The determination of

non-intentionally added substances (NIAS)

as a part of ‘safety-by-design’

Dr. Malcolm DriffieldFera Science Ltd.

malcolm.driffield@fera.co.uk

Date: July 2016, Version 1

Fera Science Ltd

Our Vision

• To be the global provider of leading edge scientific solutions,

evidence and advice across the agri-food chain.

Sustainable productivity & environmental

protection

Sustainable productivity & environmental

protection

Chemical safety &

stewardship

Chemical safety &

stewardship

Food safety, quality &

authenticity

Food safety, quality &

authenticity

Proficiency testing

Proficiency testing

Crop health protection

Crop health protection

Targeted knowledge for

informed decision making

Targeted knowledge for

informed decision making

Food packaging &

migration

Food packaging &

migration

Novel foodstuffs

Novel foodstuffs

From Farm to ForkFrom Farm to Fork

What can migrate from FCMs?

• Known ingredients such as monomers, catalysts, solvents,

suspension media, additives etc.

• Known or unknown isomers, oligomers (arguable if NIAS or

not), impurities, reaction products and breakdown

products of these ingredients

• Possible contaminants from the manufacturing process

such as recycled materials, irradiated products etc

• Contamination from indirect sources such as printing inks,

external coatings, adhesives, secondary packaging

IAS

NIAS

EU legislation - NIAS

• Framework Regulation EC 1935/2004 Article 3 and Plastics

Regulation EU 10/2011 Article 19

• FCM manufacturers are obliged to ensure the safety of their

products

• Safety of NIAS must be assessed

• Presently, there are no levels of migration or exposure set

for which compliance with this requirement can be

demonstrated

• EU 10/2011 states unauthorised substances may be used in

FCM plastics, provided they do not migrate at levels above

10 µg/kg food, and this is often used for NIAS

Protocols to follow……

• Presently no legally prescriptive guidelines on how to

assess the safety of NIAS

• ILSI Monograph - Guidance on best practices on the risk assessment

of NIAS in Food Contact Materials and Articles (July 2015)

• What approach should be taken?

Reporting

Data interpretation

Analysis

Extraction or migration exposure

Sample preparation

Initial discussions

Targeted vs. Non-targeted analysis

Specific extractionClean-up

Specific analysis

Generic extraction

Minimal clean-upGeneric analysis

IAS - Know exactly what to measure

PredictableNIAS – oligomers,

impurities, literature references

NIAS -Complete unknowns

Targeted Non-targeted

Analysis and Data interpretation

Reporting

Data interpretation

Analysis

Extraction or migration exposure

Sample preparation

Initial discussions

Identification

Quantification

0 50 100

NM

R in

teg

ral v

alu

e

Concentration

TIC vs. EIC?

vs.

Detection

Detection of NIAS

• Is chromatography required?

• Ambient mass spectrometry - rapid diagnostic information

PVC gasket

Atmospheric pressure Solids Analysis Probe (ASAP)

Driffield et al. (2015) Use of ASAP-TOF-MS to screen for plasticisers in gaskets used in contact with foods. Rapid Communications in Mass Spectrometry, 29,

1603-1610.

Direct Analysis in Real Time (DART)

HR-MS

Real time chemical imaging of the surface

FCM

Identification of NIAS – GC-MS

• Volatile and semi-volatile NIAS - GC-MS

• Identification can be carried out by comparison to

electron impact (EI) MS libraries

• Libraries contain many 10,000’s of substances but often still

unknown NIAS with no library match

(replib) Decane

30 60 90 120 150 180 210 240 270 300 330 360 390 4200

50

100

30

43

57

71

85

99113 142

n-decane

NO LIBRARY MATCH

GC-MS (EI) TIC of FCM solvent extract

GC-MS (chemical ionisation)

• Electron impact – high energy, leads to fragmentation

• Chemical ionisation – softer ionisation which can lead to

quasi-molecular species (or higher mass fragments)

• Different mechanisms possible, e.g. proton transfer

e-. + M → M+. + 2e-→ F1+. + F2 →→

e-. + CH4 → CH4+. + 2e-

CH4+. + CH4 → CH5

+ + CH3.

CH5+ + M → CH4 + [M+H]+

GC-MS EI vs. NCI for Triacetin

(mainlib) Triacetin

10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 2300

50

100

1527 32

43

61 73 86

103

116

145

158

O O

O

O

O

O EI-MS(library)

NCI-MS

OO

O

O

CH3

CH3

O

CH3

O TriacetinMW 218

[M-H]-

EI = Electron impact ionisationNCI = Negative chemical ionisation

GC-TOF-MS

• High resolution – higher selectivity

• Accurate mass – potential chemical formulae for

fragments (molecular adduct if using chemical ionisation)

• Some accurate mass libraries available but not as

comprehensive as EI libraries

OH+

O

O

= C8H5O3+ = 149.0250

GC-TOF-MS – FCM extract

OH+

O

O

= C8H5O3+ = 149.0250

EICs at m/z 149.0250 with different mass windows

TIC

m/z 148.5250-149.5250+/- 3300 ppm

m/z 148.9877-149.0623+/- 250 ppm

m/z 149.0213-149.0287+/- 25 ppm

Identification of NIAS – LC-MS

• Non-volatile and polar NIAS detected by LC-MS

• Liquid chromatography

• Reversed phase – typically used for LC-MS

• HILIC – smaller, very polar and/or basic analytes

• HR-MS

• Accurate mass, isotope information (spacing and intensity),

fragmentation (in-source or MS/MS)

• Positive and negative ionisation modes possible

Database searching

• Comparison of list of accurate masses to theoretical

database

• Can include retention time, structures, MS/MS fragment data

• Theoretical oligomers and reaction products associated with

starting materials (and impurities if known) below 1000 Da

• Should consider simulant-oligomer interactions

• Can contain tens of thousands of compounds….

• Not confirmed unless compared to authentic standard

AA TMA PA CHDM BD EG DEG PG HD HMP TMP NPG H2O MWPA+EG linear 1 1 1 210.0528EG+PA+EG linear 1 2 2 254.0790PA+EG+PA+EG linear 2 2 3 402.0951PA+EG+PA+EG cyclic 2 2 4 384.0845PA+EG+PA+EG+PA linear 3 2 4 550.1111PA+EG+PA+EG+PA+NPG linear 3 2 1 5 636.1843PA+EG+PA+NPG+PA+EG linear 3 2 1 5 636.1843PA+PG+PA+PG+PA+PG linear 3 3 5 636.1843PA+PG+PA+PG+PA+PG cyclic 3 3 6 618.1737

Halogenated-oligomers

• Identified by isotope pattern

• M:M+2 3:1 intensity = Cl

• M:M+2 1:1 intensity = Br

• Predicted formulae can indicate presence of multi-halogenated oligomers

• F is monoisotopic but presence of F atoms indicated by massdeficient masses

• If monomer contains 5xF atoms then presence of substances inGC-MS or LC-MS analysis containing 5xF, 10xF, 15xF, 20xF etcmight indicate oligomer

• Other multiples of F indicate reaction and/or breakdownproducts

• How do we differentiate between linear and cyclic

oligomers?

• Authentic standards available?

• Chromatography

• Retention time differences?

• Structural isomers?

• Functional group specific chemistry

Linear or cyclic oligomers?

BADGE.BPA = C36H40O6Cyclo-di-BADGE = C36H40O6

R

O CH3 O CH3

O O

+ no reaction

R

O F3C O CF3

O O

+

R

O

O

F3C

O

CF3

O

Acylation reactions

Driffield et al. Agilent Application Note: LC-TOF-MS as a tool to support can coating/food interaction studies

LC-QTOF-MS investigation

• Can we use MS data to differentiate between linear and

cyclic oligomers?

• Accurate mass information will be the same

• BADGE.BPA = C36H40O6, Cyclo-di-BADGE = C36H40O6

• Fragment ions characteristic of the presence of an epoxy

group?

C+

O

O

CH3

CH3

C12H15O2+

m/z 191.1067C+

OH

CH3

CH3

C9H11O+

m/z 135.0804 C+

O

CH3

CH3

OH

OH

C12H17O3+

m/z 209.1172

FCM with oligomers – cyclic or linear?

LC-QTOF-MS EIC at m/z 404.2430Database hit to potential oligomer

C+

OH

CH3

CH3

C+

OH

CH3

CH3

C+

OH

CH3

CH3

C+

O

O

CH3

CH3

C+

O

O

CH3

CH3

135.0801

Quantification

• How can we quantify if we don’t know what we are looking

for?

• Internal standards

• 2H, 13C analogues – not naturally occurring

• Range of compounds to cover the mass range of interest?

• IS that responds in positive and negative ionisation (for LC-MS)

• External standards

• Chemically similar to substances of interest

e.g. BADGE or BADGE hydrolysis products for epoxy-related

e.g. Polyester diol urethane substance for polyester-related

• Synthesis of authentic standards to confirm identity

• Retention time, accurate mass, fragmentation comparison

• Expensive and time consuming

Quantification by LC-TOF-MS and NMR

Extraction of a FCM with solvent

AA TMA PA CHDM BD EG DEG PG HD HMP TMP NPG H2O MWPA+EG linear 1 1 1 210.0528EG+PA+EG linear 1 2 2 254.0790PA+EG+PA+EG linear 2 2 3 402.0951PA+EG+PA+EG cyclic 2 2 4 384.0845PA+EG+PA+EG+PA linear 3 2 4 550.1111PA+EG+PA+EG+PA+NPG linear 3 2 1 5 636.1843PA+EG+PA+NPG+PA+EG linear 3 2 1 5 636.1843PA+PG+PA+PG+PA+PG linear 3 3 5 636.1843PA+PG+PA+PG+PA+PG cyclic 3 3 6 618.1737

0 50 100

NM

R in

teg

ral v

alu

e

Concentration

0.04 mg/6 dm2

0.03 mg/6 dm2Monomer functional group quantification

Oligomer determination

0

20

40

60

80

100

120

0 50 100 150

Are

a

Concentration

Quantification by LC-TOF-MS and NMR

• Taking into account the uncertainties associated with

each of the measurement techniques, these results are in

fact very similar

• NMR measures actual epoxy functionality, whereas the

epoxy structures by LC-TOF-MS are assumed and not

confirmed by comparison to authentic standards

• Concluded that there is 0.03 mg/6 dm2 of residual epoxy

functionality in the acetonitrile extract of the FCM

Universal detectors

• Chromatography detector where the response is independent of chemical structure

• For example GC-FID, LC-ELSD, LC-CAD

• Rely on the separation capability of chromatography as not as selective as MS detectors

• Must have good separation of NIAS in chromatogram – not usually the case

• Testing of real food samples difficult as components from food may interfere

• Still need a decision in consultation with customer on which standard(s) will be used for quantification

NIAS and ‘safety-by-design’

• Close relationship between Manufacturer of FCM and

Analytical Laboratory allows an iterative process during

new product development

• Better understanding of components of ingredients should

lead to more ‘predictable’ NIAS

• What is the effect of changing an ingredient or processing

temperature to the profile of the NIAS?

Summary

• Ever-advancing analytical instrumentation provides very

powerful tools for determination of IAS and NIAS

• As instruments become more readily available and

sensitivity increases there is the potential for more and

more NIAS to be detected

• Identification and quantification are still providing

challenges

• Evolution of current technologies to be used in new areas or

to provide different information, but also evolution of new

technologies to be used in NIAS determination

Acknowledgements

• Julie Christy

• Antony Lloyd

• Mónica García-López

• Jonathan Tarbin

• James Donarski

• Emma Bradley

• Laurence Castle