ISO/DTS 25364-1

• Status: Not Published
• ICS: 11.100.20 - Biological evaluation of medical devices
• Technical Committee: ISO/TC 194 - Biological and clinical evaluation of medical devices
• Drafting Committee: ISO/TC 194/WG 14 - Material characterization
• Current Stage: 5020 - FDIS ballot initiated: 2 months. Proof sent to secretariat
• Start Date: 28-Aug-2026
• Completion Date: 28-Aug-2026

Overview

ISO/DTS 25364-1: Chemical Characterization of Medical Devices - Part 1: Identification of Organic Extractables in Non-Targeted Analysis (NTA) is an emerging standard developed by the International Organization for Standardization (ISO) under Technical Committee ISO/TC 194. This document sets out best practices and structured guidelines for identifying organic extractables in medical devices using non-targeted analysis (NTA). The approach is science-based, focusing on objective, repeatable methods for determining substances that may leach from medical devices and potentially pose biological risks to patients.

The standard emphasizes a clear, stepwise system for classification and reporting of identified organic compounds, which supports toxicological risk assessment and strengthens compliance with regulatory requirements, particularly those related to ISO 10993-18.

Key Topics

• Identification Levels: The standard introduces structured identification levels, assigning numerical categories to the certainty of compound identification:

  • Level 0: Unknown (no information)
  • Level 1: Partially known (some structural attributes known)
  • Level 2: Tentative Identity (based on single analytical evidence)
  • Level 3: Confident Identity (corroborated by multiple data sources)
  • Level 4: Confirmed Identity (verified by a reference standard)

• Non-Targeted Analysis (NTA): NTA is defined as the process for discovering and identifying extractables without predetermined target compounds. It is a comprehensive approach using advanced analytical techniques to capture the full profile of organic substances present in medical device extracts.

• Analytical Procedures: The document outlines best practices for generating high-quality, information-rich analytical data-especially mass spectra-using methods like GC-MS and LC-MS. Emphasis is placed on spectral deconvolution, mass spectral interpretation, and robust data review by expert analysts.

• Reporting and Documentation: The identity and confidence level for each compound must be reported clearly, along with the evidence supporting identification. Thorough documentation ensures transparency and supports downstream toxicological risk assessments.

• Suspect Screening: While focused on NTA, the standard also references suspect screening analysis (SSA) - the process of checking for the presence of known suspect compounds using prior knowledge and reference data.

Applications

• Regulatory Compliance: This standard supports medical device manufacturers in meeting global regulatory requirements for chemical characterization, particularly those mandated by ISO 10993-18.

• Toxicological Risk Assessment: By providing a framework for robust identification of extractables, the standard enables accurate toxicological profiling of potential leachables, which is essential for patient safety assessments.

• Laboratory Best Practices: Analytical laboratories benefit from standardized processes that ensure repeatability, reproducibility, and consistency in chemical identification, supporting internal quality control and external audits.

• Product Development: Early integration of this identification process in product design mitigates risks associated with unknown or toxic extractables, ensuring safer medical devices and expedited regulatory approval.

Related Standards

• ISO 10993-18: Biological evaluation of medical devices – Part 18: Chemical characterization of medical device materials within a risk management process
• ISO 10993-17: Biological evaluation of medical devices – Part 17: Determination of allowable limits for leachable substances
• ISO 10993-1: Biological evaluation of medical devices – Part 1: Evaluation and testing within a risk management process
• ISO 14971: Medical devices – Application of risk management to medical devices

The implementation of ISO/DTS 25364-1 is a significant advancement for the medical device industry, ensuring safer products through standardized chemical characterization and supporting all stakeholders in aligning with international best practices. For laboratories, manufacturers, and regulatory professionals, applying this standard offers practical, traceable, and harmonized procedures for identifying and documenting organic extractables in medical devices.

FINAL DRAFT
Technical
Specification
ISO/TC 194
Chemical characterization of
Secretariat: DIN
medical devices —
Voting begins on:
2026-08-28
Part 1:
Identification of organic
Voting terminates on:
2026-10-23
extractables in non-targeted
analysis (NTA)
RECIPIENTS OF THIS DRAFT ARE INVITED TO SUBMIT,
WITH THEIR COMMENTS, NOTIFICATION OF ANY
RELEVANT PATENT RIGHTS OF WHICH THEY ARE AWARE
AND TO PROVIDE SUPPOR TING DOCUMENTATION.
IN ADDITION TO THEIR EVALUATION AS
BEING ACCEPTABLE FOR INDUSTRIAL, TECHNO-
LOGICAL, COMMERCIAL AND USER PURPOSES, DRAFT
INTERNATIONAL STANDARDS MAY ON OCCASION HAVE
TO BE CONSIDERED IN THE LIGHT OF THEIR POTENTIAL
TO BECOME STAN DARDS TO WHICH REFERENCE MAY BE
MADE IN NATIONAL REGULATIONS.
Reference number
FINAL DRAFT
Technical
Specification
ISO/TC 194
Chemical characterization of
Secretariat: DIN
medical devices —
Voting begins on:
Part 1:
Identification of organic
Voting terminates on:
extractables in non-targeted
analysis (NTA)
RECIPIENTS OF THIS DRAFT ARE INVITED TO SUBMIT,
WITH THEIR COMMENTS, NOTIFICATION OF ANY
RELEVANT PATENT RIGHTS OF WHICH THEY ARE AWARE
AND TO PROVIDE SUPPOR TING DOCUMENTATION.
© ISO 2026
IN ADDITION TO THEIR EVALUATION AS
All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication may
BEING ACCEPTABLE FOR INDUSTRIAL, TECHNO-
LOGICAL, COMMERCIAL AND USER PURPOSES, DRAFT
be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying, or posting on
INTERNATIONAL STANDARDS MAY ON OCCASION HAVE
the internet or an intranet, without prior written permission. Permission can be requested from either ISO at the address below
TO BE CONSIDERED IN THE LIGHT OF THEIR POTENTIAL
or ISO’s member body in the country of the requester.
TO BECOME STAN DARDS TO WHICH REFERENCE MAY BE
MADE IN NATIONAL REGULATIONS.
ISO copyright office
CP 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Geneva
Phone: [phone]
Email: [e-mail]
Website: www.iso.org
Published in Switzerland Reference number
ii
Contents Page
Foreword .iv
Introduction .v
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 General principles . 1
4.1 Identification in NTA .1
4.2 Identification process .2
4.3 Identification categories .2
4.4 Process recommendations .3
4.4.1 Generating analytical data for identification .3
4.4.2 Classification of identities .3
4.4.3 Securing identities .3
4.4.4 Corroborating evidence .4
4.4.5 Reporting identities .4
Annex A (informative) Discussion and description of the identification levels . 5
Annex B (informative) Identification process, suspect screening . 7
Annex C (informative) Identification process, non-targeted analysis, GC-MS (electron impact
ionization) . 9
Annex D (informative) Identification process, non-targeted analysis, LC-MS .15
Annex E (informative) Criteria for an acceptable spectral match in GC-MS and LC-MS.21
Annex F (informative) Good practices for structure elucidation .24
Annex G (informative) Corroborating information for identifications .30
Annex H (informative) Reporting the identity and identification status of a compound .34
Bibliography .36

iii
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards
bodies (ISO member bodies). The work of preparing International Standards is normally carried out through
ISO technical committees. Each member body interested in a subject for which a technical committee
has been established has the right to be represented on that committee. International organizations,
governmental and non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely
with the International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
The procedures used to develop this document and those intended for its further maintenance are described
in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the different types
of ISO documents should be noted. This document was drafted in accordance with the editorial rules of the
ISO/IEC Directives, Part 2 (see www.iso.org/directives).
ISO draws attention to the possibility that the implementation of this document may involve the use of (a)
patent(s). ISO takes no position concerning the evidence, validity or applicability of any claimed patent
rights in respect thereof. As of the date of publication of this document, ISO had not received notice of (a)
patent(s) which may be required to implement this document. However, implementers are cautioned that
this may not represent the latest information, which may be obtained from the patent database available at
www.iso.org/patents. ISO shall not be held responsible for identifying any or all such patent rights.
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and expressions
related to conformity assessment, as well as information about ISO's adherence to the World Trade
Organization (WTO) principles in the Technical Barriers to Trade (TBT), see www.iso.org/iso/foreword.html.
This document was prepared by Technical Committee ISO/TC 194, Biological and clinical evaluation of
medical devices.
A list of all parts in the ISO 25364 series can be found on the ISO website.
Any feedback or questions on this document should be directed to the user’s national standards body. A
complete listing of these bodies can be found at www.iso.org/members.html.

iv
Introduction
During their clinical use, medical devices either directly or indirectly contact tissue. As a result of this
contact, substances present in or on the medical device can be leached from the device during its clinical use.
The biological harms from these substances are evaluated using (quantitative) toxicological risk assessment,
which is driven by information about those substances. Chemical characterization per ISO 10993-18
provides that information by identifying and quantifying a patient’s exposure to leached substances. As
such characterization involve extraction of medical devices, extracted substances (extractables) are taken
as potential leachables.
ISO 10993-18 specifies a framework for the discovery, identification and quantification of extractable
constituents of a medical device, allowing the identification of biological hazards and the estimation and
control of biological risks from material constituents according to ISO 10993-17, ISO 10993-1 and ISO 14971.

v
FINAL DRAFT Technical Specification ISO/DTS 25364-1:2026(en)
Chemical characterization of medical devices —
Part 1:
Identification of organic extractables in non-targeted analysis
(NTA)
1 Scope
This document describes and recommends best practice for the identification of organic extractables or
leachables discovered during non-targeted analysis (NTA) of medical device extracts or leachates.
NOTE A closely aligned topic, suspect screening analysis (SSA), is also considered in this document, see Annex B.
This document is intended to be used in conjunction with ISO 10993-18.
2 Normative references
There are no normative references in this document.
3 Terms and definitions
For the purposes of this document, the following terms and definitions apply.
ISO and IEC maintain terminology databases for use in standardization at the following addresses:
— ISO Online browsing platform: available at https:// www .iso .org/ obp
— IEC Electropedia: available at https:// www .electropedia .org/
3.1
identified compound
organic compound which has been assigned a molecular structure and chemical name [and similar
identifying information such as Chemical Abstract Service Registry Number (CASRN)] based on the
interpretation of available analytical data
3.2
unidentified compound
organic compound which cannot be assigned a molecular structure and chemical name based on the
interpretation of available analytical data
4 General principles
4.1 Identification in NTA
The identity of an extractable or leachable is critical to toxicological risk assessment (TRA). The identity
allows the risk assessor to obtain information about the compound’s toxicological profile. Identification in a
NTA can be a challenging and subjective process.
The identification process in a NTA should:
— be standardized, reasonable and science-based;

— ensure repeatability, reproducibility and consistency;
— be practical and feasible for all compounds above the analytical evaluation threshold (AET);
— employ quality standards;
— produce a set of easily understood results (see Table 1);
— include how scientific evidence supporting identification is reported.
4.2 Identification process
ISO 10993-18 describes identification as assigning a molecular structure and chemical name (and other
designations such as CASRN) to a compound.
An identified compound has been assigned a molecular structure and chemical name. An unidentified
compound has not or cannot be assigned a molecular structure and chemical name.
Identification is based mainly on the compound’s mass spectrum. Additional analytical data, including
retention time (or index), accurate mass (empirical formula), also contribute to the identification process.
When the analytical data for the compound does not inform its identity, the compound is assigned to Level 0
(no identification possible).
When the analytical data for the compound partially informs its identity (e.g. general structural attributes
are indicative of a phthalate), the compound’s identity is partially known and is assigned to Level 1.
When there is a single piece of analytical data (e.g. the mass spectrum) which establishes the molecular
structure (e.g. by structure elucidation by an expert mass spectrometrist or by spectral matching to a
reference database), the compound’s identity is tentatively known and it is assigned to Level 2.
When there are two corroborating pieces of analytical data (see Annex G) which establish the molecular
structure, the compound’s identity is confidently known and it is assigned to Level 3.
When a Level 2 or 3 compound is subsequently verified by matching its analytical data with similar data for
a reference standard its identity is confirmed and it is assigned to Level 4 (the highest identification level).
A higher identification Level indicates a more definitive (i.e. information-rich) proposed identity. A higher
identification Level indicates the certainty of its true identity.
4.3 Identification categories
Table 1 — Identification classes and levels for NTA
Descriptive identification classes
Numerical classification
levels for reporting
General classes Detailed classes
Unknowna Level 0
Unidentified [3.2]
Partially known Level 1
b
Tentative identity Level 2
a
Identified [3.1] Confident identity Level 3
a
Confirmed identity Level 4
a
Consistent with class descriptions provided in Reference [8] and Reference [9].
b
“Tentative identity” as defined in Reference [8] and Reference [9] is inconsistent with “partially known.” USP terminology
should not be used.
These classification levels are discussed in more detail in Annex A.
The identification of organic extractables involves a systematic process of review and analysis of available
data. Such processes are described in Annex B, Annex C and Annex D.

4.4 Process recommendations
4.4.1 Generating analytical data for identification
Analytical methods should be optimized to produce interpretable and information-rich mass spectra.
Deconvolution processes should be used on experimental chromatographic data (when necessary) to
generate the purest spectra achievable. For example, by resolving co-eluting compounds that cannot be
chromatographically resolved (see Annexes C and D).
4.4.2 Classification of identities
The identity of a substance should be reported along with its classification. This communicates the
confidence in the identity (see Annex A for additional discussion).
4.4.3 Securing identities
Compounds should be identified by either mass spectral matching or by mass spectral interpretation
(structure elucidation) or a combination of both.
— Mass spectral matching:
— used alone provides a Level 2 identity for a compound if the spectral match is judged to be acceptable.
— can provide a Level 1 identity even with an unacceptable match if the available information supports
inferences concerning the compound’s structure or functionality.
— can be supported by expert review of available information, including the quality of the match.
— Quality is influenced by multiple, orthogonal and complementary factors (see Annex E). For example,
[10]
considering matches secured from the NIST/Wiley database , the following are recommended as
[11]
objective criteria for an acceptable match :
— match factor: >800 or equivalent
— reverse match factor: >800 or equivalent
— probability percentage: >50 %
— in lib score: >0
Three of these four criteria should be met for a match to be judged as acceptable. Alternate factors
and acceptance criteria may be used to establish the quality of a match if justified.
— Candidate identities with matches of sufficient quality should be subjected to expert review to consider
discrepancies or inconsistencies between the analytical and library spectra.
— The expert review should produce one of two conclusions:
— any discrepancies or inconsistencies between the analytical and library spectra are inconsequential
and thus that the match is acceptable and the candidate identity is reportable;
— any discrepancies or inconsistencies between the analytical and library spectra are sufficiently
consequential that the match is unacceptable and the candidate identity is rejected.
NOTE Mass spectral matching is inherently more effective (e.g. more likely to produce the correct identity)
when the experimental mass spectrum is matched against a database of mass spectral internally generated
by analysis of reference materials. Typically, this is a normal quality control procedure used by analytical
laboratories.
— For structure elucidation:
— expert analysis of the existing MS data should always be included (see Annex F);

— a Level 2 identity is secured, regardless of the rigor and completeness of the elucidation, unless it is
corroborated with additional information or analysis.
4.4.4 Corroborating evidence
Evidence should be corroborated to the practical extent possible to produce the highest classification and
which is most likely to be the true identity. Corroborating evidence should be sought or generated to produce
the highest Level (See Annex G).
4.4.5 Reporting identities
When reporting an identity, the evidence supporting the identity should be provided so that the identification
can be reviewed by potential users of this information (see Annex H).
In certain cases, the identification process can produce multiple Level 2 identities that are established and
justified to be acceptable. If expert review or corroborating evidence cannot establish a single identity as
the most likely candidate, then all the multiple identities should be reported.
The analytical chemist should consult with a toxicologist in the event of any doubt over an identity reported
for TRA.
Annex A
(informative)
Discussion and description of the identification levels
In the simplest sense, a compound can either be identified or remain unidentified. These two classes make
up the first tier of descriptive identification.
These two broad categories can be subdivided into sub-classes to provide greater granularity. Thus,
unidentified compounds can be categorized into two groups, a true unknown, meaning that no statement
can be made about its identity, or a partially known, meaning that some general statement can be made about
the compound’s structure. Similarly, identified compounds can be initially categorized into two groups, one
group where the identity is not corroborated with supporting data (tentative) and the other group where
the identity is supported with corroborating data.
Lastly, the sub-class “identity supported by corroborating data” can be further subdivided into two
secondary sub-classes depending on the type and quantity of corroborating data. If the corroborating data
includes a retention time (or index) and mass spectrum procured by analysis of an authentic reference
standard, then the identification class is “confirmed identity”. If the corroborating data are not secured by
analysis of a reference standard, then the identification class is “confident identity”.
Identification classes can be numerically assigned to Levels to eliminate the subjective nature of the
descriptive terms:
a) Unidentified: The circumstance where a molecular structure and chemical name cannot be assigned to
a compound.
— Level 0: Unknown: The circumstance where no structural information can be inferred from the
available information.
— Level 1: Partially known: The circumstance where only certain structural attributes or the elemental
formula of the compound can be inferred, generally with some (limited) doubt as to the correctness
of the structural or elemental formula assignment.
The value of securing a partially known identification is as follows. The ability to assign a compound
to a structural class (e.g. phthalate, siloxane, aliphatic hydrocarbon) can facilitate the decision as to
whether additional identification information is needed (e.g. an identified constituent). Furthermore,
the partially known information can be sufficient to support the conclusion that the compound of
interest is not of especially high toxicity or is not a “cohort of concern” compound.
b) Identified: the circumstance where a molecular structure and chemical name can be assigned to a
compound
— Level 2: tentative identity: an identity that is obtained and supported by a single step analysis of one
piece of information (i.e. mass spectral matching or structure elucidation based on the compound’s
analytical mass spectrum) and which lacks corroborating information, where there is some doubt
as to whether the identity is correct. The criteria for obtaining acceptable matches are discussed in
Annex E and the means of properly performing structure elucidation are discussed in Annex F.
— Level 3: Confident identity: an identity that is supported by two or more pieces of corroborating
information, where there is little or no doubt that the proposed identity is correct. A discussion of
corroborating information is contained in Annex G.
— Level 4: Confirmed identity: an identity that has been verified via matching of analytical data (e.g. the
mass spectrum and retention time or index) from the compound of interest to an analysed reference
standard, where there is essentially no doubt that the identity is correct.

In certain circumstances, the corroborating data supporting a Level 3 identity is either so plentiful or so
compelling that the identity is, for all intents and purposes, confirmed. Nevertheless, a Level 4 identity is
reserved for that circumstance of authentic reference standard match.

Annex B
(informative)
Identification process, suspect screening
Suspect screening is a process that involves establishing whether pre-specified substances are present in
an analysed sample. As suspects are pre-specified using prior knowledge, their key analytical properties
(such as retention time and mass spectrum) are known. Suspect screening is accomplished by examining
the experimental data generated by analysing a sample to establish if that experimental data contains a
response that has the same known analytical properties exhibited by the suspect (e.g. retention time and
mass spectrum). If the analytical data contains a response that can be matched to the suspect, then it has
been verified that the suspect is present in the analysed sample. If the analytical data does not contain a
response that can be matched to the suspect, then the suspect is not present in the analysed sample (within
the sensitivity limitations of the test method). Thus, suspect screening often involves matching experimental
data to analytical data contained in a database.
Suspects are established by prior knowledge. For example, extractables data can be used to establish
suspects which are specifically considered in a leachables study and an item’s compositional information
can be used to establish suspect extractables.
The suspects that are addressed in suspect screening are often substances with Level 4 identities whose
retention time and mass spectrum has been properly matched to those properties of a reference standard.
However, suspects can also include compounds from all the other identification levels, so long as the
diagnostic data for the suspects have been collected. For example, a compound with a Level 1 (partially
known) can be a suspect because its retention time and mass spectrum are known.
The process of securing an identity via suspect screening for organic extractables and leachables is shown in
Figure B.1 for gas or liquid chromatography coupled to mass spectrometric detection, gas chromatography
- mass spectrometry (GC-MS) or liquid chromatography - mass spectrometry (LC-MS). The process starts
with a suspect whose analytical properties (e.g. retention time and mass spectrum) are known. The
process proceeds by examining the experimental data generated by analysing a sample to establish if that
data contains a response that has the same known analytical properties exhibited by the suspect. If such a
response is found, then the presence of the suspect in the analysed sample is likely. The likely presence of
the suspect in the analysed sample is verified when the analytical match has been substantiated via expert
review.
When a suspect has been confirmed to be present in a sample, the identification level of the suspect in the
analysed sample is the same identification level that the compound had as a suspect. For example, when an
extractable is used as a suspect leachable and is confirmed to be present in a leachate, the leachable will have
the same identification level as the extractable. Thus, matching a leachable’s peak in an analysed sample to a
suspect extractable that had a Level 1 identity produces a leachable that also has a Level 1 identity.
To ensure that a compound is correctly identified, the two primary sources of structure identities, the
mass spectrum and the retention time, should be highly reproducible. Since the acquisition of GC-MS
spectra is highly standardized and is not significantly influenced by ionization competition effects (e.g. ion
suppression), these spectra are highly reproducible. Retention times, however, can shift for several reasons.
Therefore, control measures should be used to keep the retention times stable (e.g. retention time locking,
use of relative retention times) and the effectivity of these measures should be established by system
suitability testing.
The outcome of a suspect screening exercise is generally definitive because it is based on corroborating
data. However, as is the case with any process that involves matching, the outcome of suspect screening
should be reviewed from two perspectives:
a) The supporting mass spectral data should be reviewed by an expert mass spectrometrist to verify the
match between the analytical and the suspects’ reference spectra, and

b) The presence of the suspect in the tested sample should be reviewed considering all the available prior
knowledge (e.g. it can be necessary to verify if the proposed identity of the suspect is consistent with
the known chemical composition of the test article).

a
As necessary and appropriate to resolve co-eluting peaks.
Figure B.1 — Process flow diagram, identification process for suspect screening
by GC-MS and LC-MS
Annex C
(informative)
Identification process, non-targeted analysis, GC-MS (electron impact
ionization)
C.1 General
As opposed to suspect screening and target analysis, where suspects or targets have been predefined via
prior knowledge, NTA can be broadly defined as determining the chemical composition of a given sample
without a pre-defined list of suspects or targets. Thus, all compounds revealed via NTA identification efforts
are unidentified (until their analytical data has been interpreted) and unexpected.
In general, the identification process begins with the analysis of a sample and the generation of a
chromatographic peak which has, ideally, an interpretable and information-rich mass spectrum (and
retention time) associated with it. This mass spectrum is the foundation of the identification effort and can
lead to a compound being identified via two approaches, mass spectral interpretation and mass spectral
matching.
Securing a high-quality mass spectrum is facilitated when the mass spectrum is differentiated from, and
unobscured by, potential interferences. Although chromatographic selectivity is a primary means for
managing co-eluting substances, peak co-elution cannot generally be avoided and it can be challenging to
resolve peak responses sufficiently so that useful and uncompromised mass spectra can be obtained for
the coeluting analytes. When the analyte’s mass spectrum is compromised due to spectral contamination,
unreliable identifications can be produced. Typical sources of spectral contamination include ion signals
from coeluting compounds, column bleed, solvent tailing and electronic noise.
As visual inspection of complex chromatograms can be an ineffective means of resolving chromatographic
peaks and their associated mass spectra, the application of data processing techniques can be necessary.
When determining the mass spectrum that will serve as the basis for identification either through library
matching or mass spectral interpretation, or both, background subtraction should be the absolute minimum
approach. To support a higher quality of identification, however, application of a scrutinous deconvolution-
based approach can be necessary, as deconvolution delivers better quality mass spectra and thus reduces
the risk of misidentification.
C.2 Deconvolution
Deconvolution is the process of computationally extracting analyte signals from a complex mass
chromatogram, resulting in the elimination of background noise and the spectral separation of co-eluting
compounds. An example is given in Figure C.1.

NOTE The peak of interest (left above) is composed of two closely eluting components (mass spectrum, left
below), whose measured mass spectrum is deconvoluted into two high-quality mass spectra (right, above and below),
one for each analyte.
SOURCE Reference [12]. Reproduced with permission from the authors.
Figure C.1 — Example of deconvolution of a single signal peak detected in a
total ion current (TIC) chromatogram
The deconvolution process involves multiple steps such as noise analysis, peak shape analysis (ions
belonging to the same peak should have the same apex and peak shape) and the assembly of a deconvoluted
[12]
spectrum. It should be noted that deconvolution algorithms and related parameters can differ among
various software platforms and therefore it is possible to generate slightly different deconvoluted spectra
for the same raw data. Additionally, one should be aware that deconvolution does not provide clean mass
spectra in all cases, especially in complex chromatograms that consist of both small and large peaks or which
contain coeluting compounds that share the same m/z values. In such circumstances, the deconvoluted
spectrum can either be incomplete (deconvolution is carried too far and critical features are removed from
the mass spectrum) or contaminated (deconvolution was not carried far enough and interfering features
remain in the deconvoluted mass spectrum). As is the case with any means used to processing data, poorly
or improperly performed deconvolution can lead to misidentification or cause proposed identities to be
classified at a lower Level.
These circumstances reinforce the need for a mass spectrometry expert to review the data and in difficult
cases can require additional actions (such as the evaluation of other deconvolution settings, manual
deconvolution or other appropriate actions).
C.3 Mass spectral matching
Considering the process of identification in NTA, Figure C.2 illustrates the process workflow for identification
of a compound detected by GC-MS. Once the analytical data (mass spectrum) has been collected and
deconvoluted, there are two principle means of using the mass spectrum to propose a compound’s identity,
mass spectral matching and mass spectral interpretation. The concept of mass spectral matching was
discussed for suspect screening in Annex B. Mass spectral matching is conceptually identical in suspect
screening and NTA; the analytical mass spectrum is compared to reference mass spectra collected in a
library with the intent of uncovering library spectra that are comparable to (a good match to) the analytical
spectrum. Library compounds whose mass spectra match the analytical mass spectrum are candidate

identities. Presumably, the better the spectral match, the more likely it is that the candidate identity is the
true identity.
However, mass spectral matching differs significantly in suspect screening versus NTA. In suspect screening,
the matching occurs between the analytical mass spectrum and the known mass spectrum of the suspect.
The matching process in suspect screening therefore produces a binary and definitive (yes or no) outcome.
If there is a match, then it is concluded that the suspect is present in the tested sample. If there is no match,
then it concluded that the substance is not the suspect and thus further identification activities can be
warranted.
In NTA, the mass spectrum of the detected compound is matched with the mass spectra contained in an
external mass spectral library. The mass spectral matching likely produces an open-ended outcome. The
matching does not produce the one true identity; rather, it produces potential identities. Even if there is
only one good match, the circumstance of having a single match does not mean that the match is surely
the compound’s true identity. Furthermore, if there would be no match, it does not end the identification
process, but means that the identity need be secured by other means.
The library of mass spectra can be developed internally by analysing reference standards and collecting the
[12]
key analytical information (retention time, mass spectrum, response factor). Such a library is specific
in the sense that it contains information related only to extractables or leachables. Alternatively, a library
of mass spectra can be developed externally, typically by an independent organization that collects and
collates spectra from multiple sources which produce the spectra in varying ways. Such external libraries
are generic and general in the sense that they contain data for compounds which are not extractables or
leachables. Well-known and well-controlled large commercial reference libraries such as the NIST/EPA/NIH
[13][14] [15]
Mass Spectral Library and the Wiley Registry of Mass Spectra are widely used tools to facilitate
spectral matching of GC-MS data.
C.4 Expert assignment of identification levels
Considering the work process for spectral matching more closely (the right-hand side of Figure C.1), the
process begins with the comparison, generally machine-driven, of the analytical spectrum to the library
spectra, presumably producing a list of hits (compounds whose mass spectra are close matches to the
analytical spectrum). Should one (or more) of the hits meet the criteria for an acceptable match (discussed
in Annex E), the match is evaluated by an expert to confirm that the match is proper and to weed out false
identities (e.g. good mass spectral matches which cannot be the analyte of interest because of clear scientific
discrepancies between the library and analytical spectra). This expert review focuses on verifying the data
that supports the acceptable match but can involve some basic mass spectral interpretation (structure
elucidation), looking for fundamental inconsistencies between the library spectrum and the analytical
spectrum (e.g. an ion is missing in the analytical spectrum, compared to the library spectrum, indicating a
functional group can be missing in the proposed identity). Ultimately, this expert evaluation either rejects
the match (i.e. expert review finds one or more critical inconsistencies between the analytical and library
spectra) or confirms the match (i.e. expert review finds no critical inconsistencies between the analytical
and library spectra).
In the circumstance that the expert rejects the match as being unacceptable, identification proceeds in one
of two ways. Firstly, if no other acceptable candidates were revealed by spectral matching, then further
attempts at identification can only proceed forward as an exercise of mass spectral interpretation. If other
candidates were revealed by spectral matching (but were not taken as the lead candidate), then these other
candidates are also reviewed by the expert. Again, if the expert rejects the match or matches as being a
viable candidate, the identification efforts can proceed forward only as an exercise of mass spectral
interpretation. However, any spectral match identification the expert finds acceptable becomes a Level 2
identity as it is only supported by one piece of evidence, which is the spectral match. Additionally, a mass
spectrometrist can elevate the level of identification from a Level 2 identity to a Level 3 identity by including
additional structure supporting evidence such as retention time (or index) match, a consideration of prior
knowledge (e.g. the identification consistent with the known composition of the test article), a provable
relationship to a compound with a confident or confirmed identity, detected in the same analytical method
or by correlating the compound to a compound with a confident or confirmed identity which was detected
in an orthogonal analytical technique. As the identity would then be based upon two or more pieces of
independent information, the identity of the compound is corroborated and becomes Level 3.

In the event that two types of information are in conflict, experimental evidence should supersede the
rationalization of an identification from material information.
C.5 Confirmation of identity (Level 4)
C.5.1 Candidates that move forward with Level 3 identities are queried as to whether a Level 4 identity is
required or desired. If a Level 4 identity is required, the ability to procure a reference standard is ascertained.
If a reference standard can be procured, then this is an opportunity to verify the identity. Verification of the
identity has at least two desirable outcomes:
a) It inspires confidence that the identity is correct and that all appropriate efforts have been used to
secure and justify the identity,
b) It promotes the identified analyte from NTA to suspect screening (because the compound is placed in
the organization’s spectral library once the reference standard is analysed).
However, there are cases where the value of verifying an identity is less than the value of the resources
required to perform the verification:
— when the reference standard is not commercially available and generation of the reference standard is
difficult and costly;
— when a Level 4 identity is not necessary.
If a reference standard can be procured, then the reference standard is analysed to produce a reference
spectrum and a reference retention time (or index). If there is an acceptable match between the analytical
spectrum and retention time (or index) and the standard’s spectrum and retention time, then the Level 3
identity has been verified or confirmed and the identity is elevated to Level 4.
C.5.2 Although one presumes that verification of an identity by analysis of a reference standard is
successful (that is the identity is verified), such a desirable outcome is not always achieved. Given that the
proposed identity at this point is corroborated by both an acceptable spectral match and by the expert
review, one expects such negative outcomes to be rare (but not impossible). If the identity is not confirmed
by analysis of the reference standard, then the identity is rejected as false. At this point, one reviews the
available analytical data and decides
a) whether there is another candidate identity that can be taken through the expert review process or
b) whether the identification process need be re-started via structure elucidation (as spectral matching
produced only a false result).
Considering the importance of retention data as corroborating information, this topic is addressed in greater
detail in Annex G. While retention times are specific to a particular chromatographic method, retention
indices are more universally applicable as they are, in essence, normalized relative retention times using
linear n-alkanes for normalization. Most of a method’s operating parameters do not have a large effect on
the retention index, except for the nature (polarity) of the stationary phase.
The NIST spectral libraries augment the spectral data of reference compounds with different types of
retention indices. Experimental retention indices are curated from experimental data and are indicated per
type of stationary phase (semi-standard non-polar, standard non-polar and polar). In addition, computed
retention indices can be available.
C.6 Structure elucidation
Considering mass spectral interpretation (the left-hand side of Figure C.2), good practices for performing
structure elucidation are provided in Annex F. The structure elucidation process can have two outcomes: if
the structure elucidation fails to reveal a structure due to lack of interpretable information, the compound

of interest is classified as an unknown. If the available information can be interpreted, the outcome of the
interpretation is considered. Two outcomes are possible:
a) a structure can be proposed for the compound of interest, or
b) only the compound’s general structural features can be ascertained (e.g. the compound is established to
likely be “a phthalate”). In the latter case, the compound is classified as Level 1 while in the former case,
the compound is classified as having a Level 2 identity.
As it is generally desirable to either substantiate or verify a Level 2 identity, additional evidences are
collected and reviewed. If there are no additional evidences, then the identity remains unconfirmed at
Level 2. If one or more evidences can be collected, the identity is “elevated” to Level 3. Level 3 identities
are further reviewed to determine if they can be verified and elevated to Level 4 status via the process
d
...

ISO/TC 194/ WG 14
Secretariat: DIN
Date: 2026-05-0408-13
Chemical Characterizationcharacterization of Medical Devices —
medical devices —
Part 1:
Identification of Organic Extractablesorganic extractables in Non-
Targeted Analysisnon-targeted analysis (NTA)

DTS stage
Warning for WDs and CDs
This document is not an ISO International Standard. It is distributed for review and comment. It is subject to
change without notice and may not be referred to as an International Standard.
Recipients of this draft are invited to submit, with their comments, notification of any relevant patent rights of
which they are aware and to provide supporting documentation.

© ISO #### – All rights reserved

ISO #####-#:####(X)
2 © ISO #### – All rights reserved

All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication
may be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying,
or posting on the internet or an intranet, without prior written permission. Permission can be requested from either ISO
at the address below or ISO’s member body in the country of the requester.
ISO copyright office
CP 401 • Ch. de Blandonnet 8
CH-1214 Vernier, Geneva
Phone: + 41 22 749 01 11
EmailE-mail: [e-mail]
Website: www.iso.orgwww.iso.org
Published in Switzerland
iii
Contents
Foreword . v
Introduction . vi
1 Scope . 1
2 Normative references . 1
3 Terms and definitions . 1
4 General principles . 1
4.1 Identification in NTA. 1
4.2 Identification process . 2
4.3 Identification categories . 2
4.4 Process recommendations . 3
Annex A (informative) Discussion and description of the identification levels . 5
Annex B (informative) Identification process, suspect screening . 7
Annex C (informative) Identification process, non-targeted analysis, GC-MS (electron impact
ionization) . 10
Annex D (informative) Identification process, non-targeted analysis, LC-MS . 18
Annex E (informative) Criteria for an acceptable spectral match in GC-MS and LC-MS . 28
Annex F (informative) Good practices for structure elucidation . 31
Annex G (informative) Corroborating information for identifications . 39
Annex H (informative) Reporting the identity and identification status of a compound . 44
Bibliography . 46

iv
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards
bodies (ISO member bodies). The work of preparing International Standards is normally carried out through
ISO technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
The procedures used to develop this document and those intended for its further maintenance are described
in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the different types of
ISO documents should be noted. This document was drafted in accordance with the editorial rules of the
ISO/IEC Directives, Part 2 (see www.iso.org/directives).
ISO draws attention to the possibility that the implementation of this document may involve the use of (a)
patent(s). ISO takes no position concerning the evidence, validity or applicability of any claimed patent rights
in respect thereof. As of the date of publication of this document, ISO had not received notice of (a) patent(s)
which may be required to implement this document. However, implementers are cautioned that this may not
represent the latest information, which may be obtained from the patent database available at
www.iso.org/patents. ISO shall not be held responsible for identifying any or all such patent rights.
Any trade name used in this document is information given for the convenience of users and does not
constitute an endorsement.
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and expressions
related to conformity assessment, as well as information about ISO's adherence to the World Trade
Organization (WTO) principles in the Technical Barriers to Trade (TBT), see www.iso.org/iso/foreword.html.
This document was prepared by Technical Committee ISO/TC 194, Biological and clinical evaluation of medical
devices.
A list of all parts in the ISO 25364 series can be found on the ISO website.
Any feedback or questions on this document should be directed to the user’s national standards body. A
complete listing of these bodies can be found at www.iso.org/members.html.
v
Introduction
During their clinical use, medical devices either directly or indirectly contact tissue. As a result of this contact,
substances present in or on the medical device can be leached from the device during its clinical use. The
biological harms from these substances are evaluated using (quantitative) toxicological risk assessment,
which is driven by information about those substances. Chemical characterization per ISO 10993-18 provides
that information by identifying and quantifying a patient’s exposure to leached substances. As such
characterization involve extraction of medical devices, extracted substances (extractables) are taken as
potential leachables.
ISO 10993-18 specifies a framework for the discovery, identification, and quantification of extractable
constituents of a medical device, allowing the identification of biological hazards and the estimation and
control of biological risks from material constituents according to ISO 10993-17, ISO 10993-1, and ISO 14971.
vi
Chemical Characterizationcharacterization of Medical Devices:
medical devices —
Part 1:
Identification of Organic Extractablesorganic extractables in Non-
Targeted Analysisnon-targeted analysis (NTA)
1 Scope
This document describes and recommends best practice for the identification of organic extractables or
leachables discovered during non-targeted analysis (NTA) of medical device extracts or leachates.
NOTE A closely aligned topic, suspect screening analysis (SSA), is also considered in this document, see
Annex BAnnex B.
This document is intended to be used in conjunction with ISO 10993-18.
2 Normative references
There are no normative references in this document.
3 Terms and definitions
For the purposes of this document, the following terms and definitions apply.
ISO and IEC maintain terminology databases for use in standardization at the following addresses:
— ISO Online browsing platform: available at https://www.iso.org/obp
— IEC Electropedia: available at https://www.electropedia.org/
3.1 3.1
identified compound
organic compound which has been assigned a molecular structure and chemical name ([and similar
identifying information such as Chemical Abstract Service (CAS) Registry Number) (CASRN)] based on the
interpretation of available analytical data
3.2 3.2
unidentified compound
organic compound which cannot be assigned a molecular structure and chemical name based on the
interpretation of available analytical data
4 General principles
4.1 Identification in NTA
The identity of an extractable or leachable is critical to toxicological risk assessment (TRA). The identity allows
the risk assessor to obtain information about the compound’s toxicological profile. Identification in a NTA can
be a challenging and subjective process.
The identification process in a NTA should:
— be standardized, reasonable, and science-based;
— ensure repeatability, reproducibility and consistency;
— be practical and feasible for all compounds above the Analytical Evaluation Thresholdanalytical evaluation
threshold (AET);
— employ quality standards;
— produce a set of easily understood results, (see Table 1Table 1); );
— include how scientific evidence supporting identification is reported.
4.2 Identification process
ISO 10993-18 describes identification as assigning a molecular structure and chemical name (and other
designations such as Chemical Abstracts Service (CAS) numberCASRN) to a compound.
An identified compound has been assigned a molecular structure and chemical name. An unidentified
compound has not or cannot be assigned a molecular structure and chemical name.
Identification is based mainly on the compound’s mass spectrum. Additional analytical data, including
retention time (or index), accurate mass (empirical formula), etc. also contribute to the identification process.
When the analytical data for the compound does not inform its identity, the compound is assigned to Level 0
(no identification possible).
When the analytical data for the compound partially informs its identity (e.g. general structural attributes are
indicative of a phthalate), the compound’s identity is partially known and is assigned to Level 1.
When there is a single piece of analytical data (e.g. the mass spectrum) which establishes the molecular
structure (e.g. by structure elucidation by an expert mass spectrometrist, or by spectral matching to a
reference database)), the compound’s identity is tentativetentatively known, and it is assigned to Level 2.
When there are two corroborating pieces of analytical data (see Annex GAnnex G)) which establish the
molecular structure, the compound’s identity is confidently known, and it is assigned to Level 3.
When a Level 2 or 3 compound is subsequently verified by matching its analytical data with similar data for a
reference standard its identity is confirmed and it is assigned to Level 4 (the highest identification level).
A higher identification Level indicates a more definitive (i.e. information-rich) proposed identity. A higher
identification Level indicates the certainty of its true identity.
4.3 Identification categories
Table 1 — Identification classes and levels for NTA
Descriptive identification classes
Numerical classification
levels for reporting
General classes Detailed classes
Unknowna Level 0
Unidentified [3.2]
Partially known Level 1
b
Tentative identity Level 2
Identified [3.1]
a
Confident identity Level 3
a
Confirmed identity Level 4
[8] [9]
a Consistent with class descriptions provided in USP <1663> Reference and Reference <1664> [8] [9].
[8] [9]
b “tentative “Tentative identity” as defined in USP <1663> Reference and <1664> Reference is inconsistent with “partially
known.” USP terminology should not be used.
These classification levels are discussed in more detail in Annex AAnnex A.
The identification of organic extractables involves a systematic process of review and analysis of available
data. Such processes are described in Annex BAnnex B, Annex C, Annex C and Annex DAnnex D.
4.4 Process recommendations
4.4.1 Generating analytical data for identification
Analytical methods should be optimized to produce interpretable and information-rich mass spectra.
Deconvolution processes should be used on experimental chromatographic data (when necessary) to generate
the purest spectra achievable. For example, by resolving co-eluting compounds that cannot be
chromatographically resolved. (see Annexes CAnnexes C and DD).).
4.4.2 Classification of identities
The identity of a substance should be reported along with its classification. This communicates the confidence
in the identity (see Annex AAnnex A for additional discussion).
4.4.3 Securing identities
Compounds should be identified by either mass spectral matching or by mass spectral interpretation
(structure elucidation) or a combination of both.
— Mass spectral matching:
— used alone provides a Level 2 identity for a compound if the spectral match is judged to be
acceptable.
— can provide a Level 1 identity even with an unacceptable match if the available information
supports inferences concerning the compound’s structure or functionality.
— can be supported by expert review of available information, including the quality of the match.
— Quality is influenced by multiple, orthogonal and complementary factors (see Annex EAnnex E). ). For
[10]
example, considering matches secured from the NIST/Wiley database ,, the following are recommended
[ [11]
as objective criteria for an acceptable match [10] [11]::
— Matchmatch factor: >800 or equivalent
— Reversereverse match factor: >800 or equivalent
− Probability%:  >50%
— Inprobability percentage: >50 %
— in lib score: >0
Three of these four criteria should be met for a match to be judged as acceptable. Alternate
factors and acceptance criteria may be used to establish the quality of a match if justified.
— Candidate identities with matches of sufficient quality should be subjected to expert review to consider
discrepancies or inconsistencies between the analytical and library spectra.
— The expert review should produce one of two conclusions:
— any discrepancies or inconsistencies between the analytical and library spectra are
inconsequential and thus that the match is acceptable and the candidate identity is reportable.;
— any discrepancies or inconsistencies between the analytical and library spectra are sufficiently
consequential that the match is unacceptable and the candidate identity is rejected.
NOTE Mass spectral matching is inherently more effective (e.g.,. more likely to produce the correct identity)
when the experimental mass spectrum is matched against a database of mass spectral internally generated by
analysis of reference materials. Typically, this is a normal quality control procedure used by analytical laboratories.
— For structure elucidation:
— expert analysis of the existing MS data should always be included (see Annex F. (See Annex F).);
— a Level 2 identity is secured, regardless of the rigor and completeness of the elucidation, unless it
is corroborated with additional information or analysis.
4.4.4 Corroborating evidence
Evidence should be corroborated to the practical extent possible to produce the highest classification and
which is most likely to be the true identity. Corroborating evidence should be sought or generated to produce
the highest Level (See Annex GAnnex G).).
4.4.5 Reporting identities
When reporting an identity, the evidence supporting the identity should be provided so that the identification
Annex H).).
can be reviewed by potential users of this information (see Annex H
In certain cases, the identification process couldcan produce multiple Level 2 identities that are established
and justified to be acceptable. If expert review or corroborating evidence cannot establish a single identity as
the most likely candidate, then all the multiple identities should be reported.
The analytical chemist should consult with a toxicologist in the event of any doubt over an identity reported
for TRA.
Annex A
(informative)
Discussion and description of the identification levels
In the simplest sense, a compound can either be identified or remain unidentified. These two classes make up
the first tier of descriptive identification.
These two broad categories can be subdivided into sub-classes to provide greater granularity. Thus,
unidentified compounds can be categorized into two groups, a true unknown, meaning that no statement can
be made about its identity, or a partially known, meaning that some general statement can be made about the
compound’s structure. Similarly, identified compounds can be initially categorized into two groups, one group
where the identity is not corroborated with supporting data (tentative) and the other group where the identity
is supported with corroborating data.
Lastly, the sub-class “identity supported by corroborating data” can be further subdivided into two secondary
sub-classes depending on the type and quantity of corroborating data. If the corroborating data includes a
retention time (or index) and mass spectrum procured by analysis of an authentic reference standard, then
the identification class is “confirmed identity”. If the corroborating data isare not secured by analysis of a
reference standard, then the identification class is “confident identity”.
Identification classes can be numerically assigned to Levels to eliminate the subjective nature of the
descriptive terms:
a) Unidentified: theThe circumstance where a molecular structure and chemical name cannot be assigned
to a compound.
— Level 0,: Unknown: theThe circumstance where no structural information can be inferred from
the available information.
— Level 1,: Partially known: theThe circumstance where only certain structural attributes or the
elemental formula of the compound can be inferred, generally with some (limited) doubt as to the
correctness of the structural or elemental formula assignment.
The value of securing a partially known identification is as follows. The ability to assign a compound to a
structural class (e.g. phthalate, siloxane, aliphatic hydrocarbon) mightcan facilitate the decision as to
whether additional identification information is needed (e.g. an identified constituent). Furthermore, the
partially known information can be sufficient to support the conclusion that the compound of interest is
not of especially high toxicity or is not a “cohort of concern” compound.
b) Identified: the circumstance where a molecular structure and chemical name can be assigned to a
compound
— Level 2,: tentative identity: an identity that is obtained and supported by a single step analysis of
one piece of information (i.e. mass spectral matching or structure elucidation based on the
compound’s analytical mass spectrum) and which lacks corroborating information, where there
is some doubt as to whether the identity is correct. The criteria for obtaining acceptable matches
are discussed in Annex EAnnex E and the means of properly performing structure elucidation are
discussed in Annex FAnnex F.
— Level 3,: Confident identity: an identity that is supported by two or more pieces of corroborating
information, where there is little or no doubt that the proposed identity is correct. A discussion of
corroborating information is contained in Annex GAnnex G.
— Level 4,: Confirmed identity: an identity that has been verified via matching of analytical data (e.g.
the mass spectrum and retention time or index) from the compound of interest to an analysed
reference standard, where there is essentially no doubt that the identity is correct.
In certain circumstances, the corroborating data supporting a Level 3 identity is either so plentiful or so
compelling that the identity is, for all intents and purposes, confirmed. Nevertheless, a Level 4 identity is
reserved for that circumstance of authentic reference standard match.
Annex B
(informative)
Identification process, suspect screening
Suspect screening is a process that involves establishing whether pre-specified substances are present in an
analysed sample. As suspects are pre-specified using prior knowledge, their key analytical properties (such as
retention time and mass spectrum) are known. Suspect screening is accomplished by examining the
experimental data generated by analysing a sample to establish if that experimental data contains a response
that has the same known analytical properties exhibited by the suspect (e.g. retention time and mass
spectrum). If the analytical data contains a response that can be matched to the suspect, then it has been
verified that the suspect is present in the analysed sample. If the analytical data does not contain a response
that can be matched to the suspect, then the suspect is not present in the analysed sample (within the
sensitivity limitations of the test method). Thus, suspect screening often involves matching experimental data
to analytical data contained in a database.
Suspects are established by prior knowledge. For example, extractables data can be used to establish suspects
which are specifically considered in a leachables study and an item’s compositional information can be used
to establish suspect extractables.
The suspects that are addressed in suspect screening are often substances with Level 4 identities whose
retention time and mass spectrum has been properly matched to those properties of a reference standard.
However, suspects can also include compounds from all the other identification levels, so long as the
diagnostic data for the suspects have been collected. For example, a compound with a Level 1 (partially
known) can be a suspect because its retention time and mass spectrum are known.
The process of securing an identity via suspect screening for organic extractables/ and leachables is shown in
Figure B.1Figure B.1 for gas or liquid chromatography coupled to mass spectrometric detection, gas
chromatography - mass spectrometry (GC/-MS) or liquid chromatography - mass spectrometry (LC/-MS.). The
process starts with a suspect whose analytical properties (e.g. retention time and mass spectrum) are known.
The process proceeds by examining the experimental data generated by analysing a sample to establish if that
data contains a response that has the same known analytical properties exhibited by the suspect. If such a
response is found, then the presence of the suspect in the analysed sample is likely. The likely presence of the
suspect in the analysed sample is verified when the analytical match has been substantiated via expert review.
When a suspect has been confirmed to be present in a sample, the identification level of the suspect in the
analysed sample is the same identification level that the compound had as a suspect. For example, when an
extractable is used as a suspect leachable and is confirmed to be present in a leachate, the leachable will have
the same identification level as the extractable. Thus, matching a leachable’s peak in an analysed sample to a
suspect extractable that had a Level 1 identity produces a leachable that also has a Level 1 identity.
To ensure that a compound is correctly identified, the two primary sources of structure identities, the mass
spectrum and the retention time, should be highly reproducible. Since the acquisition of GC/-MS spectra is
highly standardized and is not significantly influenced by ionization competition effects (e.g. ion suppression),
these spectra are highly reproducible. Retention times, however, can shift for several reasons. Therefore,
control measures should be used to keep the retention times stable (e.g. retention time locking, use of relative
retention times) and the effectivity of these measures should be established by system suitability testing.
The outcome of a suspect screening exercise is generally definitive because it is based on corroborating data.
However, as is the case with any process that involves matching, the outcome of suspect screening should be
reviewed from two perspectives:
a) The supporting mass spectral data should be reviewed by an expert mass spectrometrist to verify the
match between the analytical and the suspects’ reference spectra, and
b) The presence of the suspect in the tested sample should be reviewed considering all the available prior
knowledge (e.g. it can be necessary to verify if the proposed identity of the suspect is consistent with the
known chemical composition of the test article).

a
As necessary and appropriate to resolve co-eluting peaks.
Figure B.1 — Process flow diagram, identification process for suspect screening
by GC/-MS and LC/-MS
Annex C
(informative)
Identification process, non-targeted analysis, GC/-MS (electron impact
ionization)
C.1 C.1 General
As opposed to suspect screening and target analysis, where suspects or targets have been predefined via prior
knowledge, NTA can be broadly defined as determining the chemical composition of a given sample without a
pre-defined list of suspects or targets. Thus, all compounds revealed via NTA identification efforts are
unidentified (until their analytical data has been interpreted) and unexpected.
In general, the identification process begins with the analysis of a sample and the generation of a
chromatographic peak which has, ideally, an interpretable and information-rich mass spectrum (and retention
time) associated with it. This mass spectrum is the foundation of the identification effort and can lead to a
compound being identified via two approaches, mass spectral interpretation and mass spectral matching.
Securing a high-quality mass spectrum is facilitated when the mass spectrum is differentiated from, and
unobscured by, potential interferences. Although chromatographic selectivity is a primary means for
managing co-eluting substances, peak co-elution cannot generally be avoided and it can be challenging to
resolve peak responses sufficiently so that useful and uncompromised mass spectra can be obtained for the
coeluting analytes. When the analyte’s mass spectrum is compromised due to spectral contamination,
unreliable identifications can be produced. Typical sources of spectral contamination include ion signals from
coeluting compounds, column bleed, solvent tailing, and electronic noise.
As visual inspection of complex chromatograms can be an ineffective means of resolving chromatographic
peaks and their associated mass spectra, the application of data processing techniques are oftencan be
necessary. When determining the mass spectrum that will serve as the basis for identification either through
library matching or mass spectral interpretation, or both, background subtraction should be the absolute
minimum approach. To support a higher quality of identification, however, application of a scrutinous
deconvolution-based approach can be necessary, as deconvolution delivers better quality mass spectra and
thus reduces the risk of misidentification.
C.2 C.2 Deconvolution
Deconvolution is the process of computationally extracting analyte signals from a complex mass
chromatogram, resulting in the elimination of background noise and the spectral separation of co-eluting
compounds. An example is given in Figure C.1Figure C.1.

NOTE The peak of interest (left above) is composed of two closely eluting components (mass spectrum, left below),
whose measured mass spectrum is deconvoluted into two high-quality mass spectra (right, above and below), one for
each analyte. From reference with permission.
[12]
SOURCE Reference . Reproduced with permission from the authors.
Figure C.1 — Deconvolution exampleExample of deconvolution of a single signal peak detected in a
total ion current (TIC) chromatogram
The deconvolution process involves multiple steps such as noise analysis, peak shape analysis (ions belonging

to the same peak should have the same apex and peak shape) and the assembly of a deconvoluted spectrum
[12]
[12]. It should be noted that deconvolution algorithms and related parameters can differ among various
software platforms and therefore it is possible to generate slightly different deconvoluted spectra for the same
raw data. Additionally, one should be aware that deconvolution willdoes not provide clean mass spectra in all
cases, especially in complex chromatograms that consist of both small and large peaks or which contain
coeluting compounds that share the same m/z values. In such circumstances, the deconvoluted spectrum
mightcan either be incomplete (deconvolution is carried too far and critical features are removed from the
mass spectrum) or contaminated (deconvolution was not carried far enough and interfering features remain
in the deconvoluted mass spectrum). As is the case with any means used to processing data, poorly or
improperly performed deconvolution mightcan lead to misidentification or cause proposed identities to be
classified at a lower Level.
These circumstances reinforce the need for a mass spectrometry expert to review the data and in difficult
cases maycan require additional actions (such as the evaluation of other deconvolution settings, manual
deconvolution, or other appropriate actions).
C.3 C.3 Mass spectral matching
Considering the process of identification in NTA, Figure C.2Figure C.2 illustrates the process workflow for
identification of a compound detected by GC/-MS. Once the analytical data (mass spectrum) has been collected
and deconvoluted, there are two principle means of using the mass spectrum to propose a compound’s
identity, mass spectral matching and mass spectral interpretation. The concept of mass spectral matching was
discussed for suspect screening in Annex BAnnex B. Mass spectral matching is conceptually identical in
suspect screening and NTA; the analytical mass spectrum is compared to reference mass spectra collected in
a library with the intent of uncovering library spectra that are comparable to (a good match to) the analytical
spectrum. Library compounds whose mass spectra match the analytical mass spectrum are candidate
identities. Presumably, the better the spectral match, the more likely it is that the candidate identity is the true
identity.
However, mass spectral matching differs significantly in suspect screening versus NTA. In suspect screening,
the matching occurs between the analytical mass spectrum and the known mass spectrum of the suspect. The
matching process in suspect screening therefore produces a binary and definitive (yes or no) outcome. If there
is a match, then it is concluded that the suspect is present in the tested sample. If there is no match, then it
concluded that the substance is not the suspect and thus further identification activities maycan be warranted.
In NTA, the mass spectrum of the detected compound is matched with the mass spectra contained in an
external mass spectral library. Here theThe mass spectral matching likely produces an open-ended outcome.
The matching does not produce the one true identity; rather, it produces potential identities. Even if there is
only one good match, the circumstance of having a single match does not mean that the match is surely the
compound’s true identity. Furthermore, if there would be no match, it does not end the identification process,
but means that the identity need be secured by other means.
The library of mass spectra can be developed internally by analysing reference standards and collecting the
[12]
key analytical information (retention time, mass spectrum, response factor) [12]. ). Such a library is specific
in the sense that it contains information related only to extractables or leachables. Alternatively, a library of
mass spectra can be developed externally, typically by an independent organization that collects and collates
spectra from multiple sources which produce the spectra in varying ways. Such external libraries are generic
and general in the sense that they contain data for compounds which are not extractables or leachables. Well-
known and well-controlled large commercial reference libraries such as the NIST/EPA/NIH Mass Spectral
[13] [14] [15]
Library [13] [14] and the Wiley Registry of Mass Spectra [15] are widely used tools to facilitate
spectral matching of GC/-MS data.
C.4 C.4 Expert assignment of identification levels
Considering the work process for spectral matching more closely (the right-hand side of Figure C.1Figure
C.1),), the process begins with the comparison, generally machine-driven, of the analytical spectrum to the
library spectra, presumably producing a list of hits (compounds whose mass spectra are close matches to the
analytical spectrum). Should one (or more) of the hits meet the criteria for an acceptable match (discussed in
Annex EAnnex E),), the match is evaluated by an expert to confirm that the match is proper and to weed out
false identities (e.g. good mass spectral matches which cannot be the analyte of interest because of clear
scientific discrepancies between the library and analytical spectra). This expert review focuses on verifying
the data that supports the acceptable match but couldcan involve some basic mass spectral interpretation
(structure elucidation), looking for fundamental inconsistencies between the library spectrum and the
analytical spectrum (e.g. an ion is missing in the analytical spectrum, compared to the library spectrum,
indicating a functional group maycan be missing in the proposed identity). Ultimately, this expert evaluation
either rejects the match (i.e. expert review finds one or more critical inconsistencies between the analytical
and library spectra) or confirms the match (i.e. expert review finds no critical inconsistencies between the
analytical and library spectra).
In the circumstance that the expert rejects the match as being unacceptable, identification proceeds in one of
two ways. Firstly, if no other acceptable candidates were revealed by spectral matching, then further attempts
at identification can only proceed forward as an exercise of mass spectral interpretation. If other candidates
were revealed by spectral matching (but were not taken as the lead candidate), then these other candidates
are also reviewed by the expert. Again, if the expert rejects the match or matches as being a viable candidate,
the identification efforts can proceed forward only as an exercise of mass spectral interpretation. However,
any spectral match identification the expert finds acceptable becomes a Level 2 identity as it is only supported
by one piece of evidence, which is the spectral match. Additionally, a mass spectrometrist couldcan elevate the
level of identification from a Level 2 identity to a Level 3 identity by including additional structure supporting
evidence such as retention time (or index) match, a consideration of prior knowledge (for example, ise.g. the
identification consistent with the known composition of the test article), a provable relationship to a
compound with a confident or confirmed identity, detected in the same analytical method or by correlating
the compound to a compound with a confident or confirmed identity which was detected in an orthogonal
analytical technique. As the identity would then be based upon two or more pieces of independent
information, the identity of the compound is corroborated and becomes Level 3.
In the event that two types of information are in conflict, experimental evidence should supersede the
rationalization of an identification from material information.
C.5 C.5 Confirmation of identity (Level 4)
C.5.1 Candidates that move forward with Level 3 identities are queried as to whether a Level 4 identity is
required or desired. If a Level 4 identity is required, the ability to procure a reference standard is ascertained.
If a reference standard can be procured, then this is an opportunity to verify the identity. Verification of the
identity has at least two desirable outcomes:
a) It inspires confidence that the identity is correct and that all appropriate efforts have been used to secure
and justify the identity,
b) It promotes the identified analyte from NTA to suspect screening (because the compound is placed in the
organization’s spectral library once the reference standard is analysed).
However, there will beare cases where the value of verifying an identity is less than the value of the resources
required to perform the verification:
— — when the reference standard is not commercially available and generation of the reference standard
is difficult and costly;
— — when a Level 4 identity is not necessary.
If a reference standard can be procured, then the reference standard is analysed to produce a reference
spectrum and a reference retention time (or index). If there is an acceptable match between the analytical
spectrum and retention time (or index) and the standard’s spectrum and retention time, then the Level 3
identity has been verified or confirmed and the identity is elevated to Level 4.
C.5.2 Although one presumes that verification of an identity by analysis of a reference standard is
successful (that is the identity is verified), such a desirable outcome is not always achieved. Given that the
proposed identity at this point is corroborated by both an acceptable spectral match and by the expert review,
one expects such negative outcomes to be rare (but not impossible). If the identity is not confirmed by analysis
of the reference standard, then the identity is rejected as false. At this point, one reviews the available
analytical data and decides
a) (a) whether there is another candidate identity that can be taken through the expert review process or
b) (b) whether the identification process need be re-started via structure elucidation (as spectral matching
produced only a false result).
Considering the importance of retention data as corroborating information, this topic is addressed in greater
detail in Annex GAnnex G. . While retention times are specific to a particular chromatographic method,
retention indices are more universally applicable as they are, in essence, normalized relative retention times
using linear n-alkanes for normalization. Most of a method’s operating parameters do not have a large effect
on the retention index, except for the nature (polarity) of the stationary phase.
The NIST spectral libraries augment the spectral data of reference compounds with different types of retention
indices. Experimental retention indices are curated from experimental data and are indicated per type of
stationary phase (semi-standard non-polar, standard non-polar and polar). In addition, computed retention
indices can be available.
C.6 C.6 Structure elucidation
Considering mass spectral interpretation (the left-hand side of Figure C.2Figure C.2),), good practices for
performing structure elucidation are provided in Annex FAnnex F. . The structure elucidation process can have
two outcomes: if the structure elucidation fails to reveal a structure due to lack of interpretable information,
the compound of interest is classified as an unknown. If the available information can be interpreted, the
outcome of the interpretation is considered. Two outcomes are possible:
a) (1) a structure can be proposed for the compound of interest, or
b) (2) only the compound’s general structural features can be ascertained (e.g. the compound is established
to likely be “a phthalate”). In the latter case, the compound is classified as Level 1 while in the former case,
the compound is classified as having a Level 2 identity.
As it is generally desirable to either substantiate or verify a Level 2 identity, additional evidences are collected
and reviewed. If there are no additional evidences, then the identity remains unconfirmed at Level 2. If one or
more evidences can be collected, the identity is “elevated” to Level 3. Level 3 identities are further reviewed
to determine if they can be verified and elevated to Level 4 status via the process discussed previously.
It should be noted that the above guidance on interpreting GC/-MS spectra is not all-encompassing as there
can be additional tools available for GC/-MS mass spectral interpretation which were not addressed in this
annex.
a
As necessary and appropriate to resolve co-eluting peaks.
Figure C.2 — Identification process workflow, NTA by GC/-MS
Annex D
(informative)
Identification process, non-targeted analysis, LC/-MS
Essentially, NTA analysis in LC/-MS produces grouped mass to charge (m/z) values at different retention
times. In the absence of prior knowledge concerning a compound of interest, expert interpretation of this mass
spectral information will be the sole basis for proposing an identity for that compound. It is the task of a mass
spectrometry expert to review this information, further group and correlate the associated m/z values to mass
fragments, and then to interpret this information, leading to a proposed identity for the detected compound.
The first step in the interpretation of LC/-MS data isare to group all m/z values per retention time,
...