5. 5
Thenhit“OK” for the nextwindow.The nextwindowislabeled“MSTune File”.In thisexperiment,the
file “ncich4.u”wasused.Click“OK”for the nexttwowindows.Forthe thirdmenu,the programistrying
to finda printerto print the information.Becausedeltamethrinresultsneedtobe manuallyquantitated,
any optionisfine. Click“Select”.Thenthe programgivesthe window labeled“Save MethodAs”.If thisis
justa verificationof method,nothingneedstobe changedandit isrecommendedtosimplycancel out
of the windowanddon’tsave.If the methodfile name isn’t changed,the changesdone tothe methods
will overwritethe previousmethod.If the methodfilename ischanged,thenitisokayto justsave but
knowa newfile withwhateverchangesmade willbe creates.
In the lab,a PAL auto samplersystemwasusedwhichworksexternallyandindependentlyfromthe GC.
The purpose of the PALsystemisnecessarytorun multiple sampleswithouthavingtomanuallyinject
samplesintothe inlet.The methodof the PALautosamplerusedinthe procedure waslabeled under
“MethodName”as “DG Dir Inj Pyrethroids”(Figure 5) andmustbe consistentwiththe sequence used
for the EnhancedChemStation.The systeminjects1.5 µL of the sample torun inthe GC.
The GC isnow readyto use.
Figure 5: PAL MethodFile Settings
6. 6
Procedure 2: Interpretingdata
Step1: LoadingGraph
Load up EnhancedData Analysis (Figure 6).Thiswill bringupadifferentwindow thanthe previous,but
isstill EnhancedChemStation.
File > Load data file…
Thiswill bringupthe windowinFigure 7.Path showsfile director. The box below Path showsdifferent
file namessavedinside thatfolder.Clickinganoptionhasa preview of the graphas well assome simple
information. Generally,aLOQ or LLOQ shouldbe chosenfirstbecause those peakscouldbe more
difficulttosee withinthe backgroundnoise.
Step2: Setup ParametersforIntegration
Find“Chromatogram”inthe dropdownmenus
Chromatogram> SelectIntegrator…>“RTE Integrator”
Chromatogram> MS Signal IntegrationParameters…
Thiswill bringupa windowlabeled“RTEIntegratorParameters” showninFigure 8.Most of thiswindow
can stay default.The aspectsthatare generallychanged inordertocreate a more accurate integration
Figure 6: Icon forEnhancedData Analysis
Figure 7: Data File Window
7. 7
are “Start threshold”,“Maximumnumberof peaks”,and“Baseline Preference”.Note thatchangesin
the parametersmustremainconsistentthroughoutadata setexceptfor“Maximumnumberof peaks”.
The followingisafewnotesof whateach option’sfunction is:
Start thresholdcanbe loweredtomake the peakstobe identifiedbythe program.The thresholdis
mostimportantwhenthe programis unable tointegrate the smallerpeaksfromthe backgroundnoise
because of howclose the baseline istothe peak.Thisnumberisa 0.200 default,butcanbe changed
from0.200 to 0.001.
Baseline Preference (Figure 9) isrelevantdependingonthe slopesof the graphespeciallyif thereare
twoor more peaks.The baseline isthe generallineof where the backgroundnoisefluctuatesaround.
Whenusing“Baseline dropelse Tangent”,the integrate functionwill draw aline fromthe baseline
where the peakstartsexactly horizontallytowardsthe nextpeakorthe baseline.Whenencountering
anotherpeak,the otherpeakwill be consideredseparate whenintegratingsothe areaof integration
will endwithastraightline upwardstowhere the peaksmeetandseparate the areaswiththe basisthat
the area of frombelowthe peakall the wayto the baseline.The “TangentelseBaselinedrop” option
functionsdifferently.Insteadof assumingwhere the baseline iscreatingintegrationoff of the peak
above the baseline,the programsimplydrawsabestfitline fromwhere the peakisstartingtowhere it
stops.Thismeansthat these integration numbersare lessthanthe “Baselinedropelse Tangent”.
Figure 8: RTE IntegratorParameters
8. 8
Maximumnumberof peaksisresourceful whenthe backgroundnoise isinterferingwiththe signalof
the peaksdesired.Thisfunctiondoesnotchange anythingintermsof the functionof integration,but
insteadshowsmore peaks.Thisiswhythisaspectcan be changedbetweengraphsof the same data set.
Step3: IntegratingaPeak
A fewnotesonhowto maneuveraroundthe graph. A left-click-hold-and-dragwill zoominonthe box
created. A double leftclickwill zoomoutthe graphto the previouszoomview used.
Rightclickcan be chosento be eitherManual Integration orMS graph view (Figure 9).Thisisdone by
findingthe “Tools”dropdownmenu.Tools> Options…will bringupa“SelectDA Options”and here
“Manual Integration”canbe checkedto change the rightclick,or unselectedtoshow aMS graph of
whereverselectedonthe graph. Whenusingthe Manual integration,a right-click-hold-and-drag
betweenlineswill findthe areaof the curve above towardsthe peakwhere the line thatwascreatedis.
Whennot usingManual integration,avertical line will presidewhere the mousecursoris.A double-
right-clickbringsupa graphunderneaththe initial graphtoshow the Mass Spectroscopypoints (Figure
10).
Figure 9: Tangentelse BaselinedropVersusBaselinedropelse Tangent
Tangentelse Baselinedrop Baseline dropelse Tangent
10. 10
To integrate apeak,simplyuse the integrate button.Thisbuttonisshown inFigure 11highlightedin
red. Note that thiscan alsobe done throughthe Chromatogramdropdownmenu.
Chromatogram> Integrate
It is possible thatinthe firstintegrationof lowerconcentrationgraphsmaynotresultinidentificationof
the peaksdesired.If thisoccurs,referbackto Step2 to go back intothe parametersof integrationto
change the “Start Threshold”or“MaximumNumberof Peaks”.Make sure that if “Start Threshold”is
changed.Do notchange againwhile usingthe same setof tests. Forexample,the settingsusedforthe
CalibrationandQCshouldalsobe usedfor the validationof samples.
Step4: ObtainResultsof Integration
If integrationof the peakissuccessful,the areaof specificpeaksisneededforfuture calculations.Thisis
done throughthe “Chromatogram”drop downmenu. Anexample of the formattingforthis information
isshownin Table 1.
Chromatogram> Integrate Results
Figure 11: Integrate Button
Table 1: Example of PeakIntegrationResults
11. 11
Thisinformationisto be copiedintoMicrosoftExcel foreach concentrationpoint.
Step5: OrganizingResultsof Integration
Switchback andforth fromEnhanced ChemStationandMicrosoftExcel totransferoverthe integration
resultslike shownabove.Onlythe peaksdesiredshouldbe kept. Thesepeaksare the Cis-permethrin,
and the twodeltamethrinpeaks. Delete the linesof peakswhichwere pickedupbythe programthat
are notthe desiredpeaks. Forexample,if the maximumnumberof peaksissetto4, the program will
find4 peaksno matterwhat.There are onlythree desiredpeaksthough,butthe otherpeakwill be
automaticallyselectedandneedstobe deleted. Anexampleof how adata set lookslike isshown in
Table 1.
Step6: CreatingCalibrationCurve
To create a calibrationcurve,achart of data needstobe made.Anexample of thischartis shown in
Table 3 withthe calibrationgraph.Thischart contains concentrationlevel,response ratio(RR),the cis-
permethrinpeakarea,deltamethrin1and 2 peakareas,the calculatedng/mL,andthe calculated
percenterror. The response ratio,ng/mL,andpercenterrorare all calculatedthroughexcel.
The response ratioiscalculatedwiththe equation:
𝑐𝑖𝑠−𝑃
(𝐷𝐿𝑀+𝐷𝐿𝑀2)
. Thatis cis-permethrinpeakareadivided
by the sumof the deltamethrinpeakareas. Thisequationcanbe formattedintoExcel forconvenience.
Aftercalculatingthe RR,the calibrationcurve isneededtoobtainthe y-interceptandslope tocalculate
ng/mLand percenterror.To calculate a weighted curve,GraphpadPrismwasusedinorder give the
curve a weightof
1
𝑥2
.
Weighted Calibration Curvein Graphpad Prism
AfteropeningupPrism (Figure 12),starta new project.
File > NewProject…
Figure 12: PrismIcon
12. 12
The settingscan be leftasdefault.Thiswill bringupablankdata table. Take the firsttwo columnsof the
excel chart,whichshouldbe the concentrationspointsandthe RR,andcopy themontothe data table
on the firsttwocolumns. Make sure to omitthe zeroconcentrationasthiswill interfere withthe
weightingfeature of thisprogram. Table 2 showshow the data shouldbe formatted.
Nextfind“Analyze”ina small box above the mainscreenandclickon it (highlightedinredinFigure 13).
Thiswill bringupa windowlabeled“AnalyzeData”.
In the “Type:”box,select“Curves&Regression”andselect“Nonlinearregression(curve fit)”inthe box
immediatelytothe right.All otheroptionscanstaydefault.Afterclicking“Okay”,anew screenlabeled
“Parameters:Nonlinearregression”will appearlike inFigure 14.Inthe “Choose anequation”box find
the “Polynomial:FirstOrder(straightline)”option.Nextin“Options”underneaththe previousbox,find
Table 2: PrismData Table
Figure 13: PrismAnalyze Button
13. 13
the “Method…” button.Whenclickingonthisbutton,the programwill give options fora“Weighting
method”alsoshowninFigure 13. Select“Weightby1/𝑋2, thenclick“OK”.
Thiswill bringyouto yourresultspage showninFigure 15. Copyall of the valuesfromthisresultspage
intothe Excel page. The graph can also beenseen inthe side toolbarbyclickingonthe “Graphs”folder
and selectingthe file withthe graph (Figure 16).Inthe results,the valuesdesiredare the bestfitvalues.
“A” representsthe y-interceptand“B” representsthe slope.
Figure 14: WindowforNonlinearregression (left)andMethodswindow toweighdata (right).
15. 15
Unweighted Calibration Curvein MicrosoftExcel
An unweightcalibrationcurve (Table3) can be createdin Microsoft Excel easilytogive ageneral ideaof
the plotof where pointslie.The unweightedcurve inExcel ismore influencedbyhigherXvaluesmaking
it lessaccurate. To create thiscurve,select the firsttwocolumnslike beforeexceptthe zerovalue can
be selected.Gotothe “Insert”tab of the menuandselect“Scatter”inthe groupof “Charts”.When
clicking“Scatter”a drop downof a fewscatterplotcharts will appear.Selectthe firstone graph.This
will create ascatteredplotof the differentdatapointsselected.Highlightone of these datapointsand
thenright-clickthe pointinordertobringup a drop downmenu.Select“AddTrendline…”tobringup
the “Format Trendline”window.Select “Linear”andlooktowardsthe bottomat the final three options.
Checkthe final twooptionslabeled“DisplayEquationonchart” and“DisplayR-squaredvalue onchart”.
Thenhit“Close”.Thiswill complete the unweightedcalibrationcurve forMicrosoftExcel.Forthis
experiment,these valuesare notused. The unweightedcurve made inExcel issimplyto reference to
see where the pointslie onthe graph.
HighlightDataPoints> Insert> Scatter > FirstGraph > Right-ClickaData Point> AddTrendline >Linear>
DisplayEquationonchart and DisplayR-squaredvalueonchart
Table 3: Excel Cal and QC Graph andChart
16. 16
Step7: FindingPercentError
The ng/mL iscalculatedthrough usingthe response ratio,y-intercept,andslope (Table4).The equation
is
𝑛𝑔
𝑚𝐿
=
(𝑅𝑅−𝑖𝑛𝑡𝑒𝑟𝑐𝑒𝑝𝑡)
𝑠𝑙𝑜𝑝𝑒
. In Excel,these calculationscanbe done easilybysettingupanequation.Inorder
to make Excel pull fromthe same cell whenworkingwithmore thanone setof points,use the “$” sign
inbetweenthe rowandcolumnvalue.
To calculate percenterror,the ng/mLvalue isneeded.The equationis
𝑝𝑒𝑟𝑐𝑒𝑛𝑡 𝑒𝑟𝑟𝑜𝑟 =
𝑛𝑔/𝑚𝐿 𝑎𝑐𝑡𝑢𝑎𝑙 − 𝑛𝑔/𝑚𝐿 𝑒𝑥𝑝𝑒𝑐𝑡𝑒𝑑
𝑛𝑔/𝑚𝐿𝑒𝑥𝑝𝑒𝑐𝑡𝑒𝑑
∗ 100%
The “𝑛𝑔/𝑚𝐿 𝑒𝑥𝑝𝑒𝑐𝑡𝑒𝑑” value inthe equationrepresentsthe concentrationvalue inthe columnbefore
RR. Thisconcentrationvalue isthe expectedvalue tobe measuredfromthe sample.Thisequation
comparesthe percentdifference betweenthe ng/mLmeasuredandthe ng/mLthatis expected.Referto
Table 3 fora viewof a completedCal andQCtable.
Perthe Universityof GeorgiaPhysical Pharmacy requirements,the percenterrorhastobe lessthan
±25% for the resultstobe consideredviable.If anon-range definingpointhasa percenterrorover25%,
the pointcan be takenoutand the overall percenterrorcan be recalculatedwithoutthe point.
Procedure 3: Validation
Afterrunninga Cal and QC,validationof specificconcentrationsforsamples isnecessary. Validation
informationisaverysimilartorunningCal and QC exceptthere are more factorsto calculate thanjust
percenterrorand ng/mL.A validationrunwill generallyinclude fourgroupsof concentrationpoints:
LOQ, LQC, MQC, and HQC. Each of these groupshas six samples,fourof whichhave topass.Validation
requiresdifferentfactorstopass.It adds an average ng/mL,a standarddeviation,accuracy,anda
precisionfactor.Accuracyand precision bothhave tobe lessthan25% to pass the criteriaasacceptable.
Calculating “avg ng”
Take the average of the entire ng/mLcolumn.Thiscan be done easilyinExcel withthe “AVERAGE”
functionandhighlightingthe portionof the column.
Table 4: Findingng/mLfromRR,y-interceptandslope
17. 17
Calculating “stdev”
Take the standarddeviationof the entire ng/mLcolumn.Thiscanbe done easilyinExcel withthe
“STDEV” function.
Calculating “accuracy”
Take the average of the “%error” column.Again,thiscanbe done withthe “AVERAGE”functionand
highlightingthe portionof the column.
Calculating “precision”
Precisioniscalculatedbyusingthe “stdev”and“avgng”. The formula:
𝑝𝑟𝑒𝑐𝑖𝑠𝑖𝑜𝑛 =
𝑠𝑡𝑑𝑒𝑣
𝑎𝑣𝑔𝑛𝑔
∗ 100%
19. 19
References:
All informationwastaught byDr. DarrenGullick,anddictatedandreported byJohnTang. Resources
providedbyDr.Michael Bartlettin the Universityof GeorgiaCollege of Pharmacy. All figures,tables,and
pictureswere takenonthe GC1 computerinthe Physical Pharmacylabat the Universityof Georgia,
College of Pharmacy.