This page has been archived on the Web

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please contact us to request a format other than those available.

Skip booklet index and go to page content

Internal Quality Assurance Requirements for the Analysis of Dioxins in Environmental Samples

Canadian Cataloguing in Publication Data

Dioxin Quality Assurance Advisory Committee (Canada)

Internal quality assurance requirements for the analysis of dioxins in environmental samples

(Report ; EPS 1/RM/23)
Text in English and French with French text on inverted pages.
Title on added t.p.: Exigences internes d'assurance de la qualité pour l'analyse des dioxines dans des échantillons prélevés dans l'environnement.
Includes bibliographical references.
ISBN 0-662-59298-0
DSS cat. no. EN49-24/1-23

  1. Dioxins -- Laboratory manuals.
  2. Polychlorinated dibenzofurans -- Laboratory manuals.
  3. Polychlorinated dibenzodioxins -- Laboratory manuals.
  1. Canada. Environment Canada.
  2. Title.
  3. Title: Exigences internes d'assurance de la qualité pour l'analyse des dioxines dans des échantillons prélevés dans l'environnement.
  4. Series: Report (Canada. Environment Canada) ; EPS 1/RM/23.
TD196.C5D56 1992

©Minister of Supply and Services Canada 1992
Cat. No. En 49-24/1-23E
ISBN 0-662-59298-0
Beauregard Printers Limited

Return to Table of Contents

Reader's Comments

Any comments or inquiries concerning the contents of this report should be directed to:

Chemistry Division
Technology Development Branch
Conservation and Protection
Environment Canada
River Road Environmental Technology Centre
Ottawa, Ontario
KlA 0H3

Additional copies of this report are available from:

Environmental Protection Publications
Technology Development Branch
Conservation and Protection
Environment Canada
Ottawa, Ontario
KlA 0H3

Return to Table of Contents


The accuracy and comparability of data produced in different laboratories have always been of concern to laboratory analysts, researchers, and environmental program managers in both government and industry. A Dioxin Quality Assurance Advisory Committee (DQAAC) was established in 1990 to address this issue. This committee is comprised of scientists and scientific program managers from Environment Canada and other government departments who have current and direct experience in the areas of quality assurance in analytical laboratories and methodologies for analyzing polychlorinated dibenzo-para-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs). One of the main goals of this committee was to develop a common generic quality assurance protocol for dioxin analysis in support of federal programs.

This report outlines those elements of a laboratory quality assurance program that are considered essential to ensure the reliability of dioxin data. Performance criteria by which data quality can be assessed are also established in this report. By focusing on principles and performance rather than procedural details, laboratories have the flexibility to follow specific sample processing procedures of their own choosing, and the need to develop reference methods for individual environmental matrices is simplified or eliminated.

Note: Mention of trade names or commercial products does not constitute endorsement by Environment Canada.

Return to Table of Contents


The degree of agreement of a measured value with the true or expected value of the quantity of concern (Taylor, 1990).

Blind Sample
A sample submitted for analysis whose composition is known to the submitter but unknown to the analyst. A blind sample is used to test the proficiency of a measurement process (Taylor, 1990).

Calibration Standards
a set of solutions containing known amounts of native and carbon-13-labelled PCDDs and PCDFs. These solutions are used to establish the relationship between PCDD/PCDF concentration and MS detector response over the expected range of sample concentration.

Calibration Verification Standard
A calibration standard solution of intermediate level concentration (e.g., CS3), used to assess whether the initial calibration is still valid.

Certified Reference Material
A stable, homogeneous, and well characterized reference material, one or more of whose property values are certified by repetitive analysis by several operators and different methodologies in one or more qualified laboratories of known precision and accuracy. This material is used to assess the accuracy of a measurement process.

a member of a family of compounds that differ from each other only in terms of numbers and locations of common substituent atoms, or groups of atoms, on the parent compound. There are 75 PCDD and 135 PCDF congeners.

Control Sample
A reference material of known composition that is analyzed concurrently with test samples to evaluate the accuracy and/or precision of a measurement process (Taylor, 1990).

Column Performance Test Mixture
a solution whose components allow for performance testing of a specific Gas Chromatograph (GC) column in terms of some column characteristic that is critical to data quality. Chromatographic resolution of 2,3,7,8-TCDD and 2,3,7,8-TCDF from neighbouring isomers is assessed using such test mixtures.

Glassware Proof Rinse
The composite final solvent rinse of each piece of glassware intended for use in processing a batch of samples. Proof rinse samples are analyzed before sample processing begins.

A subset of congeners whose members are isomers one to another. For example, both 1,2,3,4-TCDD and 2,3,7,8-TCDD belong to the TCDD homologue group, whereas 1,2,3,7,8-P5CDD belongs to the P5CDD homologue group.

Instrumental Detection Limit
the smallest concentration/amount of analyte, in a solution containing only the analyte(s) of interest, that produces an instrumental response which satisfies all analyte detection and identification. criteria.

Internal Standard Quantitation
a quantitation procedure that corrects target analyte concentration data for processing losses, as well as normal variations in operator and instrument performance, in a single calculation. To use this procedure, samples must be spiked with known amounts of surrogate compounds (i.e., internal standards) before processing and a Relative Response Factor must be determined for each target analyte.

A member of a group of compounds that differ from each other only in terms of the locations of a specified number of common substituent atoms, or groups of atoms, on the parent compound. For example, there are 22 possible isomers of TCDD.

Method Blank
media (e.g., filter, solvent, water) spiked with surrogates and processed in an identical manner to actual test samples.

Method Detection Limit (MDL)
The smallest test sample concentration/amount of analyte that produces an instrumental response which satisfies all analyte detection and identification criteria when the sample is processed and analyzed according to the requirements of a specified test method. Reported MDL values reflect the composite effect of sample-related variables as well as method-related variables.

The degree of agreement between the data generated from repetitive measurements under specified conditions. It is generally reported as the standard deviation (SD) or relative standard deviation (RSD).

Quality Assurance (QA)
A system of activities whose purpose is to provide the producer or user of a product with the assurance that the product meets a defined standard of quality. The system consists of two separate but related activities, quality control and quality assessment (Taylor, 1990).

Quality Control (QC)
The overall system of activities whose purpose is to control the quality of a product so that it meets the needs of users (Taylor, 1990).

Recovery Standards
Selected compounds that are added to sample extracts immediately before instrumental analysis so that surrogate (internal standard) recoveries can be calculated. The two labelled compounds that serve as recovery standards for PCDD/PCDF analysis also serve as markers for identifying specific congeners on the basis of relative retention time.

Relative Response Factor (RRF)
The quotient of a target analyte response factor (instrument response per unit weight) divided by the response factor for its corresponding labelled surrogate. An RRF value remains constant over the range of concentration for which instrument response is linear.

2,3,7,8-Substituted Congener
A dioxin or furan congener with a minimum of four chlorine substituents at the 2,3,7, and 8 positions on the parent compound. There are seven 2,3,7,8-substituted PCDD congeners and ten 2,3,7,8-substituted PCDF congeners.

Compounds whose compositions and chemical properties are nearly identical to those of target analytes, but which are distinguishable from target analytes by some means of detection (e.g., MS). Isotopically-labelled dioxin and furan congeners are used as surrogates for their native analogues. Using surrogate compounds as internal standards allows target analyte data to be corrected for whatever degree of analyte loss has occurred during the course of sample processing.

Window Defining Mixture
A solution containing the earliest and latest eluting congeners within each homologous group of target analytes on a specified GC column.

Return to Table of Contents


Several draft versions of this report have been reviewed by the following members of the Dioxin Quality Assurance Advisory Committee (DQAAC):

  • A.S.Y. Chau, Chairman, Environment Canada, National Water Research Institute (NWRI);
  • D. B. Sergeant, Fisheries and Oceans, Great Lake Laboratories for Fishery and Aquatic Science (GLLFAS); and
  • M.A. Forbes, Environment Canada, National Water Quality Laboratory (NWQL).

Valuable comments and suggestions were also provided by members of C & P Laboratory Managers Committee (LMC) and the following scientists of government and commercial laboratories:

  • R.Turle and colleagues (Canadian Wildlife Service);
  • P. Kluckner (Environmental Protection, Pacific and Yukon Region, Environment Canada);
  • D. T. Williams, B. Lau, and J.J. Ryan (Health and Welfare Canada);
  • M.G. Foster and colleagues (Zenon Environmental Laboratories);
  • B.G. Chittim (Wellington Environmental Consultants Inc.);
  • R. Smythe (Walker Laboratories);
  • D.A. Sutherland (Canviro Analytical Laboratories Ltd.);
  • C. Chan (Mann Testing Laboratories Ltd.); and
  • D.H. Schellenberg (The Research and Productivity Council).

This report was prepared for DQAAC by the following staff members of the Technology Development Branch at the River Road Environmental Technology Centre (RRETC):

  • C. Chiu, R. Halman, and
  • G. Poole.

Return to Table of Contents

Section 1: Introduction

Polychlorinated dibenzo-para-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) are now known to be widespread environmental contaminants. They are found in such diverse matrices as industrial effluent, soil, sediment, water, ash, tissue, and ambient air samples. The detection limits reported for these toxic substances are generally in the low ppt to ppq range. It is clearly important that analytical data corresponding to such ultratrace levels be of definable and verifiable quality.

The objective of this report is to promote the generation of PCDD/PCDF data of consistently high quality by:

  1. specifying the essential elements of a Quality Assurance (QA) program that all laboratories generating PCDD/PCDF data in support of federal government programs will be expected to implement as a basic minimum;

  2. establishing criteria that both laboratory analysts and program managers can use to assess data produced;

  3. specifying the documentation that will be necessary for screening assessment of data quality in terms of QA support and performance criteria.

The application of these internal QA guidelines constitutes an essential part of a complete QA program. In this report, it is assumed that the laboratory has already demonstrated satisfactory performance in appropriate interlaboratory performance evaluation studies.

This report is applicable to the analysis of samples using capillary high resolution gas chromatography (HRGC) coupled with either low resolution mass spectrometry (LRMS) or high resolution mass spectrometry (HRMS).

Detailed analytical methods and QA/QC procedures for specific matrices are described in publications listed in the References of this report (DQAC,1989; Environment Canada, 1992; Environment Canada, 1990; U.S. EPA, 1990).

Return to Table of Contents

Section 2: Sample Handling and Custody

2.1 Containers and Sample Size

Sample Bottles
Sample bottles should be amber glass and have a Teflon-lined lid. Cleaned bottles should be screened for possible contamination before use. (Pre-cleaned and certified sample bottles for dioxin analysis are also commercially available. However, the user should ensure that levels of certification are suitable for levels of analysis.)

Solid Samples
Solid samples should be collected in wide-mouth bottles. Care should be taken to exclude as much water as possible. Biota such as fish, clams etc. should be wrapped in cleaned aluminum foil and frozen.

Liquid Samples
Liquid samples should be collected in narrow-mouth bottles. The level of the liquid sample in the bottle should be clearly marked just after collection to allow for assessment of possible sample loss during shipment.

When taking either solid or liquid samples, enough sample should be taken to allow for re-analysis, if necessary.

2.2 Shipping

  1. Each sample must be clearly identified and securely labelled,

  2. Samples (except tissue) must be maintained at 1 to 5°C in the dark from time of collection until extraction. Tissue samples should be kept frozen during shipment.

  3. Samples should be forwarded to the lab without unnecessary delay and the lab should be informed of the shipping waybill number and the anticipated time of arrival.

  4. A sample submission and chain of custody record sheet should accompany the samples. As a minimum, this form should include information identifying the project; the name, signature, and telephone and Fax number of the sample collector; the sample identification number, sample description or matrix; the date, time, and place of sampling; the analytical requirements and expected concentration; the name of the intended receiver; the storage conditions during shipment; the date and time of sample receipt; and the signature of the receiver. (See Figure 5 for a sample of the submission/custody sheet.)

2.3 Receiving

  1. All information on submission sheets should be verified and completed immediately upon receipt. The chain of custody portion of the form (see Figure 5) must be updated at this time. Any discrepancies, such as missing sheets, missing samples, mislabelled samples, damaged containers and/or samples, should be noted and reported to the client immediately.

  2. Each sample should immediately be labelled with a unique assigned lab code cross-referenced to the identifying field code. This code, plus a brief description of the sample, should be entered into the lab's sample log book and/or computer.

  3. Samples (except tissue) and sample extracts should be stored in a refrigerator at 1 to 5°C. Tissue samples should be stored in a freezer at -20°C.

Return to Table of Contents

Section 3: Method Performance Tests

  1. Before sample analysis, the laboratory must demonstrate the ability to achieve acceptable recoveries of native and surrogate PCDDs/PCDFs by conducting method performance tests as described in this section. These results must be documented, available for examination, and current (obtained within a 3-month period before sample processing begins). Three test samples of blank media, spiked with known amounts of native standards and labelled surrogates, should be prepared, extracted, and cleaned up according to normal procedures. Appropriate surrogate standard congeners and their spike levels for both LRMS and HRMS analysis are described in Section 4(i). The native congener spiking solution should contain all seventeen 2,3,7,8-substituted dioxin and furan congeners. Native congener spike levels should be selected to produce sample extract concentration levels equal to or near those of the appropriate (LRMS or HRMS) CS3 calibration standard solution (see Section 5).

  2. The performance test sample (e.g., filter, XAD, water, soil, ash) should be similar in size and nature to that of actual samples and either be free of target analytes or contain known trace levels of analytes that will not significantly affect the accuracy of native-congener spike recovery data.

  3. All surrogate recoveries for each of the three tests must be within the range of 40 to 120%.

  4. For each of the three tests, recovery for each of the native PCDD/PCDF congeners (corrected for surrogate recovery) must be within the range of 80 to 120% of the spiked value (i.e., accuracy of ± 20%).

  5. No samples should be processed until the method performance tests yield acceptable results. Tests must be repeated whenever extraction or cleanup procedures are modified; whenever there is a change in the lot number of chemicals used for cleanup; and whenever method performance test results are not current (within 3 months).

Return to Table of Contents

Section 4: Sample Preparation and Processing

  1. All glassware should be washed thoroughly with lab-grade detergent and high purity solvents. Following this, analysis of a composite glassware proof rinse (hexane and dichloromethane) sample for PCDDs/PCDFs and potential interferences is recommended.

  2. All solvents and reagents should be verified to be free of contamination that might interfere with the analysis of target compounds.

  3. Solid samples should be homogenized to ensure that any aliquot used for analysis is representative of the whole sample. Splitting of aqueous samples is not recommended.

  4. Sample size is dictated by the method in use and to some extent it is selected on the basis of desired detection limits. Recommended sample size, final volume, and achievable detection limits for various matrices are provided in Table 1.

  5. Extraction and cleanup procedures for use with various matrices can be selected by the laboratory. The lab must demonstrate, however, that these procedures can yield acceptable results for accuracy and surrogate recovery (Section 3).

  6. If possible, samples expected to be similar in analyte concentration range should be processed together in order to minimize the possibility of cross-contamination.

  7. Pertinent sample processing details should be recorded and made available for review if required.

  8. Unused portions of samples and sample extracts must be preserved in a refrigerator or freezer for tissue samples (Subsection 2.3c) for possible reanalysis if required. Samples and extracts must not be discarded without the permission of the client.

  9. Before processing, each sample must be spiked with known amounts of the carbon-13-labelled dioxin/furan surrogates shown in Table 2. The degree of analyte loss during sample workup is reflected in the percentage recovery of the spiked surrogates.

    If any surrogate recovery is outside the range indicated in Table 2, the sample should be reprocessed and reanalyzed, if possible. These ranges reflect the practical experiences of several government and commercial laboratories.

    Although individual surrogate recoveries as low as 30 or 40% will be considered acceptable, consistently low or highly variable recoveries may indicate that one or more of the sample processing procedures, or the GC/MS instrumentation, is not being effectively controlled.

    Method performance tests (see Section 3) should be repeated as the starting point of any investigation, and all investigative work should be fully documented.

  10. A method blank consisting of blank media (e.g., reagent water, filter, solvents) spiked with surrogates should be processed along with each batch of up to 10 test samples to demonstrate freedom from PCDD/PCDF cross-contamination and freedom from contaminants that would interfere with PCDD/PCDF analysis.

  11. One sample in every 10 test samples should be processed and analyzed in duplicate (supplied by laboratory QA/QC personnel as blind samples if possible) to assess reproducibility. Duplicate sample analysis can be replaced or supplemented by the analysis of a control sample (certified or uncertified reference material whose PCDD/PCDF content is well characterized). Analytical precision can then be assessed on a continuing basis. Control samples should also be submitted by QA/QC personnel as blind samples if possible.
Table 1: Recommended Sample Size, Final Volume, and Achievable Method Detection Limits for Various Matrices
 TissueSediment, Soil, Sludge, AshPulp Mill EffluentPulpDrinking WaterOilAmbient Air
Sample Size10 g
5 g
1 L12 g
12 L1 g1000 m3

Note : These method detection limit (MDL) values are bases on assumption of low processing losses (high recovery) and final extracts that are free from significant levels of matrix background interference. - (March 1992)

Table 2: Suggested Surrogate Standard Spike and Recovery Criteria
Surrogate StandardAmount Spiked (ng/sample)Acceptable Recovery(%)
LRMSHRMSTissueAll Other Matrices
13C12-2,3,7,8-TCDD2140 to 12030 to 130
13C12-2,3,7,8-TCDF2140 to 12030 to 130
13C12-1,2,3,7,8-P5CDD4140 to 12030 to 130
13C12-1,2,3,7,8-P5CDF*0140 to 12030 to 130
13C12-1,2,3,6,7,8-H6CDD4140 to 12030 to 130
13C12-1,2,3,4,7,8-H6CDF*0140 to 12030 to 130
13C12-1,2,3,4,6,7,8-H7CDD4140 to 12030 to 130
13C12-1,2,3,4,6,7,8-H7CDF*0140 to 12030 to 130
13C12-OCDD8240 to 12030 to 130
* Optional but desirable - (March 1992)

Return to Table of Contents

Section 5: Gas Chromatographic/Mass Spectrometric (GC/MS) Calibration and Quantitation

  1. Optimum settings for gas chromatograph parameters and correct retention time windows for time-sequenced selected ion monitoring mode (SIM) analysis of PCDDs/PCDFs should be established by analyzing a Window Defining Mixture containing the earliest and latest eluting congeners within each homologous group of congeners.

    This mixture should be analyzed at regular intervals for routine verification of window settings, and must be analyzed after any intentional change in GC parameter settings, and after any condition or upset that requires that the GC column be disconnected.

    As shown in Table 3, the order of elution on a 60-meter DB-5 column (polymethyl [5% phenyl]silicone) is such that five retention time windows can be defined, corresponding to the five levels of chlorine substitution (4 Cl to 8 Cl) without any overlap.

  2. For 2,3,7,8-TCDD and 2,3,7,8-TCDF analysis, a Column Performance Test Mixture containing the target analyte and its neighbouring isomers at equal concentration should be analyzed daily to confirm acceptable chromatographic resolution. It is recommended that these isomers be included in the Window Defining Mixture.

    The valley between peaks representing 2,3,7,8-TCDD and its closest neighbouring isomer should be equal to or less than 25% of the 2,3,7,8-TCDD peak height. The corresponding peak valley criterion for 2,3,7,8-TCDF is 30% maximum. These criteria must be satisfied whenever analyses are performed for regulatory compliance purposes. In other cases, results for either of these two congeners must be flagged if the corresponding criterion has not been met.

    The acceptable chromatographic resolution for 2,3,7,8-TCDD and 2,3,7,8-TCDF is shown in Figure 1.

    Figure 1: Acceptable Chromatographic Resolution for 2,3,7,8-TCDD and 2,3,7,8-TCDF
    Acceptable Chromatographic Resolution for 2,3,7,8-TCDD and 2,3,7,8-TCDF
    Click to enlarge

  3. The linear calibration range for each analyte should be established by running a series of calibration standards before initial sample analysis.

    Calibration standards should contain:
    1. all seventeen 2,3,7,8-substituted PCDD/PCDF congeners;
    2. the set of labelled surrogate congeners added to samples before processing; and
    3. the labelled congeners that are added to sample extracts just before GC/MS analysis.

    These latter congeners, 13C12-1,2,3,4-TCDD and 13C12-1,2,3,7,8,9-H6CDD, are used for calculating sample surrogate recoveries. In addition to their function as recovery standards, these compounds also serve as reference congeners for assigning sample peak identities on the basis of relative retention time.

    Recommended compositions and concentration levels for LRMS and HRMS calibration standards are presented in Table 4 and Table 5.

  4. An internal standard method is recommended for quantitation of sample data. Use of this method relies upon consistent linearity of MS response over the intervals between multi-point calibrations. The method is easily integrated into automated routines for data quantitation.

    Internal standard quantitation is based on the use of Relative Response Factors (RRFs). An RRF is the ratio of analyte response factor to the response factor of the corresponding labelled surrogate. These RRFs remain unchanged over the range of concentration for which MS response is linear. Using these RRFs, along with surrogate responses from the sample run, concentrations of PCDD/PCDFs can be calculated directly, without the need to calculate surrogate recoveries. Recoveries should nevertheless be calculated separately and reported, as these values serve to indicate the overall quality of the concentration data being reported.

    Isomer-specific RRFs are used for quantitation of 2,3,7,8-substituted congeners when isomer-specific analysis is required. For homologues represented by more than one isomer in the calibration standard solutions, the "homologue-average" RRF is used to quantitate all target analytes that are not 2,3,7,8-substituted congeners. If isomer specific analysis is not required, the "homologue-average" RRFs are used exclusively.

  5. From the initial calibration data, RRF values are calculated for each native analyte at each standard concentration level. The Relative Standard Deviation (RSD) of these RRF values for each native analyte must be less than 20%. If this criterion is met, then the calibration data is considered to be linear, and mean RRF values are used for quantitation of subsequent PCDD/PCDF data.

  6. The established calibration must be verified by analyzing a calibration verification standard (LCS3 in Table 4 and HCS3 in Table 5) at least once during every 12-hour period in which samples are being analyzed.

    Verification standard results are calculated in the same manner as an actual sample, using the mean RRF values obtained from the initial calibration. The calculated concentration for each native PCDD/PCDF congener must be within 20% of its actual known value. The calculated recovery of each surrogate compound must be within the range of 75 to 125%. Remedial action is required if any native or surrogate compound fails this verification test.

  7. Whenever new calibration standard solutions are prepared, and at least twice a year, the accuracy of the calibration standards should be verified against certified reference materials, e.g., National Institute of Standard and Technology, Reference Material for 2,3,7,8-TCDD concentration (NIST,1985).

  8. Sample components are identified as PCDD/PCDF only if GC/MS data satisfy the following criteria.

    1. Peak responses for each of the two selected molecular cluster ions must be at least three times the background noise level (S/N ≥ 3).

    2. For LRMS, the chlorine isotope ratio for the two molecular cluster ions must be within ±20% of the correct ratio. For HRMS, the criterion is ±15%.

    3. Peak maxima for the molecular cluster ions must coincide within two seconds.

    4. COCl-loss must be indicated when analyte concentration is sufficient for detection (for LRMS only).

    5. For combustion emission and industrial waste samples, monitoring for chlorinated diphenyl ether interference is recommended. Reported sample concentrations of PCDFs must be flagged with an "E" if any peak recorded in the chlorinated diphenyl ether channel maximizes within two seconds of the retention time of an apparent PCDF congener, and response for the ether ion exceeds 5% of the sum of the peak areas for the two monitored furan ions.

    6. Sample components are identified as 2,3,7,8-substituted congeners if the following applicable criterion is also satisfied.

      1. In the case of a congener for which a labelled analogue is present in the surrogate spiking mixture, all native and surrogate ion peak maxima must be coincident within three seconds.

      2. In the case of a congener for which a labelled analogue is not present in the surrogate spiking mixture, peak maxima for both molecular cluster ions must be coincident with each other within two seconds, and both peaks must be within three seconds of the expected retention time for that congener. Expected retention time is determined on the basis of known relative retention times for these congeners.

  9. For HRMS analysis, the mass spectrometer should be operated in a mass drift correction mode and must be tuned, using PKF (perflurokerosene), to achieve a resolution of at least 10 000 (10% valley definition). Hard copied verification of these measurements (peak shape and peak width) must be available (see Environment Canada, 1992 for calibration procedures).

  10. If the estimation of toxicity equivalents (TEQ) is required, the International Toxicity Equivalency Factors given in Table 6 should be used.
Table 3: Elution Order of PCDD/PCDF Window Defining Mixture on a 60-metre DB-5 Column
Homologue GroupFirst Eluting IsomerLast Eluting IsomerRetention Time Window* (min.)
* Listed times are for illustrative purposes only.
TCDD1,3,6,8-1,2,8,9-25.0 to 29.9
TCDF1,3,6,8-1,2,8,9-23.4 to 29.9
1,2,3,8,9-31.5 to 34.0
1,2,3,8,9-30.0 to 34.2
1,2,3,4,6,7-35.6 to 37.3
H6CDF1,2,3,4,6,8-1,2,3,4,8,9-35.1 to 37.8
H7CDD1,2,3,4,6,7,9-1,2,3,4,6,7,8-39.9 to 40.7
H7CDF1,2,3,4,6,7,8-1,2,3,4,7,8,9-39.5 to 41.3
OCDD  47.0
OCDF  47.3

Table 4: Composition of PCDD/PCDF Calibration Solutions for LRMS
PCDD/PCDFs Standardpg/μL
Native Standards
Recovery Standards
  • a also used to assess detection limits;
  • b used daily to verify calibration and abundance ratios;
  • c 13C-labelled penta-, hexa-, hepta-, and octa-CDF congeners are not used because they may interfere with PCDD analysis;
  • d retention time marker and recovery standard for tetra- and penta- homologues;
  • e retention time marker and recovery standard for hexa-, hepta- and octa- homologues

Table 5: Composition of PCDD/PCDF Calibration Solutions for HRMS
PCDD/PCDFs Standardpg/μL

Native Standards
Recovery Standards
  • a also used to assess detection limits;
  • b used daily to verify calibration and abundance ratios;
  • c 13C-OCDF is not used because it may interfere with OCDD analysis;
  • d retention time marker and recovery standard for tetra- and penta- homologues;
  • e retention time marker and recovery standard for hexa-, hepta- and octa- homologues

Table 6: International Toxicity Equivalency Factors (I-TEF) for Congeners of Concern
Congener of ConcernI-TEF
Source: NATO-CCMS, 1988

Return to Table of Contents

Section 6: Limit of Detection

  1. The Method Detection Limit (MDL) for PCDD/PCDF analysis is defined as the smallest concentration of analyte in a sample extract that will produce clearly defined peaks with an acceptable chlorine isotope ratio, and with a signal-to-noise ratio equal to three for the molecular cluster ion exhibiting the poorer signal-to-noise ratio. Reported MDL must be corrected for surrogate recovery and is calculated as follows:

    MDL Equation: MDL = [3N * (A/H) * Qs] / (As * RRF * S)

    estimated noise level expressed as peak height;
    area/height ratio for the surrogate standard peak;
    amount of surrogate standard added;
    surrogate peak area;
    relative response factor, as defined in Section 5d;
    sample weight or volume.

  2. The noise level for each homologue group should be determined from the actual sample chromatograms. Sample chromatograms (or portions thereof) should be "magnified" if necessary so that noise levels can be assessed and measured. An example of noise determination is provided in Figure 2.

  3. If detection of any potential 2,3,7,8-substituted congeners is rejected solely due to an incorrect isotope ratio, reporting requirements are as follows:
    1. congener presence is reported as NDR, instead of ND, in the concentration column of Figure 3;
    2. a concentration value for the congener is calculated in the same manner as if the isotope ratio had been within specification and this value is reported, in brackets, in the Detection Limit columns (but is not included in the corresponding homologue-total value);
    3. in Figure 4, the ion(s) used for quantitation are bracketed; and
    4. the actual isotope ratio is recorded, in brackets, in the adjacent column.

  4. Recommended Method Detection Limits for various sample matrices are listed in Table 1.

  5. A standard solution at near-detection limit concentrations (i.e., LCS 1 in Table 4 or HCS1 in Table 5) should be analyzed daily to assess and verify instrumental detection limits.

Figure 2: Noise Determination
Noise Determination
Click to enlarge

Figure 3: Sample Data Sheet - Part I
Sample Data Sheet - Part I

Figure 4: Sample Data Sheet - Part II (for isomer-specific analysis)
Sample Data Sheet - Part II (for isomer-specific analysis)

Return to Table of Contents

Section 7: Data Reporting

  1. Submission of sample results must be accompanied by a good photocopy of the corresponding sample submission/custody form. An example of this form is shown in Figure 5.

  2. For every sample analyzed, information on sample identity, sample size, GC column, MS model, concentration of target analytes, detection limit, number of isomer peaks identified for each homologue, amounts of surrogates added, and surrogate recoveries must be reported as shown in Figure 3.

  3. For isomer-specific analysis, ion abundance ratio and the deviation from expected retention time for each target congener must be reported as shown in Figure 4.

  4. Quality Assurance/Quality Control data that must be submitted with sample results include: results of the most recent method performance tests Figure 6; multi-point calibration data (Figure 7); calibration verification data (Figure 8); method blank results (Section 4j.); and any replicate or control sample results (Section 4k).

  5. All original sample and supporting QA/QC data must be available for auditing, if required. Documentation that could be subject to audit includes Column Performance (Section 5b) and Window Defining Mixture (Section 5a) chromatograms, control charts to document performance on control samples (if they are analyzed routinely), hard copies of data for standard solutions at near-detection-limit concentrations (Section 6e), hard copies of MS chromatograms for all samples (ion intensity ratios and checkmarks for positive analyte peaks should be recorded on chromatograms).

    All data should be saved on disk or tape for at least three years after contract completion. Written consent must be obtained from the client to dispose of any records.

Figure 5: Sample Submission/Custody Sheet
Sample Submission/Custody Sheet

Figure 6: Method Performance Test Data Sheet
Method Performance Test Data Sheet

Figure 7: Multi-point Calibration Data Sheet
Multi-point Calibration Data Sheet

Figure 8: Calibration Verification Data Sheet
Calibration Verification Data Sheet

Return to Table of Contents

Section 8: Data Evaluation

The following is a summary of the key performance achievements that would support confidence in the reliability of test sample results.

  1. The laboratory has demonstrated acceptable accuracy in analyzing method performance samples (Section 3d), certified standard solutions (Section 5g), and other certified reference materials as they become available.

  2. Duplicate or control sample results are provided, and demonstrate reasonable precision (Section 4k).

  3. All surrogate recoveries are within the acceptable range (Table 2).

  4. Reported method detection limits do not greatly exceed target values (Table 1).

  5. Method blanks have acceptable recoveries, contain non-detected or insignificant levels of analytes, and are free of interferences (Section 4j).

  6. Reported sample and calibration standard results are fully documented (Section 7).

Return to Table of Contents


Dioxin Quality Assurance Committee (DQAC), "Quality Assurance Protocol for CPPA National Dioxin Characterization of Canadian Bleached Chemical Pulping Operations", (January, 1989).

Environment Canada, "A Method for the Analysis of Polychlorinated Dibenzo-para-dioxins (PCDDs), Polychlorinated Dibenzofurans (PCDFs) and Polychlorinated Biphenyls (PCBs) in Samples from Incineration of PCB Waste", Conservation and Protection, Ottawa, Ontario, Report EPS l/RM/3 (1990).

Environment Canada, "Reference Method for the determination of Polychlorinated Dibenzo-para-dioxins (PCDDs) and Polychlorinated Dibenzofurans (PCDFs) in Pulp and Paper Mill Effluents", Conservation and Protection, Ottawa, Ontario, Report EPS l/RM/19 (1992).

National Institute of Standards and Technology (NIST), "Standard Reference Material 1614, 2,3,7,8-TCDD in Isooctane", U.S. Department of Commerce (1985).

NATO Committee on the Challenges of Modern Society (NATO-CCMS), "International Toxicity Equivalency Factor (I-TEF) Method of Risk Assessment for Complex Mixture of Dioxins and Related Compounds", NATO-CCMS Report No. 176 (1988).

Taylor, J.K., "Quality Assurance of Chemical Measurements", Lewis Publishers, Chelsea, Michigan (1990).

U.S. EPA, "EPA Method 1613: Tetra through Octa-chlorinated Dioxins and Furans by Isotope Dilution HRGC/HRMS", Revision A, U.S. EPA (april 1990).

Return to Table of Contents

List of Tables

Return to Table of Contents

List of Figures

Figure 1

The figure shows two chromatograms: First chromatogram is acquired on a DB-5 column and it displays four peaks in elution order corresponding to following isomers: 1,2,3,4-TCDD at 28:11 min, 1,2,3,7-/1,2,3,8-TCDD at 28:17 min, 2,3,7,8-TCDD at 28:25 min and 1,2,3,9-TCDD at 28:37 min. None of the peaks are baseline resolved: The valley between first two peaks is ~ 65%, between second and third peak ~25% and between third and fourth peak ~20%. The ratio between second valley X and the third peak height Y is less than 0.25. Second chromatogram is acquired on a DB-225 column and displays three peak in elution order corresponding to following isomers: 2,3,4,7- TCDF at 19:59 min, 2,3,7,8-TCDF at 20:15 min and 1,2,3,9-TCDF at 20:34 min. These peaks are not baseline resolved. The valley between first and second peak is ~ 15% and between second and third peak ~ 10%. The ratio between first valley X and second peak height Y is less than 0.3.

Figure 2

The picture shows a sequence of the chromatogram and the way how the signal S to noise N for the peak is calculated. S is the peak height and N is estimated by zooming the baseline close to the peak (e.g. 10 times). Ratio S/N must be higher than 3.

Equation 1

Left side of the equation: MDL (is equal to), Right side of the equation: Numerator: 3 multiplied by N multiplied by A/H multiplied by Qs, Denominator: As multiplied by RRF multiplied by S. All terms are explained in the text.

Return to Table of Contents

Date modified: