97 views

Uploaded by Subhamoy Sikder

save

You are on page 1of 10

com

**Food Chemistry 109 (2008) 815–824
**

www.elsevier.com/locate/foodchem

Analytical Methods

**Uncertainty estimation and in-house method validation of
**

HPLC analysis of carotenoids for food composition data production

M. Gracßa Dias a,*, M. Filomena G.F.C. Camões b, Luı́sa Oliveira a

a

Laboratório de Bromatologia e Nutricßão, Centro de Segurancßa Alimentar e Nutricßão, Instituto Nacional de Saúde Dr. Ricardo Jorge

(INSA, I.P.), Av. Padre Cruz, 1649-016 Lisboa, Portugal

b

CCMM/DQB – Faculdade de Ciências da Universidade de Lisboa, 1749-016 Lisboa, Portugal

Received 23 March 2007; received in revised form 4 December 2007; accepted 12 December 2007

Abstract

The method for separation and quantitative determination of the main carotenoids in food by high-performance liquid chromatog-

raphy (HPLC) was in-house validated. Tomato (Lycopersicon esculentum M.) as food matrix was used to demonstrate its linearity,

repeatability, intermediate precision, detection and quantiﬁcation limits, sensitivity and bias. In addition, stability of carotenoids was

studied in function of temperature and time. Method accuracy was quantiﬁed through measurement uncertainties estimate based on this

validation study. Furthermore, a study was conducted to evaluate variability coming from location in an experimental ﬁeld composed by

12 subﬁelds. The use of two metal free reverse phase columns and an organic mobile phase based on acetonitrile, methanol and dichlo-

romethane enabled the separation of the six target compounds (trans-a-carotene, trans-b-carotene, b-cryptoxanthin, all-lycopene, lutein,

zeaxanthin) within a 30 min run; detection at 450 nm and external calibration allowed the quantiﬁcation of the analytes. Carotenoids

concentration and measurement uncertainty, in mg/100 g, in tomato varieties ‘‘lido” and ‘‘for salad” were, respectively, 1.0 ± 0.14

and 0.39 ± 0.056 for trans-b-carotene, 8 ± 2.0 and 2.3 ± 0.57 for all-lycopene and 0.10 ± 0.017 and 0.08 ± 0.015 for lutein; trans-a-car-

otene, b-cryptoxanthin and zeaxanthin were not detected in both varieties (detection limits, in lg/100 g, 0.81, 0.57 and 0.77, respectively).

For b-carotene and lutein, uncertainty associated with the entire process including small-scale within-region variation was statistically

diﬀerent, at a signiﬁcance level of 5%, from measurement uncertainty (which includes sampling in the laboratory).

Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: Carotenoids; HPLC; Validation; Tomato; Analytical and pre-analytical uncertainties

**1. Introduction jugated double bonds imparts them beautiful colours from
**

yellow to red.

Carotenoids are a class of hydrocarbons (carotenes) and These natural pigments are present in fruits and vegeta-

their oxygenated derivatives (xanthophylls) consisting of bles and are essential components of plant photosynthetic

eight isoprenoid units joined in such a manner that the apparatus. In human health their importance is related

arrangement of isoprenoid units is reversed at the centre not only to their provitamin A activity but also to their

of the molecule. All carotenoids may be formally derived antioxidant action (Riso, Visioli, Erba, Testolin, & Porrini,

from the acyclic C40H56 structure having a long central 2004; Weisburger, 2002). In Europe the predominant

chain of conjugated double bonds, by hydrogenation, carotenoids, both in food and serum, are a-carotene, b-car-

dehydrogenation, cyclization or oxidation or any combina- otene, b-cryptoxanthin, lycopene, lutein and zeaxanthin

tion of these processes (IUPAC, 1974). Such system of con- (Granado, Olmedilla, Blanco, & Rojas-Hidalgo, 1996;

Olmedilla, Granado, Blanco, & Rojas-Hidalgo, 1992).

*

Corresponding author. Tel.: +351 21 7526485; fax: +351 21 7508153. Among carotenoids, lycopene probably has the highest

E-mail address: m.graca.dias@insa.min-saude.pt (M.G. Dias). antioxidant capacity (Di Mascio, Kaiser, & Sies, 1989).

**0308-8146/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
**

doi:10.1016/j.foodchem.2007.12.026

their content generally varies largely of the same method and thus contributing to uncertainty. a priori. mainly because of the large base is one of the more demanding and diﬃcult aspects number of these compounds chemically very much alike of database preparation (Greenﬁeld & Southgate. Scott. tomato and . this study was way which involves the exhaustive division of the analytical planned aiming at contributing to the implementation of method into its steps to estimate all components separately an HPLC method to produce analytical data on carote- and their subsequent combination applying the law of noids composition of Portuguese fruits and vegetables to propagation of uncertainties. be included in food composition tables.g.) that lack descriptive models of the behaviour of the analyte in the analytical system. the be overcome estimating trueness indirectly by determining reliability of a substantial part of current data on food the amount of an added spike recovered from a sample carotenoids still appears questionable (Rodriguez-Amaya. with variety. it may be tion using a tomato food matrix. may not include contri. and trueness. 1995. uncertainties obtained either way should be similar. Finglas. occurring in food (around 60) and their susceptibility to Users of compositional databases require representative isomerization and oxidation during analysis (carotenoids values for the composition of the foods consumed by the are sensitive to light. evaluating measurement uncertainty and transfer steps. more valid than Camões. like price. contributing to sampling planning. 2004). 2003). determines On the other hand. evap. butions from all sources when estimating the variation. etc. soil and light intensity. a mathematical As an outcome of the arguments above. & Froidmont-Görtz. To deal with method high time and money con- overall performance of the method. variation of carotenoids content was evaluated in a tomato ods. Ideally. To our knowl. 1999). as in the present case (e. The guide was inter. measurement material (lyophilized mixture of sweet corn. 2000. namely. Quan- tary lycopene and also an important source of b-carotene tiﬁcation of the trueness (bias) and its associated measur- (Krinsky & Johnson. Silva. Hart. Dias et al.816 M. studying sample and stan- extremely complex when analytical methods include mass dards stability. Seale. the need for their analysis and the inclusion of this infor. analyte concentration) within each species. oration. & (also named bottom-up) is not. might vary in subsequent applications (in samples with dif- itative and quantitatively from food to food and that. primary objective in sampling is. of course. by in-house valida- ful tool (Analytical Methods Committee. The Guide to the expression of Uncertainty to be expected and should not be confused with variations in Measurement issued by the International Organization associated with the analytical conditions (Greenﬁeld & for Standardization (ISO. Although it is a very power. may be to document this variability as it relates to factors edge. / Food Chemistry 109 (2008) 815–824 Tomato and tomato products are the main sources of die. Small-scale within-region obtained from in-house validation of the analytical meth. 2002). considering the intermediate precision parameters obtained during laboratory validation studies. evaluating and expressing uncertainty. diﬃculties related to CRMs. analytes and concentration can mation on food composition databases. carotenoids instability and HPLC prone faults. 2003). suming. oxygen and active population for whom the database is being prepared. Method trueness was quantiﬁed using a certiﬁed reference nents (Eurolab. maturity. a secondary objective with the measurement uncertainty estimate. temperature. Sampling foods for inclusion in a compositional data- ute to this lack of reliability. extraction. a second possibility that involves the treatment of interlab. this approach has not yet been done to carotenoids such as season. On the other hand the ﬁrst approach approach (EURACHEM/CITAC. risking ‘‘double counting” when others contributions are Measurement uncertainty was estimated by a transversal studied separately. ment uncertainity with certiﬁed reference materials The possible relation between the consumption of fruits (CRMs). geography and cultivar. to collect food For consistent interpretation of an analytical method samples that are representative ensuring that changes in result it is necessary to evaluate the conﬁdence that can composition do not take place between collection and anal- be placed in it. matrix (Barwick & Ellison. 1995). stability studies were done with samples and standards. ferent composition. 2005). includes all current contributions to overall bias and vegetables containing carotenoids and the human for a particular study but it is rarely clear which of these health (Cooper. Such variations are determination.G. therefore. 2006) based on analytical method in-house valida- a process that makes use of ‘‘whole method” performance tion data. 2. limited number of matrices. This approach involving experiments that study the variety. Nevertheless. 2000). 1996) oratory information and also the use of information using tomato as food matrix. Santos. homogeneity. 1995) establishes the rules for Southgate. Since all foods are biological materials and exhibit of its accuracy (trueness and precision) combining bias natural variations in composition. but in the second Studies were carried out to in-house validating an ana- edition (EURACHEM/CITAC. Materials and methods preted for analytical chemistry by EURACHEM. additionally uncertainties coming from vari- experience shows that uncertainty estimations obtained ables not covered by in-house validation studies were esti- by the mathematical analytical approach are often too mated separately through the bottom-up approach. small due to a tendency to omit some signiﬁcant compo. which originally presented only this approach. The surfaces). the fact that carotenoids vary qual. this can be provided by the quantiﬁcation ysis. Carotenoid analysis inherent diﬃculty may contrib. 2000) it already includes lytical method based on literature (Hart & Scott.

ane. harvested in September. injection volume 50 lL and stop time organic plants starters grown in a farm at the central/south 30 min. were solution of lutein.5 mL mobile phase without DCM. The 12 lots according to literature (Britton. no. 3450 (472 nm). IKA. 2550 (445 nm). The col- surface was removed. Empower software.05 M ammonium acetate):dichlorometh- 21 °C. containing 0. assuming a behavioural similarity with the two sublots of quarters). Each lot composite sample was substances under study. Samples and sampling tem. 2460 (451 nm). glass was used when feasible and extracts were always ad”) and identical maturation state was bought at a super. manipulated under yellow light and within the shortest market.1% BHT and 0. Lido) samples produced from 1. (Grindomix tem. lycopene immediately analysed in duplicate. 2800 (444 nm). the other 9 were frozen at trile:methanol (0. b-apo-80 -carotenal. Auto Sampler. b-apo-80 -carotenal was each tomato in a lot was cut longitudinally in quarters and used as an internal standard to control the analytical pro- two parts from opposite sides were ground and homoge. Stock solutions were pre- pared according to literature (Hart & Scott. (each time with 50 mL of a mixture 1:1. a-Carotene. 1 kg of tomato (ﬁrm homogeneous ripe fruits) of cover of ordinary artiﬁcial light and direct sunlight. The residue was carefully dissolved in carotenoid. b-cryptoxanthin. was evaporated at 40 °C in a rotary evaporator. ether ether ketone) lined (Alltech. at 440 nm and 0. b-apo-80 -carotenal. A Waters HPLC sys- group was immediately grinded in a knife mill. ﬁltered through a 0. Sample peak purity was the same person with the same criteria and kept refriger. based on peak area of the . All manipulations were carried out under the studies. intermediate precision and stability was made. an adequate dilution For repeatability. were analysed. was used. Germany) and vacuum–ﬁltration. b-carotene. b-car- high-performance homogeniser (ultra-turraxÒ. b- apo-80 -carotenal from Fluka Chemicals. 9 were ana.521 of absorbance. all liquid concentrations were determined with a Hitachi U-2000 chemicals for HPLC were HPLC grade. cat. Each tomato was then cut in two parts and separated reverse phase HPLC with photodiodes detection and chro- in two groups. amber the same variety (Lycopersicon esculentum M. the remaining sample was grinded as umns used were a Waters Spherisorb ODS2 PEEK (poly previously described and 9 replicates analysed. One matograms were extracted at 450 nm. Exhaustive liquid– carotenoid were injected in the HPLC system on the ﬁrst liquid extraction (3 times energetically hand shook) was day for purity check (purity. a 2996 Photodiode Array Detector and nised. prepared by elimination of the peduncle and surrounding Carotenoids separation and quantiﬁcation was done by area. acetoni- lysed on the same day. v/v) using a batch lycopene. based on peak areas. in each spiked sample 2. cat. Individual working solutions of each fen. a 717plus GM200. 47.45 lm PTFE mem- and analysed after 24 h. ‘‘for sal. each one with half of each tomato. and lutein were purchased from Sigma Chemical Co. cess and its recovery was used to correct carotenoids con- nized as above described (assuring similarity between the centration. otene 2560 (450 nm). M.1. a-carotene. Two injections per vial were made.. region of Portugal. The other group was refrigerated at 4 °C triethylamine.1% (w/v). composed by a 600 Controller and Pump. Analytical method b-cryptoxanthin from Carl Roth.6 mm. with certiﬁed value for a-carotene. var.5 mL/min.6 mm connected to a reverse phase C18 (Vydac. 1995) and their All chemicals were at least of analytical grade. the most unfa- of each analytical sample were weighed in a balance with vourable measuring situation) using the following an acceptation criteria of 0. 250 4. b-carotene and made to transfer analytes to a petroleum ether phase which lutein and information value for lycopene and zeaxanthin. zeaxanthin and 2.007 of absor- The homogenised sample replicates (approximately 10 g bance. Edible part was prepared as above and calibration. with an unitary Sample carotenoids were identiﬁed using the HPLC sys- mass in each lot ranging from 112. The edible part (skin and seeds included) was time possible. a thin slice of the exposed tomato brane ﬁlter and degassed in an ultrasonic bath. was ﬁltered through a 0.G.5 mL of dichloromethane (DCM) and then diluted in analysed. 12 lots (1 kg each) of ﬁrm ripe tomato (Lycop. 2480 (451 nm). no. 8161).45 lm PVDF syringe ﬁlter and injected in the HPLC sys- 2.9 g to take tem software by comparing their retention time (RT) and into account the possibility of carotenoids concentration visible absorption spectra (kmax and spectral ﬁne structure being a function of the fruit size. 1995) was also used) with were harvested from 12 subﬁelds at the same day and by those of carotenoids standards.0002 g) added with magnesium extinction coeﬃcients – E1%1 cm and wavelengths: ethanolic carbonate and internal standard. 18 portions were immediately weighed. under inert atmosphere (nitrogen) and were ana.2. 5 lm. evaluation. var.05% lysed 48 h later. hexane solution of zea- extracted three times with tetrahydrofuran and methanol xanthin. spectrophotometer (acceptation criteria of 0. checked using the HPLC system software to adjust the inte- ated until the day of analysis (3–4 days) and were analysed gration. Dias et al. 5 lm. butylhydr- by participation in an interlaboratory study (baby food) oxytoluene (BHT) was used as antioxidant at a concentra- for b-carotene and by the recovery of a non food native tion of 0. Flow rate was ersicon esculentum M. 75:20:5. 201TP54).4 g to 200. For small-scale within-region carotenoids variability 100 4. Retsch) 14 s at 7000 rpm and carefully homoge. Stau. the mobile phase. v/v/v. / Food Chemistry 109 (2008) 815–824 817 carrots). The quantiﬁcation was done by external standard in 2 successive days. When sample solution analyte concentration over- came calibration curve higher point.

equally distanced) were injected daily to and within and between day variations. for the quantiﬁ. ues for total-lycopene (1. 98.55 mg/100 g). one using standard solutions and the other . 100%) and the concen. the homogeneity of variances was evaluated through an metic mean of the diﬀerent detection limits obtained with F-test at a signiﬁcance level of 1%. with b-apo- tion were determined based on two diﬀerent methods: 8’-carotenal (a carotenoid similar to those analysed and method A – based on the calibration curve (Miller & not present in test samples). product- was calculated as 3 hnoise C/hpeak. It was estimated as the arith. of carotenoids absent in the matrix were conﬁrmed by spik- thin. Dias et al. the signal output per unit concentra- noise ratio. Method in-house validation z-score is acceptable for values between -2 and 2. Vf (diﬀerence between the value obtained by the laboratory is the ﬁnal volume (mL). and information val- was obtained by calibration curve interpolation. in lg/mL (Csi) mg/100 g) and lutein (1. using six levels of calibration. 1990).97 mg/100 g). d is the dilution and 10 is CRM or BIPEA value) to standard deviation of data a unit conversion constant. and was based Linearity was checked between 0. bias was estimated by a recovery test. The quantiﬁcation limits and 98.818 M. for trans-a-carotene. in mg/100 g. 0. Indi- rectly. the sy=x Rx2i =½nRðxi –xmean Þ2 1=2 and sy/x (standard error of y- Community Bureau of Reference. sample. each one injected in duplicate. mobile phase and calibration solutions lg/mL of extract. cation limit the factor 3 was substituted by 10. 2. in sample measurements were made in 3 diﬀerent days. m is the sample mass (g). 1994(E). sa (standard error of the intercept) = and carrots) from former European Commission BCR. The detection limit obtained by method A tion. trans-a-carotene (1. the were prepared each day. 1993). spiking each On a ﬁrst approach the limits of detection and quantiﬁca. under and above. considering a 95% conﬁdence interval. Repeatability and with a mixture of the six carotenoids (concentrations intermediate precision were estimated considering within. and upper and lower calibration limits were calculated and that the curve intercepts zero. t value is taken at the conﬁdence level of 95% for (n – 2) Bias was evaluated by duplicate analysis in four diﬀerent degrees of freedom. b-cryptoxan.25 mg/100 g).G. tomato sy=x =½Rðxi xmean Þ .05–4 lg/mL. respectively. Carotenoids stability studies food matrix were conﬁrmed by preparing and analysing 10 independent standard solutions at the lowest quantiﬁca. Linearity was tested with three criteria: day.3.05 lg/mL and 4 lg/mL on the assumptions that. after weighing and before extraction. (0. adjustment.42 mg/100 g) and zeaxanthin The analyte concentration in samples.9975. where hnoise is the mean moment correlation coeﬃcient equal or higher than height of the noise and C/hpeak is the concentration of a 0.05 mg/100 g). The detection limit obtained by method B close to the calculated linear regression line. in licates each day. R is the and the best estimate of the ‘‘true” concentration (assigned b-apo-80 -carotenal recovery (%). Ten independent calibration nal of a solution with a concentration near the blank is solutions of the lowest. the standard deviation of the sig. was given by the expression 3. and at a signiﬁcance level of 1% a non-linear cali- solution giving a peak with roughly 5 times the height of bration function does not lead to a signiﬁcantly better the noise to the height of that peak ratio. C.3. sb (standard error of the slope) = days of the Certiﬁed Reference Material (CRM) 485 (lyo- 2 1=2 philised mixed vegetable preparation of sweet corn. Twenty seven Y ¼ ða tsa Þ þ ðb tsb Þ X. obtained in the interlaboratory study for the CRM or BIPEA quotient). To deal with eventual equipment failure. tem. trans-b-carotene (2. build the calibration curves by the least squares method.4. ing matrix samples at the lowest quantiﬁcation level trations of the standard solutions were corrected obtained by the above methods and comparing the relative accordingly. with certiﬁed values for residuals) = ½Rðy i ^y i Þ2 =ðn 2Þ1=2 (Miller & Miller. freshly prepared each tion was calculated. there intermediate calibration level were analysed by HPLC sys- is a normal distribution of the signal at this concentration. was given by the slope of the regression line. Signal variances of the probability of 5% of occurring error type a or b. Trueness was also checked by participat- was calculated using the following equation: ing in a round of BIPEA (Bureau InterProfessionnel C si V f d’Etudes Analytiques) proﬁciency scheme (15 participants) C¼ 10 ð1Þ for the determination of b-carotene in baby food (assigned mRd value – 2. Miller. ten of the highest and one of each roughly the standard deviation of y-residuals (sy/x). where Y is the peak area. according to ISO 5725 – 1 to 6. 1993) and method B – based on detector signal to Detector sensitivity. a is the intercept and b is the slope. 9 rep- lV s and X is the concentration (also called C below). The quantiﬁcation limit was estimated using the factor experimental calibration points lying.9%. Six levels of working calibration solutions standard deviations with a limit of 10%. all-lycopene and zeaxanthin. respectively. Linear calibration func- the diﬀerent calibration curves. / Food Chemistry 109 (2008) 815–824 individual carotenoid to sum of all chromatogram peak the relative standard deviations obtained with sample areas ratio was: for trans-b-carotene and lutein. tested by an F-test (ISO 8466-1.3 sy/x/slope. The quantiﬁcation limits of carotenoids present in the 2.4% repeatability standard deviation. Bias was quantiﬁed through z-score where Csi is the sample solution concentration (lg/mL).37 Sample solution analyte concentration. two studies tion limits estimated by the above methods and comparing were conducted. 10 instead of 3.

To evaluate the stability was applied to analyse 12 tomato (Lycopersicon esculentum of tomato sample solutions containing lutein. E1% 1 cm is the extinction cient (r2 = 0. 2000). for each analyte was calculated.7. 15. . 3. and at 4– uncertainty (uc). . 11. s2.05 lg/mL and 4 us ðyðx1 . . 23. at the conditions: 36 °C (an CHEM/CITAC. each day corresponding to a storage time (i. the combined stan- dard uncertainty us(y) of a value of y is related to the uncer. Contribution from stock standard solution con.. b-carotene M. obtaining the combined standard extreme laboratory temperature in summer). ci is a sensitivity coeﬃcient evaluated as ci = oy/ level 1%). the same matrix and uncertainty proportional to concen- (2).000 is a unit conversion constant.6. uc. . for the six analytes. tration (concentrations above the quantiﬁcation limit).3. s2. performance data of dards. 5. Chromatograms 2. 1). V 1 is the initial volume (mL) lines. the seven carotenoids in a mixture of stan- laboratory data: studies of precision.e. 7. and C0 is the concentration before storage. expressed as a standard deviation. 3. Retention times (in min- ponents. 24 h and 41 within-region carotenoids variability. on which it depends by the following equation: The diﬀerence between the variances of the signal sﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ X obtained for concentration solutions 0. Linearity tainty estimation of the independent parameters x1. After estimation of all sources of uncertainty they were bility working standard solutions at six levels of concentra. conﬁrming the high linearity veriﬁed visually oxi (partial derivative of y with respect to xi) and u(xi) is by the graphical representation of the six experimental cal- the uncertainty in xi.5.3. To evaluate standards sta. . b-apo-8’-carotenal. . utes) were 5.. The expanded uncertainty (Uc) was calculated as Uc = k uc. the minimum value of P was 0. . C 24 C 0 measurement uncertainty. ibration levels and by the statistics product-moment corre- This general procedure was applied to calculate uncer. ration and after 24 h storage at 4 °C and concentrations Assuming a similar behaviour for diﬀerent varieties of were evaluated by comparing the relative error (Er). Considering the experimental design C0 the maximum expected contribution of analytical uncer- where C24 is the concentration after 24 h storage at 4 °C tainty to the observed variance. 000 ð4Þ Although point (0. this point is included in the conﬁdence limits where V 2 is the re-dissolution solvent volume (mL) mea.8 and 25. & Miller. for the intercept (conﬁdence level 95%) in all regression sured with a 10 mL pipette. Eq. an one-way mately inﬁnite degrees of freedom (because uc collected ANOVA statistical test was applied to repeatability and information from the combination of the various sources intermediate precision results to evaluate the diﬀerences of uncertainty).G.2.997).1. 1993).. with injector repeatability relative standards deviation. lation coeﬃcients of ﬁrst order regression functions tainty associated to stock standard solution concentration (r P 0. .9985). (Cstock) which is obtained from the following equation: The parameters of the regression equations obtained for Absorbance V 2 the calibrations curves are shown in Table 1. 0) was not used to build the calibra- E1% 1 cm V 1 tion curves. xn .995) (Miller coeﬃcient and 10. b-cryptoxanthin. is the estimated u2c . 3. lycopene. 12 h. xn ÞÞ ¼ c2i uðxi Þ2 ð3Þ lg/mL was not signiﬁcant at a 1% level of signiﬁcance i¼1. .10). Analytical uncertainty and natural variation The chromatograms show that the method can success- Measurement uncertainty was estimated based on intra.n (considering the diﬀerent carotenoids. . . lysed. var. The obtained values were compared through a F-test. estimated individually by a mathematical model (bottom. region and the variance. concentration signals obtained for the six standards at Natural variation was evaluated based on small-scale t = 0 h with those obtained at t = 6 h. The validated method h. A paired two-tailed t test was done to compare the where k is the coverage factor (EURACHEM/CITAC. at 36 °C and at t = 24 h at 4 °C. Dias et al. . fully separate. between the 3 mean results of 9 replicates analysed each day. Since tomato is a perishable product and in order to where s has 11 degrees of freedom and uc has approxi- guarantee the reliability of precision studies. trans-a-carotene and trans-b-carotene. . respectively centration is not covered by precision data and it was for lutein. Lido) samples harvested in 12 subﬁelds of the same and lycopene they were analysed immediately after prepa. the internal standard added (Fig. 8 °C (remaining solutions storing conditions) were ana. . The non-linear calibration functions did where: yðx1 . C stock ðlg=mLÞ ¼ 10. xn. Zeaxanthin showed the smallest determination coeﬃ- measured with a pipette of 1 mL. xn Þ is a function of several parameters not lead to a signiﬁcantly better adjustment (signiﬁcance x1 . . M. / Food Chemistry 109 (2008) 815–824 819 using tomato sample solutions. but still acceptable (r2 P 0. and the carotenoids present in a tomato sample with the analytical process and quantiﬁcation of individual com. . 2000). Results and discussion 24 h at 4 °C and 48 h at 21 °C).2.. According to this model. . up approach). 0 h. joined according to their laws of combination (EURA- tion.0. was estimated for each ana- Er ¼ 100 ð2Þ lyte in these samples. zeaxanthin.

00 0.0182 0. For carotenoids not 3.2) 104 (0 ± 4.46 2. N = 5) (s = 0.14 lg/mL).4 ± 0.48 lg/mL). respectively.81 0. the detection and quantiﬁ.00 25. Chromatographic conditions as referred in Section 2. t-statistics parameter. c = 0.6) 104 (50.313 min.00 15.33 Regression curve data a ± tsa (lV s) (1 ± 1. (s = 0. Table 1 Limits of detection and quantiﬁcation and regression curve data Parameter trans-a-Carotene trans-b-Carotene b-Cryptoxanthin All-lycopene Lutein Zeaxanthin Detection limit (lg/mL) 0.4) 104 (42.00006. precision.71 lg/ mL).00 20.5) 104 (1 ± 1. 3.8) 104 (0 ± 1. / Food Chemistry 109 (2008) 815–824 0.045 min).0492 0.77 Quantiﬁcation limit (lg/mL) 0.76 1.759 min.51) 104 r 0.00 30.030 0. 1. c = 1.8 ± 0. c = 0. (s = 0. (s = 0. c = 0.G.0154 (lg/100 g) 0. Repeatability and intermediate precision found in tomato.00 25.16 lg/mL).600 min. (5) lycopene (t = 15.00 15. which is to noise ratio).0% for lutein under sample repeatability relative standard deviation 4.74 2.686 min. Dias et al.178 min. uncertainty should be lower than 1 lg/100 g which consid- ations were 2. (7) trans-b-carotene (t = 25. sb – standard deviation of slope.2) 104 (47 ± 2. b – slope.1. only the analytical process random errors inﬂuenced the The obtained detection and quantiﬁcation limits are pre.040 1 3 0.0115 0. suggested limit of 1 lg/100 g for carotenes in food compo- cation limits were always higher for method A (based on sition databases (Greenﬁeld & Southgate.9995 1.00 35.10 0.9995 0.002. (4) b-cryptoxanthin (t = 11.002. (6) trans-a-carotene (t = 22.05 7 1 AU 0. 3.3) 104 (1 ± 2.00 10.0399 0. zeaxan.99 2.1% a-carotene and 5.0162 0.0504 0. r – product-moment correlation coeﬃcient.51 lg/mL). (3) b-apo-80 -carotenal (t = 7.00 5.4. (2) zeaxanthin (t = 5. In order standard deviation of y-residuals (sy/x) of the calibration to conform with the above recommendation the reconstitu- curve to slope ratio) than for method B (based on signal tion volume should be 1.050 0.060 5 Standards 2 0.15 AU 0. do not allow complying with the Except for b-cryptoxanthin.0166 0. 6.0553 0.0000 0.0347 0.9985 (s = 0.3.00 30.000 5.00006.51) 104 (34 ± 1.820 M.66 0.83 0. as described.0006.9990 0.0466 (lg/100g) 2.62 lg/mL).52 1.0% for lycopene and ering matrix interferences is questionable.9996 0.305 min. 2003).00 Minutes Fig. values obtained for repeatability and intermediate sented in Table 1. Chromatogram of a standard mixture and a tomato sample solution showing carotenoids elution proﬁle – (1) lutein (t = 5.0132 0.0.0% b-cryptoxanthin.91 0. The relative standard devi.20 Tomato Sample 0.7% for b-carotene. Limits of detection and quantiﬁcation The Limits of Quantiﬁcation obtained with this analyt- ical method.00 3 Minutes 0.4) 104 b ± tsb (lV s mL lg1) (51 ± 1. sr – standard deviation of product-moment correlation coeﬃcient.7 mL.8) 104 (48 ± 1. a – intercept. sa – standard deviation of intercept.0. c = 0.010 7 6 0.3%. instead of 50 mL. (s = 0. the relative standard deviations were 5.57 0. 3. c = 0.5) 104 (0 ± 2. .00 20.020 4 5 0. experimental conﬁrmation of quantiﬁcation not practicable for the method as designed (the minimum limits was performed at the lowest concentration level feasible reconstitution volume would be 5 mL) and the obtained by the above methods. 5. Sample stability study results (see below) insured that thin all below the 10% limit. N = 5) N = 11) N = 5) N = 8) N = 11) N – number of freshly prepared curves in diﬀerent days from standards also freshly prepared.0003.00 10. 5.

21%. All 3.0777 0.0776 Overall mean (mg/100 g) 0.7.38 0.2032 2.0744 0.3578 2. / Food Chemistry 109 (2008) 815–824 821 The analytical results (raw data) of repeatability and The z-scores for CRM 485 reference material were.05. Sensitivity obtained z-scores were satisfactory. this agreement with values referred in the literature (Hart & result was expected since lycopene is the most unsaturated Scott.3267 2.0757 0.3893 0.061 for trans-a-carotene.3446 2.39 2.24).3842 0.4049 0.7% (s = 10. Bias there were non-signiﬁcant diﬀerences in the detector response to the diﬀerent concentration levels of standards Laboratory analysis of CRM 485 gave the following solutions up to 6 h at 36 °C (0. senting the worst performance.0723 0.0 6.3542 2.3713 0.4448 2. 1.0747 0.0757 0.2048 2.0747 0. 1. n = 48).16. Identical behaviour for wavelength is 450 nm. 0.2957 2.0800 0. .2763 2. Recovery values.003 standard deviation (mg/100 g) Repeatability relative 4. thin.0802 0.026 0. lutein.0760 2. approaches for trueness quantiﬁcation.051) (P = 0.3548 2. each one with lim- measurement wavelength was chosen in order to be as close itations.2 of the mean (%) Repeatability limit (r) 0.3667 2.3845 2.66 for all-lyco- and intermediate precision and stability signiﬁcance test pene.076 Relative standard deviation 4.0755 0.3791 0.4 relative standard deviation (%) Stability signiﬁcance NS*** NS*** NS*** test (ANOVA) (P = 0.G.0753 0.39.3981 0.0738 0. 0.1586 2. M.5%. all-lycopene.6.0504 0.3785 0.3743 2.3848 0.3 0.0746 Statistics Daily mean (mg/100 g) 0.0760 0.0 3.0756 0. At a signiﬁcance level of 5% and for all carotenoids.3884 0.3090 2.3922 2. At 36 °C there is a signiﬁcant thin.4680 2.3747 0.0 6.3741 2.0756 0.5.3939 0.83. Dias et al. intermediate precision as well as statistics such as mean.1862 0.0743 0.0795 0. Given that the integration algorithm added and natural carotenoids was assumed. based on the addition of b-apo-80 -car- According to results presented in Table 1 the method is otenal were 93.043 0.4398 0.15) (P = 0.19) *Stored 24 h at 4 °C. 1995) for diﬀerent vegetable matrices (87. 0. The three of HPLC system software always needs manual adjustment. carotenoids more frequently present in food matrices. 3.3156 1.0754 0.8%.3955 0.1 standard deviation (%) Intermediate precision 6.3918 0.3755 2. 2. BIPEA test material analysis gave a b-carotene diﬀerence at t = 12 h (lycopene and a-carotene). 4 °C (0.06 < P < 0.097 for trans-b-carotene. its absorption maximum being therefore s = 8.0787 0.21 < P < 0. 0.4062 0.0777 0.2679 2. the reference material did not contain b-cryptoxan- for tomato matrix are presented in Table 2. t = 24 h result of 1.86 mg/100 g.4279 2.0739 0.4103 2.0067 (mg/100 g) Intermediate precision 0.9894 0.64 for zeaxanthin. and at least up to 24 h at carotene.0800 0. 0. 1.0756 0. which are in good 33% less sensitive for lycopene than it is for a-carotene. Standards solutions stability grating chromatograms at diﬀerent wavelengths.3917 0.36 was obtained for b-carotene in BIPEA proﬁciency testing scheme participation. Table 2 Carotenoids in tomato – Precision data (raw data and statistics) Parameter trans-b-Carotene All-lycopene Lutein * ** * ** Day 1 Day 2 Day 3 Day 1 Day 2 Day 3 Day 1 Day 2* Day 3** Raw data (mg/100 g) 0.3770 2. acyclic carotenoid.7 3.0794 0.4 12.3813 0.2608 0. 3.0331 0.0728 0.53 for lutein. trans-b.71). not signiﬁcant. lycopene pre- mean results in mg/100 g: trans-a-carotene.2787 0.05.1220 0. **stored 48 h at 21 °C. with the purpose to avoid the extra work of extracting and inte.4572 2.15 0.3907 0. n = 127) and it is not statistically diﬀerent from at a wavelength around 473 nm whereas the measurement 100% at a signiﬁcance level of 0.3430 0.3935 0.8 11. repeatability 1. A z-score of 0.0800 0. ***signiﬁcance level – 5%. agree with the fact that the laboratory method as possible to the absorption maximum of the analysed performance has no signiﬁcant bias.05 and zeaxan. relative standard deviations of the mean.2330 0.3741 2.5301 2. NS.

and a rectangular distribution for 8 0.096 0. Analytical uncertainties random errors from many variables.0746 0.5847 0.61)*** (F = 6.0727 volumes of 10 mL and 1 mL.64)** ment uncertainty and C is the concentration) for diﬀerent test (F-test) analytes was 0.9714 7. deviation (s) carotenal in this matrix (n = 48). For the tomato matrix studied. / Food Chemistry 109 (2008) 815–824 (all except b-carotene) and t = 41 h (all). **critical value of F – 1.12 for lycopene and *Signiﬁcance level – 5%. Variability Signiﬁcant* Not signiﬁcant* Signiﬁcant* bined relative uncertainty (uc/C. 2 1.9303 0.9954 7.9 9. 2. Sample solutions stability Comparing the results for sample solutions immediately analysed and after storing 24 h at 4 °C.10 Uncertainty associated to bias was calculated based on Standard 0.79 0.0882 conﬁdence interval (level of conﬁdence not provided) of 7 0.0431 9. a negligible 11 0.1393 4 0.97 8. 0.97)** (F = 1. because it includes 3.9006 8.007 and a rectangular 5 0. because when assuming inﬁnite degrees of the precision and recovery studies since all these instru- freedom t-Student distribution tends to a normal distribu- ments are under regular control and several have been used.070 1.007 mL. Tomato samples stability At a 5% signiﬁcance level the one-way ANOVA test allowed to conclude that there were non-signiﬁcant diﬀer- ences between the three mean values (P values: 0. a 6 0.8093 0. Mean 0. Sample homogeneity and linear calibration curve interpola- tion uncertainties are included in the precision uncertainty because various replicates from the same sample were ana. stock standard solution concentration (Cstock) Subﬁelds was determined spectrophotometrically only at its prepara. 2.9859 0.1242 0. the com. respectively for uncertainty 9 0. therefore solutions are stable at least 24 h. tion. prior to analysis. Table 3 lysed and standards were injected each day.0804 photometer acceptation criteria of 0. Results are presented in Table 4.86. ***critical value of F 0. Uncertainties histogram. Tomato Fig. Small-scale within-region carotenoids variability Uncertainty coming from stock standard solution was Trans-b-Carotene All-lycopene Lutein studied separately because: work standards were prepared (mg/100 g) (mg/100 g) (mg/100 g) from it.9323 0.7335 0.1262 0.0900 Uncertainty was calculated taking in account: a spectro.1763 9.1309 uncertainty arriving from the extinction coeﬃcient (not 12 1.8972 7.0786 considered).1020 tion day and it involves a dissolution solvent change.022 the standard deviation of the recovery of added b-apo-80 .2% (b-carotene) and 2% (lycopene).822 M.051 for lycopene and 0.0945 coming from measurement with pipettes.0916 10 0. 0.0781 distribution for uncertainty arising from absorbance. the relative error of the carotenoids concentrations at the two diﬀerent temper- atures in the samples varied between 0. if stored in the refrigerator. 1 1.7 23 Long term inﬂuences of environmental temperature are deviation (rsd) not considered since these samples are perishable.05 mL and 0. 3.8. Dias et al. 3.88.073 for b-carotene.0586 8.10. Relative standard 9.3306 0.0406 8. these values are similar to the ones of repli- cate injections on the HPLC system.0 0.9.9460 8. 3 1. samples may be stored at 4 °C at least for 24 h and at 21 °C at least for 48 h.8697 7.088 for lutein.3067 0.G. Expanded uncertainty was calculated for a level of con- Sources of uncertainty such as those arising from bal- ﬁdence of approximately 95% considering a coverage fac- ances and volumetric measuring devices are covered by tor of 2. Standard solutions can be kept at least 24 h at 4 °C and only up to 6 h at 36 °C.8641 6. where uc is the measure.7744 0.19 for lutein). Analytical 0.7473 0. The major contribution to uncertainty – 2. .1 0.0086 measurement The results are presented in the Uncertainties Histogram uncertainty (u) shown in Fig. signiﬁcance (F = 1.8782 6. was intermediate precision as expected.15 for b-carotene.

position tables are apparently very precise (expressed with niﬁcantly diﬀerent for b-carotene and lutein and not signif. low concentration of lutein and since there are other good sources of this carotenoid. Rearranging the above expression and taking for lutein and 54.21)* <0. (b-carotene and lutein) content in tomato is aﬀected by etc. such ﬁgures may be unrealistic.8%. were 0.006 2. 3. carotenoids content and dominates analytical sources such Tomato samples can be stored at least 24 h at 4 °C or as analytical determination or sub-sampling of test materi. for the tomato matrix.4%.604– standard deviations of the intermediate precision. With n smaller analytical steps. Small-scale within-region carotenoids variability its contribution in the estimation of an overall uncertainty.582–5. 1992).05. uncertainty than consideration of the lab-based analytical The comparison of u2c with s2 through a F-test. The contribution of the analytical mea. associated uncertainties at a level of conﬁdence approxi- mately 95% are presented in Table 4. Olme.39 mg/100 g). the opti- unstable of the studied carotenoids and the method is less mum sample size can be calculated as n P (ta. . are considered. 1995) In this work repeatability limits.3 and 8 mg/100 g while the above men.0067 mg/100 g for lutein showed values of 2. suggesting that carotenoids when analytical uncertainty. Carotenoid values included in food com- to conclude that. therefore. even for a small-scale geographical variation.39 ± 0.006 8 ± 2. non-signiﬁcant diﬀerences (accuracy xmean)2 (Greenﬁeld & Southgate. they are sig.3 ± 0.015 <0. ‘‘for salad” *Values between () include small-scale within-region variation. 0.0 mg/100 g) agree with values HPLC has allowed good separation of all carotenoids observed by other authors for other tomato varieties.2% for b-carotene. 62. it may be considered not to 3. n will be 5 for sequence of such a high measurement uncertainty that can the determination of b-carotene and lycopene and approx- mask the variations deriving from luminosity/geochemical imately 20 for lutein.494 mg/100 g in 3 varieties (Granado. 48 h at 21 °C.008 1. Dias et al.1. 1.076 mg/100 g) and the relative tioned authors found. 11.) onstrated for tomato by estimating uncertainties. Based between partial and total variances can eventually be a con.57 0. p 95% ﬃﬃﬃ of the sample pﬃﬃﬃ means lie in the can be more relevant than uncertainty arriving from the interval ½l 1:96ðr= nÞ. etc. using s to estimate r surement result to the total variance (ratio of analytical involves the substitution of 1. on these results. for the same sample concentrations. growth conditions. For an accu- racy of 0.0 (1.653 mg/100 g and 1.702 mg/100 g in 11 varieties (Hart & Scott.393–0. for an accuracy of 0.1% freedom. / Food Chemistry 109 (2008) 815–824 823 Table 4 Carotenoids content of two tomato varieties trans-a-Carotene trans-b-Carotene b-Cryptoxanthin All-lycopene Lutein Zeaxanthin (mg/100 g) (mg/100 g) (mg/100 g) (mg/100 g) (mg/100 g) (mg/100 g) Tomato Lycopersicon esculentum <0. Standards and sample als within the laboratory. 6. allowed procedures alone.n1)2 s2/ sensitive for it.0 ± 0. as dem- localization (factors such as luminosity.008 0.049)* <0.056 <0. Tomato carotenoid content and uncertainty apply so many resources in its determination.38 mg/100 g for lycopene (sample con- be prone to major variations with variety. and 0.10 ± 0.008 M.008 M. variety. Localization Considering the sampling distribution of the mean for n natural variation from even a very small-scale plantation samples as normal. 2003). respectively. var. & Rojas-Hidalgo. Lycopene seems to 0. M. were.14 (0. geochemical. Lido Tomato Lycopersicon esculentum <0. var. Nevertheless. respectively. Lycopene is the most (xmean l)/xmean as a measure of the accuracy.7% and The obtained results indicate that primary sampling 3.3 mg/100 g) and 0. 4.96 for t at n 1 degrees of measurement variance to total variance u2c =s2 ) is 15.G. as tomato has a very conditions.11. many digits which may not be signiﬁcant ﬁgures) however.12. Recognizing sampling as the ﬁrst solutions can be stored at least 24 h at 4 °C and only up step in the measurement process there is a need to include to 6 h at 36 °C.273 mg/100 g.7)* 0. analysed samples centration 2. (sample concentration 0. l þ 1:96ðr= nÞ.017 (0.39 mg/100 g and 1. Blanco. the number of samples would be 19 for b-car- Carotenoid content of two diﬀerent tomato varieties and otene and lycopene. Having in view the production of results to food composi- The results from the 12 tomato samples harvested in the tion table. before grinding. than required by normal distribution.08 ± 0. icantly diﬀerent for lycopene. 0. Conclusions Trans-b-carotene content of analysed tomato samples (0. intended to be analysed. seems to be the main source of uncertainty for tomato Trueness studies showed good laboratory performance. at a signiﬁcance level of 5%.043 mg/100 g for b-carotene (sample concentration dilla.349–1. this will give more realistic estimates of the same region are shown in Table 3.

. Journal of Food Compo- Nutrition. Gracßa Dias is Krinsky. R. lymphocytes after tomato intake.). promotion. G. management and use (2nd ed. and the Improvement of the analytical method for lycopene. (1995). 221S–224S. Di Mascio. 57(1). www. (sample homogenization. J. International Organization for tomato from the region under study 5 primary samples Standardization.ul.. Some considerations in generating evaluations. M.824 M. 40.chem.. J. HPLC method for the analysis of carotenoids in foods. Silva. biological carotenoid singlet oxygen quencher.. 1350–1358. supplying var. J. Analyst. (1993). Food composition data: overshadow contributions from the analytical process Production. Geneva. H. Erba. FASEB Journal. 134(1). Switzerland. F. (1992).qmul. D. on lycopene content in tomato. 54. Miller. F. Dias et al. Olmedilla.. the measurement of the carotenoid content of vegetables and fruits.. J. & Ellison. would be suﬃcient to estimate the population mean value ISO 8466-1 (1990(E)). F. 124. Switzerland.. Statistics for analytical chemistry (3rd ed. Determination of nine carotenoids. Geneva. E. Water quality – Calibration and evaluation of with a level of conﬁdence of 95% and 10% accuracy. variety (for the two diﬀerent analysed varieties. Blanco. analytical methods and estimation of performance characteristics. S. 274.. Santos. Guide to the expression of uncertainty in measurement (GUM). A. Hart. 924–927. D. Barwick. E. (2000). (1995). International Organization for Standardization. H. M. Uncertainty of measurement: tocopherol in control human serum using two internal standards. Structure and properties of carotenoids in relation to protection. geo. J. IUPAC Commission on Nomenclature of Organic Chemistry and the Acknowledgments IUPAC-IUB Commission on Biochemical Nomenclature (1974). P. West Sussex.. & Sies. Granado. Accuracy (trueness and precision) of Based on this study. 2135–2140.carot/. (1999). I. sition and Analysis. M. J. D. T. D. Major Variations on tomato carotenoid content introduced by fruits and vegetables contributors to the main carotenoids in the small-scale diﬀerences on geographical factors. 9.. (2004). M. (2004). 532–538. terminology for the approaches to the quantiﬁcation of the measure- Eurolab. / Food Chemistry 109 (2008) 815–824 Carotenoid concentration varied largely with tomato Granado.. J. reliability of results including a list of useful references. 26. & Southgate. I. E. Visioli. Approaches to the evaluation of uncertainties associated with recov.). B. 981–990... L. G..G. Testolin. retinol. K. EURACHEM/CITAC Guide CG4: Quantifying Uncertainty in Analyt. is required in Chemistry. J. 27 p. European Journal of Clinical Nutrition. Carotenoid actions and their relation to health and disease. D. Lido tomato samples. 641–647. in order to determine carotenoids in measurement methods and results. References Olmedilla.. (2002). R. Eﬀects on cellular antioxidant Britton. 85–90.uk/iupac. (1992). (1995). & Porrini. F. Geneva. retinyl palmitate and a- Analytical Methods Committee. Journal of Agriculture and Food Chemistry. 10(12). Accreditation and Quality Assurance. V. J. Berlin.ac. grateful to Instituto Nacional de Saúde Doutor Ricardo 459–516. Retrieved on 2004-06-01: http:// The authors thank Dr.. P. 101–111. Carotenoids in health and disease: Recent scientiﬁc Rodriguez-Amaya. 233 p. 58.eurachem. Joaquim Castro Nogueira for www. Lycopene as the most eﬃcient Scott... N. Finglas. Technical Report No. I. Jorge for providing the research fellowship BIC 01/2003-II. F. European Journal of Clinical Nutrition. Blanco.). Archives of Biochemical (1996). sample extraction. Seale.. (1996). Lycopene and tomato products in health Technical Secretariat.g. A. Eurolab Weisburger. carotenoids in foods. 120. Nomenclature of carotenoids.. 50. Measurement uncertainty: Riso. E. C. & Rojas-Hidalgo. I. D. Rome. I. Molecular Aspects of Medicine. variation coeﬃcient was 113%). Granado. function. e. 227. N. (2006). F. P. & Miller. Germany. Development and evaluation of an measurement). S. Analyst. C. & Froidmont-Görtz.. A new ical Measurement (2000) (2nd ed. Interlaboratory studies of HPLC procedures for the analysis of and Biophysics. R. 205–213. H. & Rojas-Hidalgo. (2002). J. 1/2002. Cooper. Kaiser. order to enable detection of possible location inﬂuence ISO (1995).pt. Great Britain: Ellis Harwood Limited. B. chemical or luminosity. B.. Olmedilla. Experiments in Biology and Medicine. Chemistry.. G.. Journal of carotenoid data for food composition tables. & Camões.. Food most unstable of the studied carotenoids. J. Spanish diet.. Food Chemistry. R. Lycopene and vitamin C concentrations increase in plasma and ery. 13. N. 1551–1558. B. & Rojas-Hidalgo. 246–250. & Scott. Italy: FAO Publishing Management Service. (1989). research recommendations and the consumer. International Organization for Standardization. Measurement uncertainty ment uncertainty. & Johnson. Blanco. Switzerland. . 2303–2308. 45(3). (2003). ISO 5725 – 1 to 6 (1994(E)). HPLC system Hart. B. are very important and can easily Greenﬁeld. K. lycopene Carotenoid composition in raw and cooked Spanish vegetables. Food implications of its use in analytical science. (2005).. in testing – a short introduction on how to characterise accuracy and 664–671.

- Checklist for DesignUploaded byOmran Ibn Wahid
- Eurocode 2 Variable Strut Inclination Method for Shear, its Modelling Uncertainty, and Reliability CalibrationUploaded byKenneth Mensah
- Propagation of UncertaintiesUploaded byalejandro_salcedo_69
- error analysisUploaded byAbhishek Singh
- 12 IAJMR - Ogu.pdfUploaded byIndo Asian Journal of Multidisciplinary Research
- Food Chemistry Volume 76 Issue 1 2002 B.dave Oomah; Muriel Busson; David v Godfrey; John C.G Drover -- Characteristics of Hemp (Cannabis Sativa L.) Seed OilUploaded byDynamik Edibles
- Introduction to IB ChemistryUploaded byAnnika Feign
- Analytical Chemist or Associate Scientist or Research AssociateUploaded byapi-77593584
- Jurnal Stabilitas Suhu PctUploaded byAdnan
- 40C Anemometer Uncertainty AppNoteUploaded byDavid Patiño
- ASME PTC 6 REPORT_1997.pdfUploaded bySunil Koricherla
- 6Uploaded bymjamil0995
- 5991-8156EN.pdfUploaded byAlonso Hurtado
- Reliability, Validity, AccuracyUploaded bySilvia
- AntibioticUploaded byDian Indrawati Santoso
- DefinitionsUploaded byahmad atsari sujud
- TAC 357 - Problem Set 1.pdfUploaded bylucas
- Revisiting the Development of the BlighUploaded byLeonardo Merlo
- An Updated MethodUploaded byGuille Suástegui
- Calibration of Piston CylinderUploaded byFederico
- HPLC Analysis of AdiphenineUploaded byJuan Perez
- D 6032 – 96_Determining Rock Quality Designation (RQD) of Rock CoreUploaded byzuheib63
- Morphine BloodUploaded byThanh Mi
- Why System Suitability Tests Are Not a Substitute for Analytical Instrument Qualification or Calibration Part-2Uploaded byvg_vvg
- Accuracy TimingUploaded bybagus_santoso

- IPPC-MON BREFUploaded bysdiaman
- An Assessment of the Beerkan Method for Determining the Hydraulic PropertiesUploaded byAndré Ferraz
- RCT+Appraisal+SheetsUploaded byAinur 'iin' Rahmah
- h,i Elab y Curado Rapido Metodo Agua HirviendoUploaded byVictor Victorvictor
- 5081 Spicer Chapter 5Uploaded byshentiti
- Does Religiosity Promote Property Rights and the Rule of LawUploaded byCristina Colmenares
- IE241 Hypothesis TestingUploaded byAnonymous RJtBkn
- Ejercicios Sobre Intervalos de ConfianzaUploaded byLuis Herrera Avila
- Why is the Remaining Useful Life Prediction UncertainUploaded byPradeep Kundu
- ANALISIS ESTADISTICOUploaded byRaul Cuayla
- Pruebas de Hipotesis AEDC09Uploaded byPati Razo
- Capitulo 5 Simulación de Los Centros de Operaciones (1)Uploaded byDiana Cordova
- B.A. Hons. Psychology.pdfUploaded bypriyanka
- ESTADISTICA 2Uploaded byGuadalupe Melendez Alvarez
- SESION 2 estadisticaUploaded byRandleMcmurphi
- HA0801 CorrigeUploaded byjolegende
- Exercices de Statistique 2006Uploaded byAyoub Khairany
- 39gPlusApletsUploaded byNicolas Reyes Gonzalez
- Validation of Compendial ProceduresUploaded byRhea Black
- Tarea Primer Parcial Estadistica 2016Uploaded byBernard Alexander Zuñiga Davila
- Prueba 1Uploaded bygloriafields70
- OtometricsUploaded bystesag
- Clase 12Uploaded byOmar Aletz Acosta Sarmiento
- relacion55urso13_14Uploaded byAdev Adev
- NORDTEST TR 537 (2011-11) MNUploaded byMatea Malovrh
- Taller EstahsfhhdisticaUploaded byJuanma Ballen Forero
- Trabajo EstadisticaUploaded byBrajhan Chorres
- Suficiência amostral em levantamentos de D&A da vegetação.pdfUploaded byPoly Anne
- Ejercicios 1 Inferencial-1Uploaded byDarwin Mendoza
- Estatistica_Para_Leigos.pdfUploaded byFelipe Wavemaster