Analytical method validation of GC-FID for the simultaneous measurement of hydrocarbons ãC 2-C 4 Q in their gas mixture

Se validó una cromatrografía de gases precisa. acoplada con un detector de ionización de llama Uõ%zâàçT para el análisis simultáneo de hidrocarburos ligeros U%0z%óT en su mezcla gaseosa@ Los parámetros de validación se evaluaron con base en la definición de la àSO: àã% 2DJ0í. que incluye selectividad del método. precisión y repetibilidad. exactitud. linealidad. límite de detección ULOçT. límite de cuantificación ULOQT y robustez@ 9ajo las condiciones analíticas óptimas. el análisis de la mezcla gaseosa mostró que cada analito de interés fue separado adecuadamente con alta selectividad@ Se encontró también que el método fue preciso y exacto7 la linealidad fue alta y con buen coeficiente de correlación lineal UR0 ≥ J@OOOT para todos los analitos@ Se puede concluir que el método õ%zâàç es confiable y apropiado para la determinación de hidrocarburos ligeros %0z% ó en una mezcla gaseosa@ ãl método validado ha sido exitosamente aplicado a la valoración de hidrocarburos ligeros %0z%ó en muestras de gas natural. mostrando alta repetibilidad con desviación estándar relativa URçST menor al 26 y buena selectividad sin interferencias de otros posibles componentes@ ;n accurate gas chromatography coupled to a flame ionization detector Uõ%zâàçT method was validated for the simultaneous analysis of light hydrocarbons U%0z%óT in their gas mixture@ The validation parameters were evaluated based on the àSO:àã% 2DJ0í definition including method selectivity. repeatability. accuracy. linearity. limit of detection ULOçT. limit of quantitation ULOQT. and ruggedness@ Under the optimum analytical conditions. the analysis of a gas mixture revealed that each target component was wellz separated with high selectivity property@ The method was also found to be precise and accurate@ The method linearity was found to be high with good correlation coefficient values UR ≥ J@OOOT for all target components@ àt can be concluded that the õ%zâàç developed method is reliable and suitable for determination of light %0z%ó hydrocarbons in their gas mixture@ The validated method was successfully applied to the estimation of light %0z%ó hydrocarbons in natural gas samples. showing high performance repeatability with relative standard deviation URSçT less than 2@J6 and good selectivity with no interference from other possible components@ âoi avaliada uma cromatografia gasosa precisa. equipada com um detector de ionização de chama U%õzâàçT para a análise simultâneo de hidrocarbonetos ligeiros U%0z%óT em uma mistura gasosa@ Os parâmetros de validação foram avaliados baseados na definição da àSO:àã% 2DJ0í. que inclui seletividade do método. precisão e repetibilidade. exatidão. linearidade. limite de detecção ULOçT. limite de quantificação ULOQT e robustez@ 9aixo as condições analíticas ótimas. a análise da mistura gasosa mostrou que cada analito foi separado adequadamente com alta seletividade@ Também foi encontrado que o método foi preciso e exato7 a linearidade foi alta e com bom coeficiente de correlação linear UR ≥J@OOOT para todos os analitos@ Podezse concluir que o método õ%zâàç é confiável e apropriado para a determinação de hidrocarbonetos ligeiros %0z%ó em uma mistura gasosa@ O método avaliado têm sido exitosamente aplicado à valoração de hidrocarbonetos ligeiros %0z%ó em amostras de gás natural mostrando alta repetibilidade com desviozpadrão relativo menor funcionais@ ao 26 e boa seletividade sem interferências de outros possíveis componentes@

Se validó una cromatrografía de gases precisa.acoplada con un detector de ionización de llama Uõ%zâàçT para el análisis simultáneo de hidrocarburos ligeros U% 0 z% ó T en su mezcla gaseosa@ Los parámetros de validación se evaluaron con base en la definición de la àSO: àã% 2DJ0í.que incluye selectividad del método.precisión y repetibilidad.exactitud.linealidad.límite de detección ULOçT.límite de cuantificación ULOQT y robustez@ 9ajo las condiciones analíticas óptimas.el análisis de la mezcla gaseosa mostró que cada analito de interés fue separado adecuadamente con alta selectividad@ Se encontró también que el método fue preciso y exacto7 la linealidad fue alta y con buen coeficiente de correlación lineal UR 0 ≥ J@OOOT para todos los analitos@ Se puede concluir que el método õ%zâàç es confiable y apropiado para la determinación de hidrocarburos ligeros % 0 z% ó en una mezcla gaseosa@ ãl método validado ha sido exitosamente aplicado a la valoración de hidrocarburos ligeros % 0 z% ó en muestras de gas natural.mostrando alta repetibilidad con desviación estándar relativa URçST menor al 26 y buena selectividad sin interferencias de otros posibles componentes@ ;n accurate gas chromatography coupled to a flame ionization detector Uõ%zâàçT method was validated for the simultaneous analysis of light hydrocarbons U% 0 z% ó T in their gas mixture@ The validation parameters were evaluated based on the àSO:àã% 2DJ0í definition including method selectivity.repeatability.accuracy.linearity.limit of detection ULOçT.limit of quantitation ULOQT.and ruggedness@ Under the optimum analytical conditions.the analysis of a gas mixture revealed that each target component was wellz separated with high selectivity property@ The method was also found to be precise and accurate@ The method linearity was found to be high with good correlation coefficient values UR 0 ≥ J@OOOT for all target components@ àt can be concluded that the õ%zâàç developed method is reliable and suitable for determination of light % 0 z% ó hydrocarbons in their gas mixture@ The validated method was successfully applied to the estimation of light % 0 z% ó hydrocarbons in natural gas samples.showing high performance repeatability with relative standard deviation URSçT less than 2@J6 and good selectivity with no interference from other possible components@ âoi avaliada uma cromatografia gasosa precisa.equipada com um detector de ionização de chama U%õzâàçT para a análise simultâneo de hidrocarbonetos ligeiros U% 0 z% ó T em uma mistura gasosa@ Os parâmetros de validação foram avaliados baseados na definição da àSO:àã% 2DJ0í.que inclui seletividade do método.precisão e repetibilidade.exatidão.linearidade.limite de detecção ULOçT.limite de quantificação ULOQT e robustez@ 9aixo as condições analíticas ótimas.a análise da mistura gasosa mostrou que cada analito foi separado adequadamente com alta seletividade@ Também foi encontrado que o método foi preciso e exato7 a linearidade foi alta e com bom coeficiente de correlação linear UR 0 ≥J@OOOT para todos os analitos@ Podezse concluir que o método õ%zâàç é confiável e apropriado para a determinação de hidrocarbonetos ligeiros % 0 z% ó em uma mistura gasosa@ O método avaliado têm sido exitosamente aplicado à valoração de hidrocarbonetos ligeiros % 0 z% ó em amostras de gás natural mostrando alta repetibilidade com desviozpadrão relativo menor funcionais@ ao 26 e boa seletividade sem interferências de outros possíveis componentes@ Introduction Non5methane hydrocarbons ONMH°s in short/ are typically low molecular weight O°45°4 D / species in the hydrocarbon chainq The NMH°s have become important in industry and environmentq In chemical industries2 some natural sources of NMH°s Osuch as methane2 propane2 and butane/ have more popular feedstock and their trading supply are highly demanded O1/q Furthermore2 NMH°s2 generated by anthropogenic activities2 Ofuel and biomass burning2 vehicles2 solvent usage2 and oil refineries/ have been detected in the atmosphere and they have grown environmental and public health concern O2, 3/q Regardless of their importance2 it is necessary to re5 assess measurement practice in order to provide accurate and reliable data of the NMH°s concentrationq This necessity is related to the fact that accurate and reliable data are used as the basis for decision making related to both for market price in industrial purpose and regulatory enforcement for the environmental monitoring programq 9ccording to ISOGIE°43DxL2 a reliable and accurate result can only be obtained by using a validated methodq In any testing laboratory2 method validation is a part of quality assurances to declare that a high quality of analytical result is provided O4/q In general2 method validation refers to a documented procedure used by a laboratory to assure that the method performance for the determination of a particular analyte meets the required criteria O5-7/q This paper presents results on the validation of a G°5FID method for the measurement of five components of light hydrocarbons O°x5°:/ including ethylene2 propane2 propylene2 isobutane2 and n5butane in their gas mixtureq The evaluation was based on the ISOGIE°43DxL definition O8, 9/ and it was emphasized on the following validation parameters8 method selectivity2 repeatability2 accuracy2 linearity2 limit of detection OLOD/2 limit of quantification OLOQ/2 and ruggednessq The validated method was successfully employed in the assay of light °x5°: hydrocarbons in natural gas samplesq

Gas chromatography analysis
9 certain volume of gas standard was injected into a G°system under optimized analytical conditions OTable x/q The output signal was monitored using G°°hemStation version Revq 9q4DqDx O43L3/2 which was installed on a LG personal computer OProcessor 9MD Richland 9:53zDD5HD j:3DD2 LG International °orpq/q The data was estimated by automated integration of the area under the resolved chromatographic profileq

Materials
°ertified gas standards for °x5°: hydrocarbon mixture OGS54 in short/ were purchased from Mesa Specialty Gases and Equipment O°92 US9/q 9 series of GS54 Odenoted as GS54a to GS54d/2 having concentration as listed in Table 42 was used as test standard in all experimental runsq 9nother GS Odenoted as GS5x/ was only used for method accuracy assessmentq Voth certified GS54 and GS5x with relative uncertainty xR are traceable to National Institute of Standards and Technology ONIST/2 US9q 7

Instrumentation
Separation of °x5°: hydrocarbons from their mixture was performed on a packed5column O°oated G°Packing xzR SP543DD2 jDG4DD °hromosorb P9W2 zD ft x 4Gj inch SS from Supelco/ installed on a G°system Model QjPD OHewlett Packard 9gilent2 °92 US9/2 equipped with a flame ionization detector OFID/q The optimized analytical conditions for the G°5FID method are tabulated OTable x/q  Procedure for method validation =ll data obtained from the GC;FID measurement were used for the method validation% The assessment parameters xselectivity1 repeatability1 linearity1 LOD1 LOQ1 and ruggedness: were calculated by adopting some procedures1 as they can be found everywhere in published literature x4, 5, 10-16:% In a typical experiment1 the calculation procedure is described as follows] selectivity of the method was determined by injecting the gas standard xGS;/: and it was evaluated in term of retention time xt R : and selectivity factor xα:V repeatability was established by measuring the response of the GS;/ standard and expressed as percentage relative standard deviation xO RSD: of seven replications injection under the same operating condition over a short time interval xin the same day:V accuracy was evaluated by comparing the concentration of GS;/ standard against another independent gas standard xGS;N:% Furthermore1 to investigate the linearity1 a series of GS;/ standard xGS;/a to GS;/d as listed in Table /: was used% The injection of each gas standard was conducted in seven replications and then the linearity was estimated from the calibration curve% The calibration curve was constructed by plotting peak area of each component in the GS;/ standard xGS;/a to GS;/d: as a function of their concentration% The LOD and LOQ were established at a signal; to;noise ratio xSRN: of 0 and /k1 respectively1 of the chromatogram at the lowest concentration point of each component% Ruggedness was evaluated by small changing in flow rate of carrier gas during analysis xfrom q%5 to H%5 mLRmin with k%5 mLRmin flow rate different as listed in Table H:% Selectivity1 repeatability1 accuracy1 linearity1 LOD and LOQ1 and ruggedness were defined as follows]

Selectivity
The selectivity refers to the capability of GC method to discriminate and quantify the response of target component in the presence of other components as interference x5, 10:% The selectivity is the relative retention of two adjacent peaksV hence1 it is highly dependent on the change of the t R values of the two corresponding target gas components%

Repeatability
The repeatability precision of method refers to the closeness between measured values resulting from an independent measurement using the same equipment1 under the same analytical condition1 by the same operator and within short intervals of time x5, 10:% Theoretically1 the determination of the repeatability was conducted by the prediction of relative standard deviation xORSD: of precision using Horwitz function [/] x14:] where C is the concentration of gas component stated in decimal fraction% The requirement of ORSD for repeatability is between k%5 and k%H5 of a theoretical value determined by Horwitz function% In a word1 the repeatability of the method is categorized acceptable when the ORSD is less than k%qH of the OCV; Horwitz xk%qHCV;Horwitz: x7:% Accuracy Method accuracy refers to the closeness of agreement between measured and accepted xtrue: concentration of target component% The accuracy value is dependent on two factors i%e%1 the bias and precision% The bias of a method is the difference between the measured value and the value from certificate of reference standard1 which was calculated using an expression below [N] x4, 6, 7:] where is the average of measured reference standard value1 and  is value from certificate of reference standard% For assessing the method accuracy1 precision of an analytical method xσ: from repeatability and reproducibility is included% In addition1 the uncertainty value from certificate of reference standard is also included for estimating the σ value% Thus1 the value of σ is obtained by combining those three components by using the following expression [0] x6, 7:] where S b is the standard deviation from reproducibility xinter day precision:V S w is the standard deviation from repeatability xintra day precision:1 and   is the uncertainty of standard GS;N as stated in the certificate% The acceptance criteria is set according to the ISO Guide 00]Nkkk x15:% In such ISO Guide1 no bias is found if the observed bias xC∆: value falls within ± N σ at confidence level [5O [6]] Linearity Method linearity is defined as the ability of the method to demonstrate that the test results are proportional to the concentration of sample x5, 7, 10:% Investigation of method linearity for C N ;C 6 measurement was conducted by generating a calibration curve using different concentration levels of C N ;C 6 gas standards% Each concentration level was analyzed using GC;FID in six replications xn ± q:% LOD and LOQ LOD of an analytical method refers to the lowest amount of analyte that can be detected which is not necessarily quantified as an exact value% Meanwhile1 LOQ is the lowest concentration of an analyte that can be quantitatively determined with appropriate precision x5, 10:% In a GC measurement1 both LOD and LOQ are important% The LOD and LOQ were calculated based on signal to noise ratio1 which are 0]/ and /k]/1 respectively%

Ruggedness
The ruggedness of an analytical method is the method capacity to generate some results which remains unaffected by minor changes of the experimental conditions during analysis x5, 10:% In this study1 the ruggedness of the method was assessed by investigating the effect of small change on the flow rate of the carrier gas used as the mobile phase for gas component separation in the column of GC;FID system%  − Horwitz (%) = 2 (1−0.5 log ) [/] Two natural gas samples were obtained commercially from Indonesian State Oil and Natural Gas Mining Company ,PERTAMINA3 located in North of Jakarta5 The natural gas samples were analyzed by the validated method without any special treatment5

Accuracy
From the Table q7 it can be observed that the measured values of all C 1 FC 0 hydrocarbons fall within ± 1σ/ thus7 it can be concluded that no evidence of bias can be found in the analytical method used under this study5

Selectivity
As discussed above7 no interfering peak of one gas component relative to others could be observed ,Figure ;37 resulting in excellent selectivity factor ,α3 ,Table x3 with α values larger than ;5± ,1335

Assay of hydrocarbon in natural gas samples
Practically7 in every method validation process for a GC technique7 performing development of the method is the initial step7 which can be carried out by optimizing the conditions of the GC for the measurement of the target component5 Figure ; depicts a typical chromatogram of C 1 FC 0 gas component obtained under optimum analytical conditions of the GCFFID instrument ,Table 135 As can be seen from Figure ;7 all gas components were well separated with their retention times ,t R 3 as listed in Table x5 No other interference peaks could be found7 indicating that the development of the GC method was achieved successfully ,1235 Thus7 the method validation process could be conducted5

Repeatability
From the calculation result7 the . of CVF Horwitz for each individual gas component as found to be less than ±5D% CVFHorwitz as listed in

Linearity
The linearity data of the method are listed in Table GI As it can be seen from Table GQ excellent linearity was obtained for all gas components with correlation coefficient values FR 5 L equal or greater than /IµµµI ThusQ the method may fit for purpose for the determination of C 5 ;C 8 gas in their mixtureI

Assay of natural gas samples
The validated method was applied for the analysis of light hydrocarbons FC 5 ;C 8 L in two natural gas samplesI The primary analysis results indicated that the concentration of the target components FC 5 ;C 8 L in the natural gas sample was higher than the linear concentration range of the standard gas mixture FTable GLI This implies that a dilution step is requiredI ThereforeQ the natural gas samples were then properly diluted by using ultra high pure helium FµµIµµµ9 purityL with a dilution factor of GQ and the final concentration is shown in Table SI It can be seen from Table S that all the target components FC 5 ;C 8 L in the natural gas samples were detected and found at high concentrationQ except for propyleneI Propylene may also exist in the natural gas sample but it cannot be detected by the GC;FID system under the experimental condition of this studyI In additionQ Figure 5 displays a typical chromatogram of a natural gas sample after the analysis using the validated methodI Chromatogram in Figure 5 indicates that the method was selective for the analysis of C 5 ;C 8 Q and no interference from other components could be observedI

Ruggedness
As can be seen in   Note: The GS-2 was used for producing the data with concentration as listed in Table 1.

Conclusions
In this study5 the developed UJTzIR method for the analysis of the J ' TJ " hydrocarbons /including ethylene5 propane5 propylene5 isobutane5 and nTbutaneE provides good selectivity toward separation of individual gas components from their mixtureH Moreover5 the results of each validation parameter5 based on the ISOkIqJ PAL'"5 indicated that the validated method provides a sufficient evidence for proving a reliable UJTzIR method for the measurement of J ' TJ " hydrocarbon in their gas mixtureH The developed and validated method could also be extended to the analysis of real natural gas samplesH #ence5 the use of such validated method may keep the degree of user confidence regarding their analytical dataH

Figure 1 .
Figure 1.A typical chromatogram of C 2 -C 4 under optimized analytical conditions, showing a good separation property.

Table 2 .
Optimized analytical conditions of the GC-FID.

Table 3 .
Table 07 indicating that the method is repeatable5 Retention time and selectivity factor.

Table 5 .
Accuracy data of the GC-FID for the measurement of C 2 -C 4 in their mixture.

Table 6 .
Data indicating linearity of the method for all gas component and theirLOD and  LOQ values (n = 6).For any quantification process producing a value below the LOD and LOQ level may lead to yield in a high measurement uncertainty+ thus an unreliable measurement would occurI In additionQ practicallyQ the LOD and LOQ assessment are equally important in comparison to other method validation parametersI At a LOD levelQ only qualitative analysis is possible to be evaluatedQ while at a LOQ levelQ both quantitative and qualitative analysis are possibleI HoweverQ at the LOQ levelQ the quantitative analysis performed may produce inaccuracy and imprecise resultQ leading to a high uncertainty contribution on the final analytical results F8LI Table G tabulates the LOD and LOQ values for all gas componentsI As it can be seen in TableGQthe lowest LOD value was found to be 5k/ μmolHmol for ethyleneQ and the highest LOD values was found to be kk/ μmolHmol for propyleneI CorrespondinglyQ the LOQ of ethylene Fµ1/ μmolH molL and propylene F5PG/ μmolHmolL were found to have the same trend as the lowest and the highest LOQ valueQ respectivelyI Table kQ in all flow rate levelsQ both retention time and percentage peak area were found to be within acceptable limit with very low standard deviation FSDLI ThusQ small changes on the GC;FID experimental conditions in term of flow rate variation did not have any effect on the results of analytical measurementI AlthoughQ a massive change on the flow rate level of carrier gas has been reported to significantly affect theresults of a GC measurement F16LI This is a reasonably accepted finding because the concentration of ethylene F/I%S 9 molHmolL and propylene F1%I11 9 molHmolLQ as the lowest and the highest concentration among all other components FTable GLQ respectivelyQ were used as the basis for calculating the LODHLOQI Since the value of LODHLOQ obtained from an analytical measurement is generally concentration dependent+ thereforeQ the value of LODHLOQ could be decreased by decreasing the concentration of the component used for LODHLOQ calculationI

Table 7 .
Results of the ruggedness study.

Table 8 .
The final concentration (, mol/mol) of C 2 -C 4 in natural gas samples a n = 3 (triplicate) and b ,RSD.