preciselyethods 200.7pell out a liixes.

saying that e labs that king water If you h

e levels yorence correve to proveall it an intedemonstrateu do not exlet’s just cu

andard contmply not an

not most)another. So ard, while exuddle the

nator here hm (e.g., FA

deal with ill provide is not biase

ow does onS “cocktails”stand up toence relateon that fits aontract Lab

ach. It depent, the wavckground c“X” may suY” at wavele “2” mayent “Y”, but n interferent.

m

, d e s e y 7 ist of 5 or

IECs are analyze that are

have no u see in ection is e that by erference e that at

xperience ut to the aining all

n option. ) of the jumbling xpedient,

spectral

is that ACT MSF),

spectral data to d due to

ne devise ” that will o scrutiny ed bias? all. That’s Program

pends on velengths orrection ffer from

ength “1”, y prove suddenly .

6 (6010)

If verifyinanalyzinelementcalibratsuggest becausedifferent

2. DefensibconcenThis onehas not ICS stanppm, tinterfereverified.you anawhich c(or eve“X”, yohave aany coor algorwith “Xthe levthe samthe correctivimmediastandardfound infor over

3. If you do

interfereThe anatree falls

ng lack of inng a single ts, how couion? It does different e each of tt analytes an

bility extendtration levels

e should be proved to bndards conthat is the

ence correc Consequealyze a sam

contains 30 pen 21 ppm)ou no lonassurance rrection fac

rithm associa” hold trueel identified

mple. Thereappropr

ve is ately analyd of “X” at

n the sample or under-co

on’t test it, yoe. alogy of thes in the fores

terference is standard uld that possn’t. And the

groupings he methodsnd waveleng

s only to thes tested. self-explana

be the case. tains elemee limit to ction for “ently, if—dowmple ppm ) of nger that

ctors ated e at d in

efore riate

to yze a sin a level hig

e to evaluateorrection.

ou cannot p

e age old qst….” fits quit

s difficult whcontaining ssibly work e two metho

of standars recommengths.

e

atory, but th If one of yo

ent “X” at which yoX” has bewn the road

ngle elemegher than the the potent

prove it does

question, “Ifte well here.

en all for

ods rds

nds

hat our 10

our en

d—

ent hat tial

sn’t

f a A

greacoyeinThrabedubiofmbi InprdeNi(Vbebaspinhafaat20thM“rm Yorethandearan

4. In

anThangodi(ICstaFeenmRethsa5%

reat examparth metal. ommon eleet only rarelyterference

he reason: “are anyway”eing used uring wasteiosolids will f Ce, which,

many targetiosolids meta

recent trairogram, a eveloped anine of the 1

V) results on tecause oackground

pectral correterference. ad not even

act that it is t the main w00.7. They hey “typicall

Mo”. The simrare” eleme

more commo

our auditor ecent single hen the assond review emonstrate re significann ICS.

nterferents shn ICV. he old CLP pnd (gulp!) ICood enoughifferent thanCV) standaandard withe at 200 pnough? An

mean accepeally? Isn’t han that? Thample! At t%....or 475

ple here is CCerium intements usingy have we stesting (or

“No one eve”. Not so trheavily in pewater treaincreasingly if uncorrectt analytes, als data.

ining on ICP challengend incorpora2 labs who this “challenof inapp

correction ection from We found tn incorpora clearly listedwavelength didn’t worry don’t enc

mple reality isents are beon in the env

is going to element int

ociated LOQthem to ssignificant in

nt, they need

hould be rec

protocol forCS-AB is + 20h. ICS stan

n an initial card…right?

h Al, Ca andppm and thn Al standardptance criteICP techno

hat’s broadehose levels, to 525 pp

Cerium (Cerferes with a

g 200.7 waveen any labverification)er asks for itrue anymorphosphorus atment. T

y contain higted, will inter

resulting in

P conductede PT sampated into the reported V

nge” PT sampropriately

points or m Molybden

hat a numbted Mo, ded as an inteidentified in

ry about it counter highs that even tecoming mvironment.

o ask for yoterference st

Qs for those see what nterferencesd to be test

covered as w

r evaluation 0% of true vandards are alibration veYou spike a Mg at 500 p

hen + 20% d at 500 pperia of + 1

ology more aer criteria tha we should e

pm for a 5

), a rare at least 12

elengths, b perform ) for Ce. t and it’s e. Ce is removal

Therefore gh levels rfere with n biased

d by our ple was e session. anadium ple failed

placed lack of

um (Mo) er of labs

espite the erference n method because levels of the more ore and

our most tudy and elements elements s. If they ted using

well as in

of ICS-A alue. Not really no

erification an ICS-A ppm and

is good pm would

00 ppm! accurate an any PT expect +

500 ppm

standardappropr So for recover

5. What’s

what is. The flawinterfereon recohow spshould Nare tryinon V is aabsent detecteinterferecorrectewhich issame asan initiamind thajust as te

6. Too man

interfereEveryonminimizestandardsamplesbillable”That’s multiplewe seecram eeases yanalyseworkloais particall the solution. Consideand Ti. analytesLOQ. SdetermiB is inadthe batproblem The problines at

d. That’s riate.

interferentsy should be

NOT presen w of the CLence check overies of spiked analytNOT be presng to verify tadequate, wfrom the sa

ed should ence that ed. The cons NOT presens the conce

al calibrationat overly neelling as ove

ny cooks spoents producee wants

e the analyds and s (i.e., ” samples)…why instea standard

e labs tryinverything in

your workloas standpointd from a daularly the cainterferent a.

er an ICS co Boron (B)

s, yet it is So the alarne that corr

dequate. Thattle. WHICH

matic?

blem here is249.677 and

not only re

spiked int + 5%.

nt is more im

LP protocol standards is

piked analytes impact sent in the shat your cor

wouldn’t youmple? Then be atthas not b

ncentration ont in an ICS

entration of tn blank (ICB)

gative concrly positive o

oil the broth e chaos. to ysis of

QC “non-

…right? ad of mixes,

ng to nto one samad from tht, but it only

ata quality stase when labanalytes into

ontaining ju is one of detected w

rm bells goection for inat’s great, bH interferent

s that B has td 249.772. Th

easonable,

to each IC

mportant th

for evaluati that it focuses rather thelements th

sample. If yrrection for Mu want V to n any V thatributable been propeof any analy should be tthat analyte. And keep

centrations aones.

and too ma

mple. Sure…he number increases yotandpoint. Tbs try to jumbo a single I

st Cr, Fe, M the unspikwell above

o off and yterferences ut it’s only ht correction

two prominehe 249.677 li

it’s

CS,

an

ng ses an

hat you Mo be t is to

erly yte the e in p in are

any

…it of

our This ble CS

Mo, ed its

you on

half n is

ent ine

hanm0.sapasuYowwminYesccoIC

7. AcfoWuninceshpoworcoth

8. ThinThlemdrThMthminbeSoesne Bounina nesocole

as interferenm below the.01-0.1 nm aafe, it may narticularly hi

uffers similar ou can

wavelength…wavelength more imp

terferences es, it’s comcheme foorrection ca

CP analysis w

cceptance or unspiked a

While in thnspiked ana practice tertainly, anyhould not beoint, the on

whether its LOr if there isorrection an

he affected a

he special nterferences.here are soegitimately inmight be a l

rinking watehere might

Mg as well ashe water ha

minimal Ca astead. But e expectedo what dostablished teeded?

oth 200.7 anequivocal) terference c responsibileeded. In olution mayommon ca

evels they typ

nce from Moe peak and above. Whilnot be if theigh. The othinterference

always …right? M

provide adeportantly, are there onmplex, and or evaluaan be challewas easy?

criteria cananalytes. heory estabalytes in an Ithat may noy unspiked e outside onus is on thOD (and thus a problemnd/or interfeanalyte.

case: My . ome labs wnterference ab that soleer from a pbe naturally

s bicarbonatas been softeand Mg, but

ultimately, td to be interoes a lab hat correct

nd 6010D ( indicate correction is lity to “prothis case a

y be one tions they pically see.

o and Cr wfrom Ti and le 0.1 nm m

e Fe concenher wavelene from Ti, Fe,

use Maybe. Do

equate LODwhat

n that wavel devising aating inteenging. But w

not exceed

blishing critCS should bot be realis analyte inf + LOQ. F

he lab to dus LOQ) is um with bac

erence corre

samples h

whose sam free. An ely analyzespublic watey occurring te and carbened, there a prevalenthese samplrference fre do if thetion factors

and 6010D that even

used, the laove” that an appropr

that contfind at the

within 0.05 Fe within

may seem ntration is ngth for B and Mo.

another oes that D? And potential ength?

a quality erference who said

the LOQ

teria for be + LOD, stic. But

n an ICS From that etermine

unrealistic ckground ection for

have no

ples are example s finished

er supply. Ca and onate. If may be ce of Na es would e…right? ey have are not

is pretty n if no

ab still has none is

riate ICS ains the

e highest

9. Do it daRun yosolutionsyou seejust will labs) dSure, thefor runnrepeatethe endefensibon your interfere

So that’s thdefensible can’t be mwant us toresolution, awhy the cmade ICS-solutions jusknow what about teveryone shICS-A, conand Al at and Fe at but there nAt least anlikely an ICSwell.

Tfois

D

ily for best reour series s (because that one ICnot cut it foaily for bese referenceing these le

edly that evend of the bility of each ability to doences are un

he recipe finterference

ore prescripo?) becauseand wavelenoncept of -A and (nst doesn’t m Judge Judythat! Pehould analytaining Ca,about 500 about 200

eeds to be n ICS-B and S-C and ICS

This document is inound in statute or as not finally determ

in litigation with Department of Natu

esults. of ICS

by now CS solution or 99% of st effect.

e methods hess frequentlerything is ion

day, the h’s days ana

ocument on nder control.

or cooking e check ptive (and woe every labngths are dipurchasing ope, not

make sensey says

erhaps yze an , Mg, ppm ppm, more. more S-D as

ntended solely as gadministrative rule minative of any of th

the State of Wiscoural Resources in an

NewLabNotes is p

have provisioy if you pron order. But

validity aalyses depeeach day th.

up a qualprotocol. Would you reab’s instrumefferent. Tha those readgoing ther….and we

uidance and does nare referenced. Thhe issues addressedonsin or the Departny manner address

and administrativ

Lawsletter of the Labopublished by the W

and Regist

SteveEnvironmental Sc

(608)

RicLabN(608)

ons ove at nd nd

hat

ity, We ally ent, at’s dy-re!) all

not include any mais guidance does nd. This guidance doment of Natural Resed by this guidancve rules to the relev

abNotes oratory Certification

Wisconsin DNR Labotration Program.

e Geis, Chief cience Services Sec) 266-0245

ck Mealy

Notes Editor ) 264-6006

andatory requiremenot establish or affeoes not create any esources. Any reguce will be made by vant facts.

n Program oratory Certification

ction

ents except where rect legal rights or o rights enforceablelatory decisions maapplying the gover

n

requirements obligations and e by any party ade by the rning statutes