AOAC SPIFAN Nutrients ERP Method Reviews (September 7, 2019)

Animated publication

AOAC Stakeholder Panel on Infant Formula and Adult Nutritionals

(Method Reviews)

Meeting at: Sheraton Denver Downtown Hotel 1550 Court Pl, Denver, CO 80202 September 6 -12, 2019

Saturday, September 7, 2019

AOAC INTERNATIONAL 2275 Research Blvd., Suite 300 Rockville, MD, 20850

UNITED STATES dboyd@aoac.org 301.924.7077 x126

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method Amino-04 (Sep 2019)

Title: Determination of free and/or Total Proteinogenic Amino Acids and Taurine by pre-column derivatization and UHPLC Author: George Joseph, Asha Varughese (AsureQuality Limited) Reviewer Name: Reviewer 1 Summary of Method: The method has 2 parts –determination of total and determination of free amino acids. Since sample preparation procedures are different these parts should be considered as 2 deferent methods. For total amino acids: The samples are hydrolyzed in 6N HCl for 24 H in presence of phenol, 3,3’ – dithiodipropionic acid (DTDPA) and internal standard Norvaline. DTDPA converts Cysteine and Cystine to S- 2-carboxyethylthiocysteine (CYSx) derivative. After hydrolysis the amino acids and CYSx are derivatized using 6-amino-N-hydroxysuccinimidyl carbamate (AQC), separated on reversed-phase UHPLC column and detected at 260 nm. This method is not applicable to determination of Tryptophan as this amino acid is destroyed during the hydrolysis. Asparagine is converted to Aspartic acid and Glutamine to Glutamic acid during the hydrolysis. For free amino acids: Amino acids are extracted using 0.5 N HCl with Norvaline as internal standard, derivatized using 6-amino-N-hydroxysuccinimidyl carbamate (AQC), separated on reversed-phase UHPLC column and detected at 260 nm. Method Scope/Applicability: Method is applicable to determination of proteinogenic amino acids and taurine in infant, pediatric and adult formulas (powders, ready-to-feed and concentrates).

General comments about the method:

The method is clearly written and is easy to follow. Method utilizes well-established methodology and uses common analytical equipment. No proprietary reagents are used. For SLV of determination of total amino acids all SPIFAN matrixes, NIST SRM 1849a, SRM 1869 and 2 additional matrices (goat milk infant formula and whey protein concentrate) were used.

For SLV of free amino acids analysis spikes of NIST SRM 1869 at 3 different levels were used.

1

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

It would be useful for the authors to include actual concentrations of amino acids found in SPIFAN matrices as well as actual spike levels used for spike-recoveries studies for total amino acids.

Method Clarity: Clearly written, lists all the reagents, standards and supplies

Pros/Strengths: For total amino acids • Uses established methodology and commonly available equipment

• Fast chromatographic analysis with good peak separation • Separates Citrulline from proteinogenic amino acids • Method describes all steps and procedures to prepare reagents, standards and samples • Linearity, accuracy and repeatability for SPIFAN matrices are evaluated and compared with SMPR For free amino acids • Uses established methodology and commonly available equipment • Fast chromatographic analysis with good peak separation • Separates Citrulline from proteinogenic amino acids • Method describes all steps and procedures to prepare reagents • Linearity and other parameters are evaluated using spikes of SRM 1869 and compared with SMPR Cons/Weaknesses For total amino acids Not suitable for Tryptophan For free amino acids Only spikes of SRM 1869 were used for SLV Cystine and Cysteine are not separated

Supporting Data

• Performance Characteristics:

Analytical Range: Calibration range Low level calibration standards: 0.1-0.5 mg/100 mL High calibration standards: 2-10 mg/100 mL

2

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

LOQ: Calculated using standard deviation of low amino acids area response and slope of the calibration curves. All LOQ met SMPR

Accuracy/Recovery: Total amino acids analysis

Accuracy was determined by analyzing NIST SRM 1849a and 1869. Results for all amino acids were within reference range and within 97%-100% from respective mean Reference Mass Fraction Value. The only exception was result for Tyrosine in NIST SRM 1849a. The results met SMPR Accuracy was also confirmed by analyzing lactose spiked with amino acids at 3 levels. Recoveries ranged from 93-107 % meeting SMPR. The only exception were lower recoveries for Methionine. Actual spiking levels are not listed Free amino acids analysis Accuracy was evaluated by analyzing NIST SRM 1869 spiked at 3 levels. Recoveries met SMPR Precision (RSD r ): Total amino acids analysis Repeatability (RSD r ) and Intermediate Reproducibility (RSD iR ) were calculated for 17 matrices. RSD r ranged from 0.4 to 2.7, RSD iR ranged from 1.5 to 4.8. All results met SMPR Free amino acids analysis Repeatability was evaluated by analyzing SRM 1869 spiked at 3 different levels. The data for RSD r met SMPR Reproducibility (RSD R ): NA

• System suitability:

System suitability requirements and quality control procedures are listed

Recommendation:

I recommend the method for determination of total proteinogenic amino acids and taurine for First Action AOAC Official method. More SPIFAN matrices or matrices containing free amino acids should be evaluated before recommending the method for free amino acids analysis.

3

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method AMINO-04 Title: Determination of free and/or total proteinogenic amino acids and taurine by pre-column derivatization and UHPLC Author: George Joseph and Asha Varughese Reviewer Name: Reviewer 2 Summary of Method: The method can be able to determinate free and / or proteinogenic amino acids from Infant Formula and Adult and Pediatric Nutritional Formulas by different sample preparation. 1- Method for determination of total proteinogenic amino acids and taurine The analytical method is capable of quantitative determination for 22 standard amino acids except tryptophan which is destroyed during the acid hydrolysis. Asparagine is determined as aspartic acid and glutamine as glutamic acid. The cystine and cysteine are converted to S-2-carboxyethylthiocysteine (CYSx) and the derivative is separated from the other amino acids The samples are hydrolysed in 6M HCl for 24 hours at 110±5°C in presence of phenol, 3-3’- dithiodipropionic acid (DTDPA) and internal standard norvaline. Phenol is added to prevent halogenation of tyrosine. DTDPA is added to convert cystine and cysteine to S-2-carboxyethylthiocysteine (CYSx) and quantified against derivatised calibration standard. After neutralization, amino acids and CYSx are derivatized with 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate (AQC) and are separated using reverse phase UHPLC with UV detection at 260nm. The analytical method is capable of quantitative determination for 24 standard amino acids including tryptophan. The amino acids cystine and cysteine are not separated and are quantified together as total. Citrulline which is present in some matrices is separated from other amino acids and included in the method performance evaluation Samples are dissolved in 0.5 M HCl along with added internal standard. The sample is then derivatized with 6-aminoquinolyl N hydroxyl succinimidyl carbamate. Amino acids are separated using quaternary pump Dionex UHPLC and detection by UV absorbance at 260nm. Norvaline is used as internal standard in the quantification of amino acids. 2- Method for determination of free aminoacids The method is suitable for the determination of free L-α-amino acids and taurine.

1

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Method Scope/Applicability: This method is applicable to infant and adult/pediatric nutritional formulas and other matrices such as infant cereals and pet foods.

General comments about the method: The method is very sensitive and can go down to approximately ten times lower than the SMPR requirements. The sensitivity of method is a direct reflection of its signal to noise ratio which ensures guaranteed method performance at the lower levels of amino acids in these matrices. The method offers baseline separation of citrulline which is an alpha amino acid generally present in Infant Formula and Adult / Pediatric Nutritional products. The separation of citrulline eliminates the risk of interference of this compound with other amino acids. Hydroxyproline is an important component of collagen and often this amino acid is used for quantitation of collagen .

Method Clarity: The method is clear, well written and easy to follow.

Method Safety Concerns: Not special safety concerns are needed. The analyst may be familiar with safety practices for all the chemicals and equipment utilized.

Pros/Strengths: • The method is simple and does not include any proprietary chemicals or instruments and can be performed on any basic UHPLC system. • The method incorporated an optional fluorescence detector as a precautionary measure to manage matrix interference and confirmation. • The method allows the determination of several amino acids in one single analysis. Supporting Data • General Comment : Certified Reference Materials NIST SRM 1849a and 1869 were used for the evaluation of method accuracy and precision. A sample of Whey Protein Concentrate and a sample Goat Milk based Infant Formula were also included in the validation. A total of 17 matrices were used for the single laboratory validation study.

2

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Method Optimization : Reagent and calibration blanks were analysed in all analytical batches to ensure that responses corresponding to reagent blanks are clearly identified in the multi-analyte chromatogram. There were no significant interferences observed at the retention time of target amino acids. The calibration and reagent blanks were prepared and analyzed at frequent intervals in all analytical sequences to ensure the selectivity of the method.

• Performance Characteristics: Analytical Range :

SMPR : for all analytes 0.4 – 2500 mg/100g

In both methods, a linear dynamic ranges (LDR) of calibration for all the amino acids were established with five levels of working standards. Freshly prepared working standards were used in all the analytical sequence of the validation and the levels of all the amino acids in all matrices were well within the calibration range. The correlation coefficients (r2) of all the calibrations during the validation were not less than 0.999. LOQ: SMPR: ≤ 0.4 mg/100g

The LOQ for free amino acids are between 0.01 and 0.06

The LOQ for proteinogenic amino acids are on sample as is basis between 0,02 and 0.17 and for reconstituted basis are between 0.002 and 0.017.

Accuracy/Recovery: SMPR: for all analytes in mg/100g

0.4 – 5.0 ±10% 5.0 – 50 ±8% 50 – 2500 ±5%

NIST SRM 1849a and 1869: 97 to 110%

1- Method for determination of total proteinogenic amino acids and taurine Accuracy of the method was validated using standard reference materials (NIST SRM 1849a and 1869) and spiked recovery studies. The NIST reference materials (SRM 1849a and 1869) were analysed in duplicate on six separate days (12 independent results). 2- Method for determination of free aminoacids Accuracy of the method was validated using standard reference material (NIST SRM 1869) and spiked recovery studies. The NIST reference material (SRM 1869) was analysed on three separate sequences. The methionine results for the reference materials were within 90 to 110%.

3

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Precision (RSD r ): SMPR: for all analytes in mg/100g

0.4 – 5.0 ≤ 4% 5.0 – 50 ≤ 3% 50 – 2500 ≤ 2%

1- Method for determination of total proteinogenic amino acids and taurine

Repeatability was determined in 17 different matrices in duplicate on each of the 6 days by two different analysts. The RSDr was calculated on all the 12 data points (6 sets) on all the 17 SLV matrices. The highest mean RSDr of all matrices was observed for CYSx and was not more than 2.7% and is well below the lowest set limit of 3% as per SMPR.

2- Method for determination of free aminoacids

Repeatability was determined by analysing the laboratory recovery samples prepared for accuracy validation. The %RSD was calculated on each levels and the mean RSDr of all the three levels for all the free amino acids were below the lowest set limit of 3% as per SMPR. The system precision of the analytical procedure was determined using six replicate injections of a mixed working standard which consists of all the target amino acids. The relative standard deviation of area ratio and retention time was found to be not more than 1%.

Intermediate Reproducibility (RSD iR ):

1-Method for determination of total proteinogenic amino acids and taurine

Intermediate reproducibility (within laboratory reproducibility) was calculates from the six sets of duplicares of 17 different matrices using one way Analysis of Variance (ANOVA). The highest mean RSD iR for all matrices was observed for taurine (0.48) and is bellow the set limit of 5% on SMPR.

• System suitability: The system precision of the analytical procedure was determined using 6 replicate injections of a mixed working standard which consists of all the target amino acids. The relative standard deviation of area ratio and retention time was found to be not more than 1%.

Recommendation:

I recommend move the method to first action.

4

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method Fluor-03

Title: Determination of Fluoride in Infant and Adult/Pediatric Nutritional Formula by Ion Chromatography with sequentially suppressed conductivity detection Authors: Theresa Steurer, Dr Elke Suess, Dr Christian Emmenegger, Dr Michael Klein, Dr Hari Narayanan Reviewer Name: Reviewer 1 Summary of Method: This method determines the quantity of fluoride by ion chromatography with sequential suppression conductivity detection. Fluoride is extracted from the sample into water. Inline dialysis is used to extend column lifetime and sequential suppression with a dose-in gradient is used to increase sensitivity by removing matrix interferences. The chromatographic peak height for fluoride in the sample is compared to a calibration curve prepared in water to determine the concentration. Method Scope/Applicability: Quantitation of dissolved fluoride in infant, pediatric, and adult nutritional formulas. Demonstrated performance in powders and ready-to-feed liquids. General comments about the method: This method was reviewed in a general sense by the Nutrients ERP in March 2019. Several recommendations were made to the method authors, and additional information and experimental results were included in the write up. Method Clarity: Overall the method is written clearly and easy to follow. Pros/Strengths: • Sample preparation approach is simple, involving extraction into water with vortexing. • Option provided for use of dialysis to extend column lifetime. Cons/Weaknesses • The peak resolution and shape in the samples is significantly different than in the calibrants, and the authors recommend using a horizontal baseline with peak height approach for integration. This approach could become problematic during MLT with inexperienced users. However, we have observed similar types of chromatographic challenges when analyzing infant formula for fluoride by a similar approach. • The instrument set up seems complex and finding laboratories to participate in an MLT using this or similar equipment may pose a challenge. Supporting Data • General Comment: The method authors made an attempt to utilize all SPIFAN matrices, including 12 that contained measurable levels of fluoride and 7 that did not. All samples were used for spike recovery studies to demonstrate performance in the broad range of matrices.

1

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Method Optimization: Authors have demonstrated that the inline dialysis system is not required for method performance, but is used to enhance column life by removing high molecular weight substances that are co-extracted with the fluoride. Authors have done many spike tests to exclude potential interferences from various other anions (chloride, acetate, lactate, formate) as well as disaccharides (lactose, lactulose) and mixtures of sugars, organic acids, and inorganic ions without affecting recovery or observed coelutions. Authors have provided supporting references for use of horizontal baseline and peak height approach for integration. Analytical Range: 10 to 3000 mcg/100 g, broken into two ranges for higher accuracy (10-200 mcg/100 g for low, 200- 3000 mcg/100 g for high) SMPR = 30-200 mcg/100 g LOQ: 3.6 ± 0.8 mcg/100 g calculated over matrix (SRM 1849a and SPIFAN Child Formula Powder) signal- to-noise ratio. Well below SMPR SMPR = 30 mcg/100 g

• Performance Characteristics:

Accuracy/Recovery:

Spike recoveries in powders was 80-105%; RTF was 103-108%. Lowest recovery found for spike of SRM 1849a (all spike levels: 40, 80, 160 mcg/100 g) Spike recoveries in SPIFAN kit samples at 50 and 100 % native levels was 91-120%. Lowest for FOS/GOS powder (both levels 91%) and highest for high protein RTF (117-120%). SMPR (30-100 mcg/100g) = 80-120%, (>100 mcg/100g) = 90-110%

Precision (RSD r ):

Repeatability determined on spiked samples of SRM 1849a (3.8%), ready-to-feed liquid (2%), and adult ready-to-feed (0.8%) SMPR (30-100 mcg/100g) = 8%, (>100 mcg/100g) = 5%

Reproducibility (RSD R ):

Intermediate precision determined on all samples in duplicate on 6 days by 2 analysts with 2 different columns. 5-12 % RSDR observed. Highest for powder sample C1.1 Child Formula Powder (9.4%) SMPR (30-100 mcg/100g) = 15%, (>100 mcg/100g) = 10%

• System suitability:

3 blanks are used to equilibrate the system with column gradient then again between calibrants and check standards at beginning of run to check for carry over. Should be no carryover. Unclear

2

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

if blanks are injected with every set of check standards or only with first – should be clarified in final method document. 4 check standards of fluoride in water (5, 10, 160, 500 mcg/kg) are injected in duplicate between calibrants and samples, and should be repeated every 12 h and at the end of each analytical batch. During validation, recoveries of check standards were 99-113% with RSDs of recoveries between replicates of the same check standard were 4-9% Check standards are also used to confirm stability of the dialysis transfer time ( ± 5 s). Fluoride retention time should be 7.5 ± 0.7 min. Recommendation: This method demonstrated performance that meets SMPR 2014.016 for fluoride in infant and adult/pediatric nutritional formulas. While I have reservations about the quality of the separation including the peak evaluation approach (horizontal baseline and peak height), the authors have done due diligence to explore potential problems and demonstrated method performance. As a result, I recommend this method for Official First Action.

3

Reviewer 2 – Fluor-03

AOAC SPIFAN ERP Reviewer Form OMA#/Nutrient Code Fluor-03 Method Title

Determination of fluoride in infant formula and adult/nutritional formula by ion chromatography with sequentially supressed conductivity detection. Fluoride is extracted from the sample matrices by mixing with ultrapure water. The samples are mixed and analyzed directly using Inline Dialysis with a cellulose acetate membrane. The concentration of fluoride is determined from the polynomial regression of the fluoride in the standards compared to the fluoride response as peak height detected in the samples. To enable the detection of low quantities the signal to noise ratio is effectively reduced using sequential suppression. The application of a Dose-in gradient enables the column clean-up from negatively charged organics. Carry-over is minimized by an automatic thorough cleaning step of the equipment including the dialysis cell. Theresa Steurer et al. The method authors should be acknowledged for their efforts in having addressed the comments made by the ERP at the previous SPIFAN meeting. The proprietary nature of the dialysis cell and the risk of chromatographic interference on integration are of some concern Fluoride in Infant and Adult/Pediatric Nutritional Formula

Method Author

Summary of Method

Method Scope/Applicability General Comments about the method

Method Clarity

Good.

Pros/Strengths

A chromatographic method is a modern approach to the challenges of fluoride analysis. Despite trials to identify potential interferences, the fluoride peak remains unresolved at the baseline. Using peak heights does mitigate some of the concerns on this point. The method as written requires the use of patented in-line dialysis, which renders method proprietary. However, the authors suggest the it can be removed and not undermine confidence in method performance, although this was not demonstrated for samples. If the inline dialysis is unnecessary for the method to perform as validated, we suggest this step is removed from the protocol. However, to date there is insufficient data to demonstrate the equivalence of the method with and without the dialysis unit. The use of height is an acceptable alternative to area for peaks that are not baseline resolved. Not clear what analytical range is, however standard curve exceeds the limits specified in SMPR. Values of 3.1 or 4.5 ug/100g, meets the SMPR <30 ug/100g. 1 out of 4 samples tested (117%, 120%) did not meet the SMPR for recovery requirement of 90- 110%. 1 out of 12 samples tested (9.4% RSDr) do not meet the SMPR for repeatability of< 8%. chromatogram is illustrative of challenges in obtaining baseline resolution. Is this typical, or best-case or worst-case? How does this compare to the chromatograms for SPIFAN kit samples? Examples to review would be desirable.

Cons/Weaknesses

General Comment(s)

Method Optimization

Analytical Range

LOQ

Accuracy/Recovery

Precision (RSDr)

Reproducibility (RSDR) No reproducibility data available. System Suitability Illustration of the NIST SRM sample

Recommendation

First Action could be recommended provided the authors demonstrate that the patented dialysis unit is not necessary for the successful application of the method to samples.

File Upload

Create your own form today!

Powered by

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

Expert Review Panel for Infant Formula and Adult Nutrition Evaluation of Method Fluor-04

Title: Fluoride Determination in Milk, Soy, and Water-Based Products Using Ion Selective Electrode and Direct Measurement Technique Author: Renée Erney and (Charles) Pete Black, Abbott Nutrition Reviewer Name: Reviewer 1 Summary of Method: A six-point standard curve is prepared, over a concentration range of 0.02 – 2.0ppm, in a matrix of 1% v/v conc. HCl and 0.5M sodium citrate. The response of each standard (in mV) is recorded after allowing the electrodes to stabilize under specific time and mixing conditions. A calibration curve is prepared using log(mV) v concentration (ppm). Powder samples are prepared using the standard 25g + 200g water dilution; RTF samples are used as is. 50g of such solution is treated to match the matrix described for the standards, the HCl being added to solubilize inorganic F, and the citrate buffer to standardize pH. The response of the unknown is measured under the same conditions as the standards and is back-calculated using the calibration and dilutions to express ppm F in the original sample. Method Scope/Applicability: General comments about the method: Method Clarity: No issues with the way in which the method is written. Method Safety Concerns: No issues. Pros/Strengths: • No special equipment required beyond ISEs. • Different ISE combinations were used with consistent results. Cons/Weaknesses • See below. Supporting Data • General Comment: Information was taken from the SLV document dated 2 Aug 2019. As described below, the LoQ of the method was determined to be well below that of the SMPR; nonetheless 4 of the matrices A total of 14 fortified (i.e. non-placebo) SPIFAN matrices were analyzed.

1

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

were still below its quantification limit; of the 10 remaining samples, 8 fell below the SMPR analytical range.

The results for all samples that were successfully quantified were used in the SLV summary data.

Method Optimization:

• Performance Characteristics:

Analytical Range: SMPR: Range: 30 – 200 mcg/100g reconstituted powder

LOQ: SMPR: 30 mcg/100g reconstituted powder  Method PLOQ estimated at 4 mcg/100g reconstituted powder based on lowest standard concentration in standard curve (0.02ppm) and typical sample dilutions.

Accuracy/Recovery: SMPR Requirements:

 30 – 100 mcg/100g reconstituted powder: 80 – 120%  > 100 mcg/100g reconstituted powder: 90 – 110%

Spiking was performed on 7 sample matrices, 1 of which was below the PLOQ, 4 of which were below the SMPR analytical range, and 2 of which were in the lower analytical range. Spiking was performed at 2 different levels (corresponding to the SMPR ranges), was repeated three or more times, was performed by different analysts, and the assays performed using different electrode configurations. Recovery results were reported as a grand mean for each sample (rather than results by spike level) and fell within the range of 93.5 – 109.9%, meeting the SMPR requirements.

Precision (RSD r ): Requirements:

 30 – 100 mcg/100g reconstituted powder: ≤ 8%  > 100 mcg/100g reconstituted powder: ≤ 5%

As stated, of the 14 samples tested, 10 produced reportable results, 2 of which were in the SMPR analytical range. Samples were analyzed between 9 and 15 times each, with most samples analyzed 10 or 11 times. RSDr values ranged from 1.7 – 4.6%, meeting the SMPR requirement, even though the F content of most samples was below the SMPR analytical range. The validation report indicated that liquid samples needed additional homogenization before they could meet the SMPR requirement, and that many powder samples were also treated similarly.

Reproducibility (RSD R ): Requirements:  30 – 100 mcg/100g reconstituted powder: ≤ 15%  100 mcg/100g reconstituted powder: ≤ 10%

2

FOR EXPERT REVIEW PANEL USE ONLY NOT FOR DISTRIBUTION

RSD IR values ranged from 3.1 – 11.3%. These values met the SMPR requirements, as their concentrations were in (or below) the lower concentration range of the SMPR. • System suitability: The correlation coefficient for the standard curve before must be greater than 0.995; calibration must be checked every 2 hours and is considered stable if within 2% of the initial mV reading. Recommendation: There appear to be several variables which can influence the response of the electrode to standard / sample solutions, including: o The amount of time the electrode is exposed to the solution before the response observation is made; o The rate at which the solution is stirred:  Incorporating oxygen into the sample can make the response unstable;  The heat produced by the spin plate can change the temperature of the solution, and affect the electrode response; o The way in which acid is added to samples, causing clumping (probably localized protein denaturation); o The degree to which the sample preparation (reconstitution or RTF aliquot) is homogenized: whether further treatment is needed. Lastly, it was noted that Adult Nutritional High Protein and High Fat liquid matrices were diluted 1:3 with water prior to analysis: however it was not explained on what basis this was necessary, or how it would be known for other matrices whether or not this treatment was needed. The nuance of each of these characteristics is likely well understood by the method authors, and when new analysts are trained or the method is set up in a new laboratory the details are conveyed. These points notwithstanding, the method met the SMPR requirements for LoQ, analytical range, precision, and accuracy, and the analytical community would benefit from its publication:

MOVE FOR APPROVAL AS FIRST ACTION.

3

Reviewer 2_Fluor-04 AOAC Expert Review Panel: First Action Method Review Reviewer Information Required to validate your review.

Remember, the main purpose of your review is to ensure the method conforms to the applicable SMPR. View and download most SMPRs here:

SPDS SMPRs SPSFAM SMPRs

Method Review Title of Method * Fluoride Determination in Milk, Soy, and Water‐based Products Using Ion Selective Electrode and Direct Measurement Technique

AOAC Candidate Method Number (e.g. ALN-01) * Fluor-04 Applicable SMPR *: SMPR 2014.016

I. Summary of the Method I. Summary of the Method *

Fluoride is determined quantitatively using ion selective electrodes. Samples and standards are treated with concentrated hydrochloric acid to help solubilize suspended inorganic material, and sodium citrate buffer is added to adjust the pH and the ionic strength. The fluoride concentration is determined using an external calibration technique II. Review of the Method Only 1. Does the applicability of the method support the applicability of the SMPR? If not, please explain what is missing. * 13 SPIFAN samples and one SRM were measured. Its applicability meets the SMPR requirement. 2. Does the analytical technique(s) used in the method meet the SMPR? If not, please specify how it differs from what is stated in the SMPR. *

Yes. There is no requirement of any specific technique in the SMPR. 3. Are the definitions specified in the SMPR used and applied appropriately in the method? If no, please indicate how the terms are used. * Yes 4. Does the method, as written, contain all appropriate precautions and warnings related to the method's reagents, components, instrumentation, or method steps that may be hazardous? If no, please suggest wording or option(s). * Yes, the method description contains a Safety Precaution section, in which the potential safety hazards are given. 1. Are the definitions specified in the SMPR used and applied appropriately in the supporting documentation (manuscripts, method studies, etc...)? If not, please explain the differences and if the method is impacted by the difference. * Yes 2. Is there information demonstrating that the method meets the SMPR Method Performance Requirements using the Reference Materials stated in the SMPR? If not, then specify what is missing and how this impacts demonstration of performance of the method. * There is no reference materials specified in the SMPR. The SRM 1869a has been tested in the SLV report, but there is no certified value available for an evaluation of accuracy. 3. Is there information demonstrating that the method performs within the SMPR Method Performance Requirements table specifications for all analytes in the SMPR applicability statement? If not, please specify what is missing and whether or not the method's applicability should be modified. * In the SLV report, Table 1 shows the repeatability and intermediate precision for 11 SPIFAN samples, including a SRM 1869a, in the concentration range from 4.85 ug/100g to 135 ug/100g. The repeatability results meet the SMPR requirement in different analytical ranges (30-100, and above 100 ug/100g). Table 1 also shows the accuracy obtained from spiking experiment. The accuracy results are within the 90-110%. However, the spiking levels in those samples were not shown in Table 1. Table II shows the linearity over the standard range of 2-200 ug/100g. III. Review of Information in Support of the Method III. Review of Supporting Information

Table 3-12 show the repeatability and intermediate precision for different samples. The repeatability results for these samples meet the SMPR. There is no data supporting the LOD/LOQ. However, based on the Table 1 results, a number of samples have fluoride concentration below 10 ug/100g, which is well below the SMPR requirement of LOQ at 30 ug/100g. The repeatability of the measurements of those low fluoride samples is quite good, indicating excellent sensitivity. IV. General Submission Package IV. General Submission Package 1. Based on the supporting information, were there any additional steps in the evaluation of the method that indicated the need for any additional precautionary statements in the method? * None 2. Does the method contain system suitability tests or controls as specified by the SMPR? If not, please indicate if there is a need for such tests or controls and which ones. * This method requires to establish separate calibration curve for each pair of electrodes, and a standard check must be done every 2 hours. The standard check reading must be within 2%. The SMPR requires to have a blank check samples, and two check standards at the lowest point and midrange point. The method could be modified to meet this SMPR. 3. Is there information demonstrating that the method system suitability tests and controls as specified in the SMPR worked appropriately and as expected? If no, please specify. * The system suitability tests work appropriately. 4. Based on the supporting information, is the method written clearly and concisely? If no, please specify the needed revisions. * Step E. 1. f) and g) needs some clarification. Specifically, what is the total volume approximately? The rest of the method is clear. 5. Based on the supporting information, what are the pros/strengths of the method? * The pros/strength of the method include: 1) It is simple, easy to use method. The sample preparation is simple. The device is easy to operate. 6. Based on the supporting information, what are the cons/weaknesses of the method? * The cons/weaknesses of the method is the potential interference:

Besides the temperature and the pH that could affect the results, cations, especially calcium and magnesium, could form insoluble salts with fluoride. Their solubility products (Ksp) are very low. The existence of these cations could reduce the free fluoride ion levels in sample and affect the results. 7. Any general comments about the method? * It seems that the temperature can affect the measurement results. It would be useful to investigate the effect of temperature variation in the ruggednesses testing. The recovery data from spiking experiments looks good, but it would be complete to have another approach to verify the accuracy of the method. This method measures the free form of fluoride, not the bounded form of fluoride. The SMPR does not specifically say the total fluoride, but it is believed that the total fluoride is required to be measured. Recommendation for the Method V. Final Recommendation Do you recommend this method be adopted as a First Action and published in the Official Methods of Analysis of AOAC INTERNATIONAL? Please specify rationale. * I am not ready to recommend this method be adopted as a First Action method because: 1) The Accuracy of this method need to be checked by a second approach in addition to the spiking approach. 2) The potential interference from cations (Ca2+, Mg2+, Al3+, etc.) need to be investigated. Insoluble salts could be formed and lead to reduced fluoride concentration. 3) Additional data on the LOD/LOQ assessment is needed.

Made with FlippingBook - Online catalogs