See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/325649673

Identification of non volatile migrant compounds and NIAS in polypropylene

films used as food packaging characterized by UPLC-MS/QTOF

Article  in  Talanta · June 2018

DOI: 10.1016/j.talanta.2018.06.022

CITATIONS READS

21 1,801

3 authors:

Paula Vera Elena Canellas

University of Zaragoza University of Zaragoza
35 PUBLICATIONS   499 CITATIONS    38 PUBLICATIONS   647 CITATIONS   

SEE PROFILE SEE PROFILE

Cristina Nerín
University of Zaragoza
384 PUBLICATIONS   9,296 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

PET2007‐09‐C05‐04 INIA‐Plan Especial Teruel, 2007-2010 del Ministerio de Educación y Ciencias Español y del Fondo Europeo de Desarrollo Regional. View project

valorization of opuntia ficus indica View project

All content following this page was uploaded by Elena Canellas on 04 February 2019.

The user has requested enhancement of the downloaded file.

 Talanta 188 (2018) 750–762

Contents lists available at ScienceDirect

Talanta
journal homepage: www.elsevier.com/locate/talanta

Identification of non volatile migrant compounds and NIAS in T

polypropylene films used as food packaging characterized by UPLC-MS/
QTOF
⁎
Paula Vera, Elena Canellas, Cristina Nerín
Analytical Chemistry Department, GUIA Group, I3A, EINA, University of Zaragoza, María de Luna 3, 50018 Zaragoza, Spain

A R T I C LE I N FO A B S T R A C T

Keywords: Migration of non volatile compounds from twenty six PP films used as food contact materials has been studied in
NIAS four simulants (ethanol 95% and 10%, acetic acid 3% and Tenax ®) and analyzed by UPLC-MS/QTOF. Seventy
PP six compounds have been identified, where 76% of them were non-intentionally added substances (NIAS)
Food packaging coming from degradation of additives used, such as methyl or ethyl or hexyl-3-(3,5-di-tert-butyl-4-hydro-
Migration
xyphenyl) from irganox 1076 and irganox 1010 degradation; or impurities such as N,N-bis(2-hydroxyethyl)
UPLC-MS/QTOF
amines, or compounds of unknown origin, like hydro-ceramides. The most common compounds found were
glyceryl monostearate or monopalmitate, erucamide, irganox 1010, irgafos 168, irgafos 168 OXO, N,N-bis(2-
hydroxyethyl) tridecylamine and N,N-bis(2-hydroxyethyl) pentadecylamine.
Six films didn’t comply with the European Regulation Nº 10/2011/EU, where irganox 1010 and the group of
N,N-bis(2-hydroxyethyl) amines exceeded their SMLs. Other films surpassed the maximum concentration re-
commended by Cramer for the compounds of class II (degradation products) or III (amide compounds) when
ethanol 95% was used as simulant.

1. Introduction European Regulation No 10/2011/EU [13] where the positive list of

Polypropylene is a widely used material to manufacture films and
containers for food packaging. Among its main characteristics the fol-
lowing ones can be highlighted: its optimum cost/benefit ratio, versa-
tility in its uses, good processability, lightness, moisture barrier, orga-
noleptic and chemical properties, thermal and rupture resistance and
transparency [1].
There is a lot of bibliography about the use of PP to package different
types of food like biscuits, bread, candy, snackfoods and dried foods
mainly [2,3]. More over, specific applications such as quince jelly, cheese
[3], pasta [4], screw cap for fruit juices [5] or soda, meat or fish can cover
[6], or also for utensils used for food [7]. Recently, PP has been also used
as alternative to polycarbonate for baby bottles [8–11] due to the pro-
blems with the migration of bisphenol A from this material.
Nowadays, there is a great concern and compromise from companies
and industry manufactured of these plastics, about the study of potential
migrants. Within these components susceptible to migrate; monomers,
additives, catalyst and degradation products could be found. To guarantee
their safety, the materials must fulfill the European Regulation Nº 1935/
2004 [12] about materials and articles into contact with food and also
monomer, starting substances and additives authorized and their specific
migration limits (SML), as well as, the conditions (time, temperature and
simulants) for migration tests are established.
Two different types of methodologies (target or non-target analysis)
have been used for the study of these potential migrants in PP. Target
analysis is usually applied to look for known migrated substance,
usually to those intentionally added substances (IAS) or to some well
known non-intentionally added substances (NIAS). For example, mi-
gration of antioxidants commonly used such as BHT [14], irganox 1010
[15] and irgafos 168 [16,17] in different conditions, including the
microwave heating or high-pressure processing [18]. Or also, light
stabilizers like chimassorb 944, 81 and different types of tinuvin
[19–21]. These analyses were carried out by GC-MS and HPLC-MS for
volatile and non volatile compounds, respectively. Other important
migrants determined were mineral oil saturated hydrocarbons (MOSH),
and polyolefin oligomeric saturated hydrocarbons (POSH) by a com-
prehensive two-dimensional gas chromatography (GCxGC) [4,22,23].
Non- target analysis is a tedious and difficult task, where the first
aim is basically to identify any chemical compound present in the mi-
grant sample by a screening assay.

⁎
Corresponding author.
E-mail addresses: pvera@unizar.es (P. Vera), elenac@unizar.es (E. Canellas), cnerin@unizar.es (C. Nerín).

https://doi.org/10.1016/j.talanta.2018.06.022
Received 16 January 2018; Received in revised form 6 June 2018; Accepted 7 June 2018
Available online 08 June 2018
0039-9140/ © 2018 Elsevier B.V. All rights reserved.
P. Vera et al. Talanta 188 (2018) 750–762

Table 1
Characteristics of PP films studied such as code, food simulants tested and foodstuff intended to be in contact.
Code Food simulants Foodstuff contact

ETOH95% Tenax ETOH10% AC3%

PP1 X X X Cheese, salmon and tuna, cream and butter

PP2 X X X Fresh fish and meat and pizza
PP3 X X X Pizza, Fresh vegetables, pasta, mayonnaise and mustard sauce and butter
PP4 X X X Sugar, mayonnaise and mustard sauces
PP5 X X X Poultry meat, cheese, pasta and processed fruit
PP6 X X X Fresh fish, cream and fried potatoes
PP7 X X X Cheese, cream and butter
PP8 X X X Fresh fish and meat and different sauces
PP9 X X X Poultry meat, pizza and fried potatoes
PP10 X X X Fresh fish, mayonnaise and mustard sauce and butter
PP11 X X X X Cacao, coffee, chocolate, cheese, biscuits, pastry, bread and cakes
PP12 X X X Fresh fish and meat and fried potatoes
PP13 X X X Cheese, different sauces and butter
PP14 X X X Fresh fish and meat and fried potatoes
PP15 X X X X Pasta, fresh fish, mayonnaise and mustard sauce
PP16 X X X X Biscuits, cacao, coffee, meat, vinegar, chocolate, pastry and bread
PP17 X X X X Pasta, Cheese, aromatic herbs such as mint, tea, chamomile and processed fruit
PP18 X X X X Peanuts, Almonds, hazelnuts… Sugar, biscuits, pastry and bread
PP19 X X X X Walnuts, pine kernels, preparations for soups, sauces or broths and infusions
PP20 X X X X Different types of nuts, biscuits and preparations for soups, sauces or broths
PP21 X X X X Pasta, cacao, fresh vegetables, preparations for soups, chocolate, pastry and bread
PP22 X X X X Chestnuts, almonds, biscuits, cacao and cakes
PP23 X X X Sugar, mayonnaise and mustard sauce, and chocolate
PP24 X X X Cheese, different type of fish, fresh vegetables, processed fruit and preparations for soups and sauces
PP25 X X X Pizza, poultry meat, fried potatoes, mayonnaise and mustard sauce and vinager
PP26 X X X Pizza, cheese, cream

There are several studies about non target analysis from PP, especially
for volatile compounds. Several NIAS were detected in baby bottles
[10,11,24], in films, laminate or granulate PP [5,6,25–28], like different
types of phthalates, alkanes, degradation products such as 7,9-di-tert-
butyl-1-oxaspiro(4,5)deca-6,9-diene-2,8-dione, methyl 3-(3,5-di-tert-butyl-
4-hydroxyphenyl) propionate, 3,5-di-tert-butyl-4-hydroxybenzaldehyde or
3,5-di-tert-butyl-4-hydroxyacetophenone, among others, were found. All
these compounds were analyzed and identified by GC-MS assisted by NIST
or WYLEY library. However, few works based on the combination of GC
coupled to time-of flight mass analyzer (MS/QTOF) for PP samples were
found [29–31]. This mass analyzer is considered the best choice for non
target analysis, due to its high capacity of performing accurate mass
measurements and good sensitivity in full scan, increasing the identifica-
tion efficiency. The high resolution MS provide the accurate MS, which is
the essential tool to identify non volatile compounds, as a library of MS
spectra is not available for these compounds, although NIST Peptide MS
Library, EPA Tandem Mass Spectrometry and NIST 17 Tandem MS Library
have been developed [32]
Much less bibliography was found about non target analysis for non
volatile compounds from PP samples [24,30,31,33]. Only few com-
pounds were identified, for example N,N-bis(2-hydroxyethyl)tridecy-
lamine as a NIAS, or an optical brightening agent, 2,5-bis(5′-tert-butyl-
2-benzoxaolyl) thiophene. This work was focused on non-target study
of migration of non-volatile compounds from a wide number of PP
samples from different manufacturing companies. The main purpose
was to increase the knowledge of possible NIAS that could appear, and
so, to give a complete overview of all compounds that could come from
this type of plastic. This study was carried out by UPLC-MS/QTOF as a
powerful tool for this type of analysis [34–37].
2. Materials and methods
2.1. Reagents
The standards triallyl pentaerythritol (1471-17-6), NX 8000
(882073-43-0), trihexyl trimellitate (1528-49-0), tinosorb OMC (5466-
77-3), oleamide (301-02-0), chimassorb 81 (1843-05-6), glyceryl
monostearate (31566-31-1), dinonyl phthalate (84-76-4), irganox 3114
(27676-62-6), diisodecyl phthalate (26761-40-0), irganox 1010 (6683-
19-8), erucamide (112-84-5), irganox 1076 (2082-79-3), irgafos 168
(31570-04-4), palmitic acid (57-10-3), stearic acid (57-11-4), 3,5-di-
tert-butyl-4-hydroxybenzaldehyde (1620-98-0), methyl 3-(3,5-di-tert-
butyl-4-hydroxyphenyl) propanoate (6386-38-5), benzenepropanoic
acid, 3,5-bis(1,1-dimethylethyl)-4-hydroxy- (20170-32-5), glyceryl di-
hexadecanoate (26657-95-4), N,N-bis(2-hydroxyethyl) dodecylamine
(1541-67-9), octadecanamide, N,N′-1,2-ethanediylbis- (110-30-5), oc-
tadecanamide (124-26-5), 13-docosenamide,N-hydroxy-, (13Z)-
(119042-56-7), 11-eicosenamide, (11Z)- (10436-08-5) tetradecanamide
(638-58-4), hexadecanamide (629-54-9), 9-octadecenamide, (9Z) (301-
02-0), 3,5-di-tert-butyl-4-hydroxybenzaldehyde (1620–98-0), hexamide
(628-02-4), 2,6-di-tert-butyl-4-(1-methylpropyl)-phenol (17540-75-9),
formic acid, acetic acid and tetrahydrofuran were purchased from
Sigma-Aldrich Química S.A (Madrid, Spain). All were of analytical
quality. Ethanol, water and methanol of HPLC grade were supplied by
Scharlau Chemie S.A (Sentmenat, Spain). Tenax TA 80/100 mesh was
supplied by Supelco (Bellefonde, USA).
2.2. Market samples
Migration from twenty six different types of PP intended for food
contact materials was studied. All materials were supplied by different
European companies in order to have a representative sampling of PP
used in the EU market to package different types the food.
None information about their formulations was supplied due to con-
fidentiality reasons. Only, the final intended of these materials was known,
what means, the type of food intended to be in contact with them.
The materials under study are shown in Table 1.
2.3. Migration assays
The migration assays were carried out with four different simulants
according to Commission Regulation 10/2011/EU [13]. For each ma-
terial, the simulants tested were chosen depending on their final uses
according to the intended food

751
P. Vera et al.
Ethanol 10% (simulant A) and acetic acid 3% (simulant B) were
selected when the plastics were intended for hydrophilic food char-
acter. Ethanol 95% (simulant D) and Tenax ® (simulant E) were selected
for fat and dry foods respectively [13].
The simulants chosen for each sample and the final use of these
materials in the market are shown in the Table 1.
The migration tests were performed by immersion of the plastics in
the liquid simulants (A, B or/and E), following the proportions estab-
lished in the 10/2011/EU [13] legislation (6 dm2/1 kg simulant). And
then, the assay was kept in an oven at 60 ºC for 10 days. All samples
were monomaterial and monolayer PP.
For migration experiments with Tenax ® as food simulant, 2 × 4 cm
cut-outs of PP under study were full covered with 0.32 g of Tenax®
forming a uniform layer (4 gTenax/dm2) according to UNE-EN-14338
[38]. Then, they were placed over an aluminum foil and inside a Petri
dish and kept in the oven at 60 ºC during 10 days. After this period,
Tenax® was extracted two consecutive times with 3 mL of ethanol
[35,36]. Previously, Tenax was purified by Soxhlet extraction with
acetone and ethanol during 6 h and further dried in the oven at 100 ºC
After that, A and B simulants were directly analyzed by UPLC-MS-
QTOF. However, D and E simulants were previously concentrated x6
under a gentle stream of N2 and analyzed by the same methodology
subsequently described.
2.4. UPLC- MS/QTOF analysis
Chromatography was carried out in an Acquity™ system couple to
Xevo G2 QTOF as detector, both supplied by Waters (Milford, MA,
USA). The detector consisted of an API source (atmospheric pressure
ionization) with an electrospray interface (ESI) coupled to a Xevo G2
mass spectrometer with a hexapole, a quadrupole, a collision cell and a
time of flight as analyzer (QTOF).
For chromatographic separation an Acquity UPLC BEH C18 column
of 17 µm particle size (2.1 mm x 100 mm) also from Waters (Milford,
MA, USA) was used. The solvents used as mobile phase were water
(phase A) and methanol (phase B), with or without 0.1% formic acid for
positive and negative ionization, respectively. The column flow was
0.3 mL/min and the column temperature was 35 ºC. The gradient used
was 98–2% phase A - phase B finishing at 100% phase B after of 15 min.
The volume of injected sample was 5 µL.
Instrumental parameters were as follows: the electrospray probe
was used in positive (ESI+) and negative (ESI-) modes and in sensi-
tivity mode. The mass range considered was from 10 to 1000 Da. The
corona voltage was 2.5 kV for (ESI+) and 0.5 kV for (ESI-). Two cone
voltages were used for each injection: 30 and 75V, in order to obtain a
greater number of compounds detected. The source temperature was
150 ºC, the desolvation gas temperature 450 ºC and the desolvation gas
flow 650 Lh−1.
MSE mode was selected for the acquisition, which alternates be-
tween two functions: function 1 acquiring at low-energy to obtain the
exact mass precursor ion spectra and function 2 acquiring at elevated-
energy exact mass fragment ions. The collision ramp energy was from
15 to 40 V.
MassLynx v.4.1 software (Waters, Milford MA, USA) was used to
analyze the samples and CromaLynx (Waters, Milford MA, USA) was
used to deconvolude the spectra.
2.5. Identification of migrant compounds from plastics
The first step for the identification was to determine which chro-
matographic peaks came from the sample. For this purpose, a previous
visual comparison between the migration chromatograms of each
sample with their respective blanks of simulants was carried out. The
different peaks found were called “markers”. In addition, to complete
the number of “markers” for each sample, a Chromalynx XS ® software
tool was used. This is capable to find different peaks which appear in
752
Talanta 188 (2018) 750–762
the sample and not in the blank, even though they are not visually
detected. The parameters selected for Chromalynx XS software tool
from Waters were: non targeted analysis, mass tolerance ± 20 mDa,
retention time ± 0.2 min and ions abundance above 1000.
After this step, a list the “markers” was obtained for all plastics
under study. Each marker was defined by its retention time and its
precursor ion with its exact mass.
Subsequently, the identification of the “markers” was carried out.
Then, the elemental composition was determined, assisted by the
Elemental Composition tool within MassLynx 4.1. This software pro-
vides the most probable formula for each precursor ion, considering
that the molecules are formed with the most common elements (C, H,
O, N, Cl, S, P and Na as adduct) selected by the analyst. Once the dif-
ferent options of the elemental composition were known, the molecular
structures were searched in chemical databases such as Chemspider
[www.chemspider.com] or Scifinder [scifinder.cas.org] obtaining a
number of candidates. The selection of these candidates was made ac-
cording to chemical criteria based on knowledge of the polymers and its
possible additives as well as on the experience
Finally, to confirm the “marker” compounds, Mass Fragment ®
software was utilized. This tool compare the spectra of our candidates
above selected, with the unknown compounds, both fragmented by
function 2 (high energy) of MSE. If their fragments coincided, the
identification was given as “very probable”
When the pure standards were commercially available, this identi-
fication was confirmed, comparing the retention time and mass spectra
to the “marker”.
2.6. Quantification of migrant compounds from PP films
For the quantification, calibration curves were prepared using the
pure standard solutions in ethanol. If the standard was not available,
the quantification was carried out with a standard of similar molecular
structure. These were analyzed by UPLC-MS/QTOF with the metho-
dology above described. Three replicates of each concentration level
were analyzed to determine the reproducibility.
For A, B and D2 simulants, the concentration of migrant compounds
was expressed as miligrams of compounds per Kilogram simulant.
To calculate the migration values for Tenax as solid simulant in mg/
Kg, the absolute mass migrating in mg was calculated. This value was
divided by the 0.08 dm2 used in the migration test and the ratio 6 dm2/
1 kg of simulant (regulation 10/2011/EU) [13].
2.7. Risk assessment
Risk assessment of the identified migrant compounds was applied.
First, the migrants were searched in the positive list of the legislation
10/2011/EU [13], and the values found were compared their specific
migration limits (SML). For those non listed compounds, the TTC
(Cramer's list) (software Toxtree®) was applied in order to classify its
toxicity. Cramer classifies the compounds according to their chemical
structures into three categories (class I, class II and class III that cor-
respond to low, medium and high toxicity, respectively)
Besides, Cramer recommends maximum values of human exposure
for each toxicity class [39]. The values of estimated daily intake (EDI)
for class I, II and III are 1.8, 0.54 and 0.09 mg per person per day,
respectively. According to the European rules [13,40], the worst pos-
sible scenario of intake is that, an adult consumes every days 1 kg food
packaged in 1 dm3. Therefore, the maximum migration recommended
for each class is I, II and III, 1.8; 0.54 and 0.09 mg/Kg respectively.
3. Results and discussion
Migration from twenty six different films based on PP was evaluated
in this work. These samples were used to package different types of food
like fresh meat, fish, pizza, cheese, nuts, biscuits, fried potatoes,
P. Vera et al.
chocolate, coffee, different types of sauces…They were supplied by
different European plastic companies in order to obtain a representative
sampling of these materials that can be present in the market for food
package contact.
Firstly, the migration of these materials was carried out in different
simulants depending on their intended use. After that, these were
analyzed by UPLC-MS/QTOF with their corresponding control analysis.
Then, untarget analysis was carried out to detect and identify the mi-
grants from these plastics. This complex and tedious work was neces-
sary due to the lack of knowledge of their compositions and also, the
possibility of finding new compounds from degradation, reaction by
products or impurities (NIAS).
After the identification, the migrant compounds were quantified
and then, the risk assessment was applied following the above men-
tioned procedure.
3.1. Identification of migrants from PP films
The simulants, after migration test, were analyzed by UPLC-MS/
QTOF in order to identify non volatile migrant compounds from these
PP films studied
Two migration chromatograms from PP18 sample to ethanol 95%
and its respective blank are shown in the Fig. 1. Fifteen migrated
compounds were detected which don’t appear in the blank. They were
highlighted with numbers and detailed in Table 2.
All migrated compounds detected in all samples and simulants
studied are organized by their retention times in Table 2. Their mole-
cular masses, MS fragments and their molecular formula are also
shown.
A total of seventy four compounds were found including known
additives and also NIAS. Some of these NIAS seemed to have their
origin in these additives due to their similar structures.
3.1.1. Migrant additives in PP samples
The first part of Table 2 (compounds 1–16) lists the compounds
commercially used for the manufacture of polymers (which had brand
names), whose functions are as antioxidants, such as irganox 1010 and
1076, irgafos 168, chimassorb 81 and irganox 3114 also used as
Fig. 1. Migration chromatogram of sample PP18 to ethanol 95% as simulant analyzed by UPLC-MS/QTOF to 30V, ESI+.
753
Talanta 188 (2018) 750–762
polymer stabilizer. Besides, several plasticizers were found like trihexyl
trimellitate, dioctyl and diisodecyl phthalate. Some slip agents as eru-
camide or oleamide and clarified agent like NX 8000. Also a surfactant,
emulsifier and UV absorber like glyceryl monostearate and tinosorb
OMC respectively. And also, palmitic and stearic acid as lubricants
[6,10,11,28,41,42].
3.1.2. Migrant unknown compounds in PP samples
The second part of this table (compound numbers 17–74), was
completed with the unknown compounds (NIAS), which have been
organized and grouped by different families as follows:
3.1.2.1. Family 1: Family formed with reference structure. Family 1
(Table 2, compounds from 17 to 26) was formed by the compounds
that had a similar structure constituted by a group 3,5-di-tert-butyl-4-
hydroxyphenyl (C14H22O), which was assigned the name of “reference
structure” and it is shown in Table 3.
Inside this group, the compounds 17 and 18 (Table 2) had the re-
ference structure bonded to an aldehyde group. The mass difference
between both compounds was 14.0159 m/z that corresponded to the
gain of CH2 (both compounds shown in Table 3). Their identifications
were corroborated by the availability of the 3,5-di-tert-butyl-4-hydro-
xybenzaldehyde (compound 17) as standard. Their main mass frag-
ments in both compounds were [MH+ ]- 56.0624 m/z by the lost of
C4H8 and the mass 57.0704 m/z. Both compounds were found in bib-
liography as degradation products derived from phenolic antioxidants
[5,10,43,44].
The rest of molecules of this family 1 (19−26) formed Na adducts
and all of them had the “reference structure” above named bonded with
a propanoate group. This new system called “structure of repetition” is
also shown in Table 3. It must be emphasized that this structure was
also common in the antioxidants irganox 1076 and irganox 1010
(Table 3) above described.
In some of these new compounds, this “structure of repetition” was
bonded for example to a methyl, ethyl (gain of CH2) or hexyl (gain of
C5H10) (Tables 2 and 3; compounds number 19, 20 and 21 respec-
tively). Only the compound number 19 (methyl 3-(3,5-di-tert-butyl-4-
hydroxyphenyl)) was corroborated by its retention time and its mass
 Table 2
Migrants found in migration solutions from PP samples, time retention time (RT), mass detected (m/z), molecular formula, MS fragments obtained, candidate name with its CAS and remarks like its use (if it is known) and
in whose samples were found these compounds.
P. Vera et al.

N RT and m/z Ion Molecular formula MS fragments Candidate name Remarks Samples found

+
1 8.84_279.1573 [MNa ] C14H24O4 167.0388 and 151.0768 Triallyl pentaerythritol, CAS: 1471-17-6 Crosslinking agent PP26
2 9.22_507.2730 [MNa+] C29H40O6 398.4269 and 133.1017 NX 8000, CAS: 882073-43-0 Clarified agent PP3, PP4, PP6, PP7and PP9
3 9.23_485.2879 [MNa+] C27H42O6 207.0293 and 73.0653 Trihexyl trimellitate, CAS: 1528-49-0 Plasticizer PP6
4 9.31_313.1786 [MNa+] C18H26O3 122.0771 Tinosorb OMC, CAS: 5466-77-3 UV absorbers PP6
5 9.92_304.2617 [MNa+] C18H35NO 265.2531, 247.2426 and Oleamide, CAS:301-02-0 Non-ionic surfactant PP1, PP2, PP15 and PP26
97.0891
6 10.08_327.1960 [MH+] C21H28O3 215.0708 and 137.0239 Chimassorb 81, CAS:1843-05-6 Antioxidant PP1 and PP2
7 10.17_381.2985 [MNa+] C21H42O4 378.4257, 341.3060 and Glyceryl monostearate, CAS: 31566-31-1 Emulsifier PP4, PP10, PP11, PP16, PP18, PP19,
323.2269 Without of glyceryl group PP20 and PP21
8 10.21_441.2992 [MNa+] C26H42O4 315.1554 and 149.0239 Dinonyl phthalate, CAS: 84-76-4 Plasticizer PP8 and PP9
9 10.25_806.4918 [MNa+] C48H69N3O6 765.5073 and 289.1552 Irganox 3114, CAS: 27676-62-6 Antioxidant for processing and PP19 and PP22
long-term thermal stabilization
10 10.40_469.3274 [MNa+] C28H46O4 329.1735 and 149.0239 Diisodecyl phthalate, CAS: 26761-40-0 Plasticizer PP8 and PP9
11 10.47_1199.7827 [MNa+] C73H108O12 541.1458 and 281.0536 Irganox 1010, CAS:6683-19-8 Antioxidant All samples except PP25
12 10.52_360.3250 [MNa+] C22H43NO 321.3032 and 303.2926 Erucamide, CAS: 112-84-5 Slip agent PP5, PP15, PP19, PP21, PP22, PP23
and PP24
13 11.83_553.4596 [MNa+] C35H62O3 475.4151 and 419.3525 Irganox 1076, CAS: 2082-79-3 Antioxidant PP25
14 13.71_647.4586 [MH+] C42H63O3P 591.3967 and 441.2922 Irgafos 168, CAS: 31570-04-4 Antioxidant All samples except PP26
15 9.95_255.2336 [MH-] C16H32O2 59.0133 Palmitic acid, CAS: 57-10-3 Lubricant PP16 and PP21
16 10.24_283.2637 [MH-] C18H36O2 59.0133 Stearic acid, CAS: 57-11-4 Lubricant PP16 and PP21

Family 1: Group with reference structure

17 8.78_ 235.1702 [MH+] C15H22O2 179.1072 and 57.0704 3,5-Di-tert-butyl− 4-hydroxybenzaldehyde Confirmed with standard PP17, PP21 and PP22
CAS: 1620–98–0
18 8.65_249.1861 [MH+] C16H24O2 193.1229 and 57.0704 Ethanone, 1-[3, 5-bis(1,1-dimethylethyl)-4-hydroxyphenyl], CAS: PP21

754
14035–33–7
19 9.27_315.1936 [MNa+] C18H28O3 219.1749 and 259.1410 Methyl 3-(3,5-di-tert-butyl-4-hydroxyphenyl) propanoate CAS: 6386–38–5 Confirmed with standard PP10, PP11, PP12, PP13, PP14,
and 107.0487 PP15, PP16, PP17, PP18, PP21,
PP23 and PP24, PP25
20 9.43_329.2103 [MNa+] C19H30O3 219.1749 and 259.1410 Ethyl 3-(3,5-di-tert-butyl-4-hydroxyphenyl) propanoate, CAS: 36294-24-3 PP10, PP11, PP12, PP13, PP14,
and 107.0487 PP15, PP16, PP17, PP18, PP23,
PP24 and PP25
21 9.91_385.2709 [MNa+] C23H38O3 219.1749 and 259.1410 Hexyl 3-(3,5-di-tert-butyl-4-hydroxyphenyl) propanoate, CAS: 34938-67-5 PP25
and 107.0487
22 8.90_345.2044 [MNa+] C19H30O4 305.2119 and 263.1920 Benzenepropanoic acid, 3,5-bis(1,1-dimethylethyl)− 4-hydroxy-, 2- PP17, PP23, PP24 and PP26
hydroxyethyl ester
CAS: 40388–51–0
23 8.40_277.1805 [MNa+] C17H26O3 235.1728 and 161.0966 Benzenepropanoic acid, 3,5-bis(1,1-dimethylethyl)− 4-hydroxy- Confirmed with standard PP23, PP24 and PP26
[MH-] CAS: 20170–32–5
24 7.99_419.2404 [MNa+] C22H36O6 379.2479 and 317.1429 Benzenepropanoic acid, 3,5-bis(1,1-dimethylethyl)-4-hydroxy-, 3-hydroxy- PP23, PP24 and PP26
[MH-] 2,2-bis(hydroxymethyl)propyl ester CAS: 26347–98–8
25 9.43_679.4181 [MNa+] C39H60O8 639.4261 and 317.1429 Benzenepropanoic acid, 3,5-bis(1,1-dimethylethyl)− 4-hydroxy-, 1,1′-[2,2- PP23, PP24 and PP26
bis(hydroxymethyl)− 1,3-propanediyl] ester
CAS:36913–60–7
26 10.04_939.5989 [MNa+] C56H84O10 899.6064 and 317.1429 Benzenepropanoic acid, 3,5-bis(1,1-dimethylethyl)− 4-hydroxy-, 1,1′-[2-[[3- PP23, PP24 and PP26
[3,5-bis(1,1-dimethylethyl)− 4-hydroxyphenyl]− 1-oxopropoxy]
methyl]− 2-(hydroxymethyl)− 1,3-propanediyl] ester
CAS:84633–54–5

Family 2: With glycerol molecule

27 10.31-36_297.2051 [MNa+] C15H30O4 294.2104, 257.2126 and Glyceryl monolaurate CAS:27215–38–9 PP16
239.1329
28 10.50_325.2360 [MNa+] C17H34O4 322.3589, 285.2435and Glyceryl myristate CAS: 27214–38-6 PP11 and PP16
267.1632
(continued on next page)
Talanta 188 (2018) 750–762
 Table 2 (continued)

N RT and m/z Ion Molecular formula MS fragments Candidate name Remarks Samples found
P. Vera et al.

+
29 9.96_353.2665 [MNa ] C19H38O4 350.3794, 313.2823 and Glyceryl monopalmitate CAS: 26657–96-5 PP1, PP2, PP3, PP4, PP10, PP11,
295.1963 PP16, PP18 and PP21
+
30 10.51_409.3283 [MNa ] C23H46O4 406.4379, 369.3379 and Glyceryl eicosanoate CAS:30208-87-8 PP18
351.2397
31 10.81_437.3588 [MNa+] C25H50O4 434.4495, 397.3663 and Glyceryl docosanoate CAS:30208-87-8 PP18
379.2886
32 11.18_465.3932 [MNa+] C27H54O4 462.4839, 425.4007 and Glyceryl tetracosanoate CAS:100830-53-3 PP18
407.3230
+
33 11.21-26_479.3722 [MNa ] C27H52O5 439.3756 and 239.1455 Glyceryl didodecanoate CAS: 27638-00-2 PP11 and PP13
34 11.60_507.4036 [MNa+] C29H56O5 467.4111and 239.1455 Glyceryl ditridecanoate CAS: 121957-71-9 PP11
35 12.12-19_535.4353 [MNa+] C31H60O5 495.4307 and 239.1455 Glycerol dimyristate CAS:53563-63-6 PP11
36 12.70-79_563.4670 [MNa+] C33H64O5 523.4745 and 239.1455 Glyceryl dipentadecanoate CAS:121957-69-5 PP11 and PP13
37 13.44-58_591.4956 [MNa+] C35H68O5 551.5031 and 239.1455 Glyceryl dihexadecanoate CAS:26657-95-4 Confirmed with standard PP11, PP13 and PP16
38 14.19_619.5288 [MNa+ ] C37H72O5 579.5363 and 239.1455 Glyceryl diheptadecanoate CAS:26657-95-4 PP16 and PP18
39 10.57_325.2351 [MH+] C19H32O4 322.2451, and 267.1632 Glyceryl bonded unsaturated hydrocarbon chain (C16) PP13 and PP14
40 10.76_353.2670 [MH+] C21H36O4 350.4022 and 295.1952 9,12,15-Octadecatrienoic acid, monoester with glycerol, Glyceryl PP13 and PP14
monolinolenate
CAS: 26545-75-5
41 10.95_381.3038 [MH+] C23H40O4 378.3138 and 323.2336 Glyceryl bonded unsaturated hydrocarbon chain (C20) PP5
Family 3: Dihydroxy alquilamines
42 5.44_218.2137 [MH+] C12H27NO2 200.2032, 102.0926 and N,N-Bis(2-hydroxyethyl) octylamine CAS: 15520-05-5 PP5
88.0772
+
43 6.59_246.2480 [MH ] C14H31NO2 228.2375, 102.0926 and N,N-Bis(2-hydroxyethyl) decylamine CAS: 18924-65-7 PP5
88.0772
44 7.54_274.2828 [MH+] C16H35NO2 256.2704, 102.0926 and N,N-Bis(2-hydroxyethyl) dodecylamine CAS 1541-67-9 Confirmed with standard PP5, PP20 and PP22
88.0772

755
45 7.9-8.2_288.2910 [MH+] C17H37NO2 270.2805, 102.0926 and N,N-Bis(2-hydroxyethyl) tridecylamine CAS:18312-57-7 PP3, PP4, PP5, PP6, PP7, PP8, PP9,
88.0772 PP10, PP11, PP12, PP13, PP14,
PP15, PP16 and PP18
46 8.41_302.3140 [MH+] C18H39NO2 284.3035, 102.0926 and N,N-Bis(2-hydroxyethyl) tetradecylamine CAS: 18924-66-8 PP5
88.0772
47 8.54-8.64_316.3240 [MH+] C19H41NO2 298.3135, 102.0926 and N,N-Bis(2-hydroxyethyl) pentadecylamine CAS: 24910-32-5 PP3, PP4, PP5, PP6, PP7, PP8, PP9,
88.0772 PP10, PP11, PP12, PP13, PP14,
PP15, PP16 and PP18
48 8.71_330.3433 [MH+] C20H43NO2 312.3328, 102.0926 and N,N-Bis(2-hydroxyethyl) hexadecylamine CAS: 18924-67-9 PP5
88.0772
+
49 8.84_358.3715 [MH ] C22H47NO2 340.3610, 102.0926 and N,N-Bis(2-hydroxyethyl) octadecylamine CAS: 10213-78-2 PP5
88.0772
50 8.96_398.4376 [MH+] C26H55NO2 380.4271, 102.0926 and N,N-Bis(2-hydroxyethyl) didodecylamine CAS: 15214-78-5 PP5
88.0772

Family 4: Ceramide and dihydroceramide

51 9.93-10.06_470.4566 [MH+] C29H59NO3 256.2640 and 227.2006 C11-Dihydroceramide PP11 and PP13
52 10.19_498.4888 [MH+] C31H63NO3 256.2646and 227.1908 C13-Dihydroceramide PP11 and PP13
+
53 10.30_526.5199 [MH ] C33H67NO3 256.2644and 227.1919 C15-Dihydroceramide PP11 and PP13
+
54 9.44_540.5365 [MH ] C34H69NO3 256.2646 and 227.1923 C16-Dihydroceramide CAS:5966-29-0 PP5
55 10.39_554.5531 [MH+] C35H71NO3 256.2646 and 227.1932 C17-Dihydroceramide PP11 and PP13
56 10.48_582.5813 [MH+] C37H75NO3 256.2646 and 227.1920 C19-Dihydroceramide PP11
57 9.54_594.5852 [MH+] C38H75NO3 256.2639 and 227.1919 C20-ceramide, CAS: 7344-02-7 PP5

Family 5: amides bonded by ethylene

58 10.58_503.4586 [MNa+] C30H60N2O2 519.4535 and 983.9274 Tetradecanamide,N,N′-1,2-ethanediylbis-, CAS: 5136-46-9 PP10
59 10.91_531.4916 [MNa+] C32H64N2O2 547.4865 and Pentadecanamide,N,N′-1,2-ethanediylbis- CAS:175031-37-5 PP10
1039.9934
60 11.30_559.5181 [MNa+] C34H68N2O2 575.5130 and Hexadecanamide, N,N′-1,2-ethanediylbis- CAS: 5518-18-3 PP10 and PP18
1096.0464
(continued on next page)
Talanta 188 (2018) 750–762
P. Vera et al. Talanta 188 (2018) 750–762

spectrum with its standard commercially available. The rest of mole-

cules had the same fragmentation pattern with the accurate common
masses like 259.1410, 219.1752 and 107.0487, m/z.
Fig. 2 shows the spectrum of fragmentation of methyl 3-(3,5-di-tert-

All samples except PP26

PP10, PP18 and PP19

PP10, PP18 and PP19

butyl-4-hydroxyphenyl), with their fragmented structures according to
Mass Fragment tools.
Samples found

These NIAS compounds detected may come from degradation of

irganox 1076 or irganox 1010 fragmenting their ester chains by dif-
PP19 ferent parts [10,25,26,31,41].
PP19
PP15

PP15

PP15

PP15

PP15

PP15
PP7
PP7

PP5
Among the compounds 23, 24, 25 and 26 shown in Table 3 only the
compound number 23 could be corroborated with its standard. The rest
of these compounds were identified by Mass fragment tools obtaining
Confirmed with standard. It

the same fragmentation pattern with the common mass 317.1429 m/z
Confirmed with standard
Confirmed with standard

Confirmed with standard

and the mass [MNa+]-39.9925 m/z (by the lost of Na adduct and OH)
These last compounds seemed to be degradation products from ir-
used as lubricant

ganox 1010 fragmented by different parts. Their structures showed a

clear relationship, where the oxygen of ester was bonded to hydrogen
Remarks

forming an O-H bond and producing the rupture of the molecule ir-
ganox 1010 in four, three, two or one parts, which correspond to mo-
lecules numbers 23, 24, 25 and 26 respectively.

3.1.2.2. Family 2: With glycerol molecule. The next group comprised the
compounds with structures with glycerol. This group was divided into
9,12-Octadecadienamide,N,N-dimethyl, (9Z,12Z)- CAS: 2501-33-9

the compounds whose molecular formula were CxH2xO4 (Table 2 from

9,12-Octadecadienamide,N,N-diethyl, (9Z,12Z)- CAS: 3140-46-3
Heptadecanamide, N,N′-1,2-ethanediylbis- CAS: 173862-96-9

number 27–32) corresponding to an ester of an acid chain bonded to a

glycerol molecule. And those compounds (Table 2 from number 33–38)
Octadecanamide, N,N′-1,2-ethanediylbis- CAS: 110-30-5

13-Docosenamide,N-hydroxy-, (13Z)- CAS: 119042-56-7

Ethanol, 2-(tetradecylmethyl amino) CAS: 56975-12-3

whose molecular formula was CxH2x-2O5, where glycerol was bonded to

Ethanol, 2-(dodecylmethyl amino) CAS: 35841-91-9

two same ester chains. The difference of mass between the consecutive
molecules was m/z 28.0298 that corresponds to the gain of C2H4.
Tetracanamide,N,N-diethyl- CAS: 57303-20-5
Dodecanamide,N,N-diethyl- CAS: 3352-87-2

Compounds from 39 to 41 did not form Na adducts. Their molecular

11-Eicosenamide, (11Z)-CAS: 10436-08-5

structures responded to the formula, CxH2x-6O4. They may be an ester of

Tetracosanamide CAS: 1611477-48-5

an acid but with three unsaturated hydrocarbon chains bonded to the

Irgafos 168 OXO CAS: 95906-11-9
Eicosanamide CAS: 51360-63-5

glycerol molecule.
Docosanamide CAS: 3061-75-4

Only the compounds glyceryl monostearate, glyceryl palmitate and

N,N-Dimethyllinoleamide

glyceryl dihexadecanoate were found as standards and they were cor-

N,N-Diethyllinoleamide

roborated by checking their retention times and their mass spectra. The
mass fragments of the rest of compounds were similar to those obtained
Candidate name

for the compounds 27–32 (Table 2) the common abundant fragments

[MNa+]-2.9980 m/z, [MNa+]-39.9925 m/z (lost of Na adduct and OH)
and [MNa+]-58.0702. And for the compounds 33–38 (Table 2)
[MNa+]-39.9925 (lost of Na adduct and OH) and the mass fragment
239.1455.
293.2844, 292.3004 and

293.2844, 292.3004 and

293.2844, 292.3004 and

293.2844, 292.3004 and

226.2483 and 214.2411
254.2820 and 242.2828

336.3266 and 332.3317

607.3916 and 551.3287

It is known that the glyceryl monostearate (compound 8) is widely

used as additive of anti-fog and anti-static additive in plastics [28,42].
However, application of the rest of glycerol molecules is not known.
603.5451 and

631.5772 and
MS fragments

1152.1106

1208.1748

Only the glyceryl monolaurate as antistatic plasticizer is known for

116.1100
116.1100
291.2562

319.2998

275.2739

275.2739

275.2739

275.2739

PVC, PS and cellulosic plastics (that it was not the case under study)
and glyceryl monopalmitate was found in bibliography as anti-foggy
[6]. Therefore, the detection of these compounds in some samples can
Molecular formula

be related to the impurity of glyceryl monostearate as additive.

C36H72N2O2

C38H76N2O2

C22H43NO2

C42H63O4P
C15H33NO
C17H37NO
C16H33NO
C18H37NO
C20H37NO

C22H41NO

C20H39NO

C20H41NO

C22H45NO

C24H47NO

3.1.2.3. Family 3: Dihydroxy alquilamines. Family 3 consisted of

dihydroxy alquilamines. This group was formed by the compounds
from number 42–50 of Table 2. Their molecular formula responded to
CxH2x+3NO2 and its structure consisted of an amine bonded to two
[MNa+]
[MNa ]

[MNa ]

[MNa+]

[MNa+]

[MNa+]
[MNa+]

[MNa+]

[MNa+]

[MNa+]
+

[MH+]
[MH+]
[MH+]
[MH+]

ethanol molecules and also an alkyl hydrocarbon chain. The mass

difference between the consecutive compounds was m/z 14.023, which
Ion

corresponded to the gain of CH2 in the alkyl hydrocarbon chain.

9.70 y 9.86_376.3191

The identification of compound number 44 was corroborated by the

63 7.63–7.76_244.2641
64 8.12–8.23_272.2948

standard N,N-bis(2-hydroxyethyl) dodecylamine which was commer-

11.82_587.5502

12.46_615.5823

67 10.07_330.2772

10.33_358.3089

10.25_332.2935

10.47_334.3086

10.78_362.3396

10.82_388.3554

11.31_685.4346
Others compounds

65 8.23_256.2635
66 8.68_284.2939
Table 2 (continued)

cially available, checking its retention time and its fragmentation pro-
RT and m/z

file.
Within this family, the smaller compound had a chain of eight
carbons N,N-bis(2-hydroxyethyl) octylamine and the bigger compound
with twenty two carbons was N,N-bis(2-hydroxyethyl) didodecylamine.
61

62

68

69
70

71

72

73

74
N

Most of these compounds were formed with an even number of carbons

756
P. Vera et al. Talanta 188 (2018) 750–762

Table 3
Structure of NIAS for the Family 1.
Family 1: Reference
Compound number 17 Compound number 18
structure

Family 1: Structure of
Irganox 1076 Irganox 1010
repetition

Compound number 19 Compound number 20 Compound number 21 Compound number 22

Compound number 23 Compound number 24

Compound number 25 Compound number 26

in their alkyl hydrocarbon chains, except the compounds 45 and 47 that
had thirteen and fifteen carbons respectively. Furthermore, it should be
highlighted that these last two compounds were found in 15 samples of
PP.
All these had the same fragmentation pattern, with common accu-
rate masses such as 70.0646, 88.0768, 102.0903 and 106.0876 m/z.
Besides of the most abundant fragment whose mass corresponded to
[MH+] - 18.0088 (m/z) ([MH+]-OH).
In bibliography the presence of these migrants is attributed to a
possible impurity/reaction product/breakdown product of the additives
used. This consideration was obtained from FDA [33] or also found in
bibliography like antistatic additive [28]
3.1.2.4. Family 4: ceramide and dihydroceramide. The family four was
constituted by the compounds from 51 to 57 of Table 2. Ceramides are a
family of waxy lipid molecules which are composed of sphingosine (an
18 carbon amino alcohol with an unsaturated hydrocarbon chain) and a
fatty acid. They respond to the molecular formula CxHx-1NO3. The
difference between the dehydroceramide and ceramide molecules is
that unsaturated hydrocarbon chain is substituted by saturated chain.
Therefore, the hydroceramide molecules have two hydrogen atoms
more in their formula, responding to CxHx+1NO3 formula.
In the samples studied, only C20-ceramide (Table 2, compound 57)
was detected. However six hydroceramides were found. Most of them
were formed with odd number of carbons in their fatty acid chains

757
P. Vera et al.
Fig. 2. Spectrum of methyl 3-(3,5-di-tert-butyl-4-hydroxyphenyl for high energy collision and their fragmented structures according to Mass Fragment tolls.
(gaining a mass of m/z 28.0312 corresponding to C2H4) and only one
with even carbon number (C16-dydroceramide, Table 2, compound
54). Due to the absence of any of these standards, they were checked by
MassFragment ® analysis and the mass spectrum obtained in the QTOF
at high collision energy, where the fragmentation of these markers was
similar to these candidates.
It is known, that some hydro-ceramides are widely used as a func-
tional ingredient for cosmetics [45,46], but nothing has been found
about their presence in plastics. For this reason, they can be considered
NIAS where its origin is unknown.
3.1.2.5. Family 5: amides bonded by ethylene. Other group consisted of
the compounds formed by two amides of the same acid bonded by
ethylene (compounds from 58 to 62). These compounds formed sodium
adducts [MNa+] and their molecular formula responded to CxH2xN2O2.
The mass difference between the consecutive molecules was 28.0312
produced by the gain of C2H4, It meant the presence of a methyl in each
chain of amide.
Octadecanamide, N,N′-1,2-ethanediylbis- was confirmed with its
standard by retention time and its spectrum. The most abundant masses
found in their spectrum corresponded to the formation of Na-dimer and
the common fragment [M]+ 15.9949 m/z (their oxidations)
Within this group, the biggest compound was octadecanamide,
N,N′-1,2-ethanediylbis-, which is a synthetic wax used as a dispersing
agent or internal/external lubricant [47]. The application of the rest of
compounds of this series was unknown, so it is not unreasonable to
think that they may be degradation products from this lubricant losing
C2H4.
3.1.2.6. Other compounds. The last part of Table 2 contains other
compounds. The most probable molecules for 63 and 64 (Table 2),
corroborated by Mass Fragment® analysis, were formed by an amine
bonded to an ethanol, a methyl and also an alkyl hydrocarbon chains.
Checking the results above described, it was found a great similarity
between these ones and the family number 3, where an ethanol was
substituted by a methyl.
758
Talanta 188 (2018) 750–762
The compounds 65 and 66 responded to [MH+] ionization and both
compounds gave a mass difference of 28.0304 (gain C2H4), as before. Its
molecular formula was CxH2x+1NO that would correspond to the amide
family. Due to the absence of pure standards it was not possible to
accurately identify both compounds but the most probable migrants
according to MassFragment ® are shown in Table 2. The compounds 67
and 68, with formula CxHx-3NO and mass difference between both also
of 28.0317 m/z (gain of C2H4), were dimethyl and diethyl of linolea-
mide respectively as the most probable molecules.
The compounds 69 and 70 were confirmed with their standards.
Other three compounds (71, 72 and 73) with CxHx+1NO as mole-
cular formula were detected whose mass difference was also
28.0312 m/z (gain of C2H4). In this case, they would be primary amides
corroborated by their fragmentations with the standard octadecana-
mide obtaining the same fragmentation pattern with the common
fragment masses 293.2844, 292.3001 and 275.2739 m/z.
All these amides above described could come from the impurities or
degradation products from erucamide and oleamide widely used as slip
agents [42].
Finally, the compound 74, that appeared in almost of samples, and
whose molecular formula was C42H63O4P, corresponded to the gain of
an oxygen regarding irgafos 168 (C42H63O3P). To corroborate its
identification, 10 ppm of irgafos 168 standard was oxidized with tet-
rahydrofuran during 1 day at 40 ºC in an oven [48] and analyzed by
UPLC-MS/QTOF. The chromatographic peak of irgafos 168 disappeared
and a new peak appeared at the same time and mass fragmentation as
compound 74, confirming that it is the oxo-derivative from irgafos 168.
3.2. Quantification of migrants
Once all migrant compounds were identified, the migration con-
centrations were quantified in each simulant. Table 4 shows the com-
mercial standards used, their limits of detection and quantification and
its reproducibility. It is also detailed what standard was used to quan-
tify the compounds identified whose compound was not commercial
available.
P. Vera et al. Talanta 188 (2018) 750–762

Table 4
Commercial standards used with their limits of detection (LOD) and quantification (LOQ), lineal range and reproducibility obtained. The other identified compounds
quantified with these standards.
Standards LOD µg/kg LOQ µg/kg Lineal range µg/ Calibration curve Reproducibility Quantification of other
ethanol ethanol kg ethanol identified compounds

Triallyl pentaerythritol 16 53 53–517 y = 16.5x− 52.1 0.9996

NX 8000 11 38 38–245 y = 2.87x+ 76.1 0.9924
Trihexyl trimellitate 13 43 43–550 y = 11.3x+ 22.1 0.9994
Tinosorb OMC 10 33 33–500 y = 8.3x+ 34.2 0.9982
Oleamide 75 253 253–7037 y = 0.41x+ 231 0.9900
Chimassorb 81 26 88 88–1020 y = 3.41x− 121 0.9965
Glyceryl monostearate 11 38 38–2630 y = 3.65x+ 178 0.9908 Compounds 27–32
Dinonyl phthalate 6 19 19–620 y = 4.79x+ 55 0.9928
Irganox 3114 94 314 314–5390 y = 0.05x+ 35 0.9900
Diisodecyl phthalate 4 13 13–434 y = 4.56x+ 58 0.9934
Irganox 1010 28 93 93–992 y = 0.48x+ 16.9 0.9985 Compounds 25 and 26
Erucamide 12 39 39–1120 y = 2.79x+ 33.9 0.9966 Compounds 51–57
Irganox 1076 13 42 42–1060 y = 1.52x+ 28.0 0.9918
Irgafos 168 27 91 91–2240 y = 1.32x+ 68.7 0.9979
Palmitic acid 238 792 792–5362 y = 0.84x+ 459 0.9900
Stearic acid 210 700 700–4740 y = 0.27x+ 548 0.9900
3,5-Di-tert-butyl− 4-hydroxybenzaldehyde 11 35 35–1029 y = 13.2x− 80.8 0.9991 Compound 18
Methyl 3-(3,5-di-tert-butyl−4- 10 35 35–1020 y = 2.56x+ 7.72 0.9998 Compounds 20, 21 and 22
hydroxyphenyl) propanoate
Benzenepropanoic acid, 3,5-bis(1,1- 16 52 52–1560 y = 4.25x+ 5.00 0.9986 Compound 24
dimethylethyl)− 4-hydroxy-
Glyceryl dihexadecanoate 57 189 189–2460 y = 2.97x− 162 0.9980 Compounds 33–41
N,N-Bis(2-hydroxyethyl) dodecylamine 12 39 39–1120 y = 35.9x+ 705 0.9979 Compounds 42–50 and
compounds 63 and 64
Octadecanamide, N,N′− 1,2-ethanediylbis- 11 37 37–1082 y = 1.39x+ 43.2 0.9976 Compounds 58–61
Octadecanamide 10 35 35–490 y = 8.36x+ 148 0.9975 Compounds 65–68 and
compounds 71–73
13-Docosenamide,N-hydroxy-, (13Z)- 9 30 30–450 y = 7.42x+ 55 0.9985
11-Eicosenamide, (11Z)- 12 40 40–475 y = 8.22 x + 111 0.9995
Irgafos 168 OXO 82 275 275–2980 y = 14.7x+ 796 0.9965

As can be seen, good results were obtained in terms of detection and
quantification limits, linearity and reproducibility. The LOD values
were between 10 µg/Kg, for most of the compounds, and 240 µg/Kg
(palmitic and stearic acid) which were detected in negative mode,
where the ionization was less sensitive.
Considering the high number of compounds identified, and the
great number of migration values obtained by twenty six samples stu-
died in at least three simulants, we decided to use a principal compo-
nent analysis (PCA) performed with Unscrambler v.9.7. to facilitate the
interpretation of data.
Fig. 3 shows the score plots (left graphics) and correlation loadings
plots (right graphics) for the different simulants used. The correlation in
all cases is very high with values of accumulated variance between 71%
and 99%.
For ethanol 95% used as food simulant, when the results were
plotted as scores (graphic 1, left), 3 groups of samples were differ-
entiated. The group 1 included the samples PP19, PP20 and PP11 where
the highest migration values of glyceryl monostearate obtained (be-
tween 14.5 ± 0.8 and 33.3 ± 1.2 mg/ Kg of simulant) were dis-
criminated for this group. This fact was represented in the correlation
loadings plot (right) where the compound 7 was located near the
border. The group 2 was constituted by the samples PP5 and PP15
which had the highest values of erucamide migration (compound 12)
24.1 ± 1.7 and 38.3 ± 2.1 mg/Kg of simulant respectively. Other
discriminatory migration values for this group were the migration of
the even dihydroxy alquilamines (42, 43, 44, 46, 48, 49 and 59) and
some ceramides (54 and 57) shown in graphic 1, at right. The rest PPs
were grouped in group 3.
The score and correlation loadings plots obtained for the simulants
ethanol 10% and acetic 3% were very similar. In both cases, the values
of migration were the lowest compared to other simulants and closer
between these ones. As shows Fig. 3 (Graphics 2 and 3, left) the samples
are plotted together in 4 groups. The first one was formed by the
samples PP6 and PP7 characterized by the highest values of odd di-
hydroxy alquilamines for these simulants (compounds 45 and 47 in
graphics 2 and 3, left). Values were between 0.07 ± 0.01 and
2.25 ± 0.33 mg/Kg of simulant. The second group was constituted by
the samples PP3 and PP9, they were also discriminated by the migra-
tion of these odd dihydroxy alquilamines (compounds 45 and 47), with
values lower than those mentioned above and between 0.04 ± 0.01
and 1.21 ± 0.06 mg/ Kg of simulant. The group 3 only consisted of
sample PP5 whose migration values of even dihydroxy alquilamines
were the highest shown in the graphics 2 and 3, right, with values
between 0.07 ± 0.01 and 5.41 ± 0.33. The rest of PPs were within
the group 4.
Finally, the graphics 4 expressed the score and correlation loadings
plots obtained for Tenax. Three groups of samples were constituted for
this simulant. In the first one, the samples PP 16 and 18 were grouped
due to their highest and discriminatory migration values of compounds
glyceryl monopalmitate and both odd hydroxyamides (compounds 29,
45 and 47 showed graphic 4, right). Also some glyceryl compounds (28,
32, 33 and 37) and the break-down product as ethyl 3-(3,5-di-tert-
butyl-4-hydroxyphenyl) propanoate (20) were characteristic of these
samples. On the other hand, the samples PP11, PP19 and PP20 were
together in group 2 with the highest migration values of glyceryl
monostearate (7) as occurred for ethanol 95%.
3.3. Risk assessment
In order to apply the risk assessment to these polymers, all values of
migration were compared to their SML (10/2011/EU [13]) or to the
maximum values according by Cramer for those non listed compounds
in the EU regulation 10/2011/EU.
Thirty three compounds appeared in the legislation 10/2011/EU:
oleamide, erucamide, irganox 1010, irgafos 168, palmitic and stearic
acid, hexa-octa-decanamide N,N′-1,2-ethanediylbis-, 11-eicosenamide,

759
P. Vera et al. Talanta 188 (2018) 750–762

Migration results for ethanol 95 % as food simulant

Migration results for ethanol 10 % as food simulant

Migration results for acetic 3% as food simulant

Migration results for Tenax as food simulant

Fig. 3. Scores plots and correlation loading plots corresponding to the principal components PC1 and PC2 for the different simulants studied.

760
P. Vera et al.
(11Z) and the family 2 called “with glycerol molecule” and all of them
are authorized without SML. The family 3 called “dyhydroxy alquila-
mines” have SML expressed as group where the sum of the compounds
[N,N-Bis(2-hydroxyethyl) (C8-C18) amine] must not exceed 1.2 mg/Kg.
However, the samples PP5, PP6 and PP7 didn’t comply with the legis-
lation for none simulant as these compounds surpass this limit. Also
PP11 for ethanol 95% and PP3 for ethanol 95% and acetic 3% exceeded
this SML.
Other compounds like NX8000, irganox 3114 and irganox 1076 had
a SML of 5, 5 and 6 mg/Kg respectively. The NX8000 and irganox 3114
concentrations were below their specific migration limit in all samples.
However, irganox 1076 was found in the sample PP25 with a con-
centration of 10.2 ± 1.85 mg/Kg for ethanol 95% surpassing its SML
(6 mg/Kg).
Concerning the non-listed compounds in the positive list of the leg-
islation 10/2011/EU, their concentrations were compared with the
maximum values according by Cramer classes. Six samples overcame the
maximum concentration recommended by Cramer for the simulant
ethanol 95%. PP25 due to the compound of Class II, hexyl 3-(3,5-di-tert-
butyl-4-hydroxyphenyl) propanoate with a concentration of
4.19 ± 0.06 mg/Kg. And benzenepropanoic acid, 3,5-bis(1,1-dimethy-
lethyl)-4-hydroxy-, 1,1′-[2,2-bis(hydroxymethyl)-1,3-propanediyl] ester
and benzenepropanoic acid, 3,5-bis(1,1-dimethylethyl)-4-hydroxy-, 1,1′-
[2-[[3-[3,5-bis(1,1-dimethylethyl)-4-hydroxyphenyl]-1-oxopropoxy]me-
thyl]-2-(hydroxymethyl)-1,3-propanediyl] ester with concentrations of
2.05 ± 0.22 and 0.89 ± 0.03 for sample PP26. The samples PP5, PP10,
PP11 and PP15 didn’t comply with Cramer limits for different types
compounds of Class III such as the group of dihydroceramides and tetra-
penta and heptadecanamide,N,N′-1,2-ethanediylbis- in ethanol 95% as
simulant, although they complied for the aqueous simulants.
4. Conclusions
Non-volatile migrants from twenty six PP films to be used as food
packaging materials have been identified and quantified. The study has
been carried out using UPLC/MS/QTOF, which has demonstrated to be a
powerful technique for the elucidation of unknown compounds, based on
the possibility of obtaining the accurate mass of precursor and their
fragment ions. A total of seventy five compounds were identified. Sixteen
were additives like antioxidants, plasticizers, lubricants…and the rest, fifty
seven, were NIAS. They could come from rupture of these additives due to
their similar structures, like derivatives of irganox 1010 and 1076 or the
group of glycerol molecules. Also impurities, as N,N-bis(2-hydroxyethyl)
amines, or compounds of unknown origin, like hydro-ceramides. These
results reinforced the necessity of analyzing in depth the migration sample
after exposure in order to identify the NIAS compounds.
The quantification of these identified compounds has been carried
out at low detection limits around 10 ppb. The results of the migration
assays showed that 35% of the films didn´t comply, six of these ex-
ceeded the SML for European legislation 10/2011/EU [13] (five sam-
ples due to group N,N-bis(2-hydroxyethyl) amine and one sample due
to irganox 1076). Six samples overcame the maximum concentration
recommended by Cramer for the compound hexyl 3-(3,5-di-tert-butyl-
4-hydroxyphenyl) propanoate derivative of irganox 1010 and 1076
among others.
And finally, despite the complexity of NIAS identification, this work
provides a lot of new compounds that could be important for future
works. And also, it opens up a wide range of possibilities for the plastic
manufactured companies that can reformulate and substitute some
additives which are related with these NIAS found.
Acknowledgements
The authors acknowledge Projects AGL-2012-37886, AGL2015-
67362-P from MINECO (Spain) and European Social Funds given to
GUIA group T-10 (University of Zaragoza).
761
Talanta 188 (2018) 750–762
References
[1] S. Allahvaisi, Polypropylene in the industry of food packaging, in: F. Dogan (Ed.),
Polypropylene, InTech, 2012.
[2] C. Bolognesi, L. Castle, J.P. Cravedi, K.H. Engel, P. Fowler, R. Franz, K. Grob,
R. Gurtler, T. Husoy, S. Karenlampi, W. Mennes, M.R. Milana, A. Penninks,
V. Silano, A. Smith, M.D.T. Pocas, C. Tlustos, D. Wolfle, H. Zorn, C.A. Zugravu,
E.P.F.C. Mat, Safety assessment of the substance ground sunflower seed hulls, for
use in food contact materials, EFSA J. 14 (7) (2016).
[3] M.S. Dopico-Garcia, J.M. Lopez-Vilarino, M.V. Gonzalez-Rodriguez, Antioxidant
content of and migration from commercial polyethylene, polypropylene, and
polyvinyl chloride packages, J. Agric. Food Chem. 55 (8) (2007) 3225–3231.
[4] L. Barp, M. Suman, F. Lambertini, S. Moret, Migration of selected hydrocarbon
contaminants into dry semolina and egg pasta packed in direct contact with virgin
paperboard and polypropylene film, Food Addit. Contam. Part a-Chem. Anal.
Control Expo. Risk Assess. 32 (9) (2015) 1542–1551.
[5] C. Kirchnawy, J. Mertl, V. Osorio, H. Hausensteiner, M. Washuttl, J. Bergmair,
M. Pyerin, M. Tacker, Detection and identification of oestrogen-active substances in
plastic food packaging migrates, Packag. Technol. Sci. 27 (6) (2014) 467–478.
[6] M. Rani, W.J. Shim, G.M. Han, M. Jang, N.A. Al-Odaini, Y.K. Song, S.H. Hong,
Qualitative analysis of additives in plastic marine debris and its new products, Arch.
Environ. Contam. Toxicol. 69 (3) (2015) 352–366.
[7] Y. Kawamura, R. Yonezawa, T. Maehara, T. Yamada, Determination of additives in
food contact polypropylene, J. Food Hyg. Soc. Jpn. 41 (2) (2000) 154–161.
[8] M. Onghena, E. Van Hoeck, N. Negreira, L. Quirynen, J. Van Loco, A. Covaci,
Evaluation of the migration of chemicals from baby bottles under standardised and
duration testing conditions, Food Addit. Contam. Part a-Chem. Anal. Control Expo.
Risk Assess. 33 (5) (2016) 893–904.
[9] M. Onghena, N. Negreira, E. Van Hoeck, L. Quirynen, J. Van Loco, A. Covaci,
Quantitative determination of migrating compounds from plastic baby bottles by
validated GC-QqQ-MS and LC-QqQ-MS methods, Food Anal. Methods 9 (9) (2016)
2600–2612.
[10] M. Onghena, E. van Hoeck, P. Vervliet, M.L. Scippo, C. Simon, J. van Loco,
A. Covaci, Development and application of a non-targeted extraction method for the
analysis of migrating compounds from plastic baby bottles by GC-MS, Food Addit.
Contam. Part a-Chem. Anal. Control Expo. Risk Assess. 31 (12) (2014) 2090–2102.
[11] C. Simoneau, L. Van den Eede, S. Valzacchi, Identification and quantification of the
migration of chemicals from plastic baby bottles used as substitutes for poly-
carbonate, Food Addit. Contam. Part a-Chem. Anal. Control Expo. Risk Assess. 29
(3) (2012) 469–480.
[12] Regulation (EC) No 1935/2004 of the European Parliament and of the Council of 27
October 2004 on materials and articles intended to come into contact with food and
repealing Directives 80/590/EEC and 89/109/EEC, Regulation (EC) No 1935/2004
of the European Parliament and of the Council of 27 October 2004 on materials and
articles intended to come into contact with food and repealing Directives 80/590/
EEC and 89/109/EEC.
[13] COMMISSION REGULATION (EU) No 10/2011 of 14 January 2011 on plastic
materials and articles intended to come into contact with food, COMMISSION
REGULATION (EU) No 10/2011 of 14 January 2011 on plastic materials and ar-
ticles intended to come into contact with food.
[14] Y.N. Xia, M. Rubino, Effect of cut edge area on the migration of BHT from poly-
propylene film into a food simulant, Polym. Test. 51 (2016) 190–194.
[15] H.T. Chi, Y. Liu, X. Gao, W.L. Liu, G.H. Hu, Y.J. Luo, Study on migration model of
antioxidants in food contact polypropylene plastics, Chin. J. Anal. Chem. 43 (3)
(2015) 399–403.
[16] J. Alin, M. Hakkarainen, Type of polypropylene material significantly influences the
migration of antioxidants from polymer packaging to food simulants during mi-
crowave heating, J. Appl. Polym. Sci. 118 (2) (2010) 1084–1093.
[17] J.A. Garde, R. Catala, R. Gavara, R.J. Hernandez, Characterizing the migration of
antioxidants from polypropylene into fatty food simulants, Food Addit. Contam. 18
(8) (2001) 750–762.
[18] C. Caner, B. Harte, Effect of high-pressure processing on the migration of anti-
oxidant Irganox 1076 from polypropylene film into a food simulant, J. Sci. Food
Agric. 85 (1) (2005) 39–46.
[19] L. Coulier, H.G.M. Orbons, R. Rijk, Analytical protocol to study the food safety of
(multiple-)recycled high-density polyethylene (HDPE) and polypropylene (PP)
crates: influence of recycling on the migration and formation of degradation pro-
ducts, Polym. Degrad. Stab. 92 (11) (2007) 2016–2025.
[20] C.F. Li, Y. Li, Z.N. Chen, F. Liang, X.H. Chen, S.J. Wu, Y.T. Li, X.Y. Sun,
Simultaneous determination of antioxidants and ultraviolet absorbers by ultra-
performance liquid chromatography in food simulants, Food Anal. Methods 7 (9)
(2014) 1755–1762.
[21] L. Incarnato, P. Scarfato, M.R. Milana, R. Feliciani, D. Acierno, Migrational beha-
viour of food grade PP, J. Polym. Eng. 20 (2) (2000) 121–141.
[22] S. Biedermann-Brem, N. Kasprick, T. Simat, K. Grob, Migration of polyolefin oli-
gomeric saturated hydrocarbons (POSH) into food, Food Addit. Contam. Part a-
Chem. Anal. Control Expo. Risk Assess. 29 (3) (2012) 449–460.
[23] G. McCombie, K. Hotzer, J. Daniel, M. Biedermann, A. Eicher, K. Grob, Compliance
work for polyolefins in food contact: results of an official control campaign, Food
Control 59 (2016) 793–800.
[24] M. Onghena, E. Van Hoeck, J. Van Loco, M. Ibanez, L. Cherta, T. Portoles,
E. Pitarch, F. Hernandez, F. Lemiere, A. Covaci, Identification of substances mi-
grating from plastic baby bottles using a combination of low-resolution and high-
resolution mass spectrometric analysers coupled to gas and liquid chromatography,
J. Mass Spectrom. 50 (11) (2015) 1234–1244.
P. Vera et al.
[25] J. Alin, M. Hakkarainen, Microwave Heating causes rapid degradation of anti-
oxidants in polypropylene packaging, leading to greatly increased specific migra-
tion to food simulants as shown by ESI-MS and GC-MS, J. Agric. Food Chem. 59 (10)
(2011) 5418–5427.
[26] J.S. Felix, F. Isella, O. Bosetti, C. Nerin, Analytical tools for identification of non-
intentionally added substances (NIAS) coming from polyurethane adhesives in
multilayer packaging materials and their migration into food simulants, Anal.
Bioanal. Chem. 403 (10) (2012) 2869–2882.
[27] C. Nerin, D. Acosta, Behavior of some solid food simulants in contact with several
plastics used in microwave ovens, J. Agric. Food Chem. 50 (25) (2002) 7488–7492.
[28] M. Lahimer, N. Ayed, J. Horriche, S. Belgaied, Characterization of plastic packaging
additives: food contact, stability and toxicity, Arab. J. Chem. (2013).
[29] L. Cherta, T. Portoles, E. Pitarch, J. Beltran, F.J. Lopez, C. Calatayud, B. Company,
F. Hernandez, Analytical strategy based on the combination of gas chromatography
coupled to time-of-flight and hybrid quadrupole time-of-flight mass analyzers for
non-target analysis in food packaging, Food Chem. 188 (2015) 301–308.
[30] M.J. Martinez-Bueno, S. Cimmino, C. Silvestre, J.L. Tadeo, A.I. Garcia-Valcarcel,
A.R. Fernandez-Alba, M.D. Hernando, Characterization of non-intentionally added
substances (NIAS) and zinc oxide nanoparticle release from evaluation of new an-
timicrobial food contact materials by both LC-QTOF-MS, GC-QTOF-MS and ICP-MS,
Anal. Methods 8 (39) (2016) 7209–7216.
[31] B. Geueke, Dossier-non-intentionally added subtances (NIAS), Food Packag. Forum
(2013).
[32] M. Vinaixa, E.L. Schymanski, S. Neumann, M. Navarro, R.M. Salek, O. Yanes, Mass
spectral databases for LC/MS- and GC/MS-based metabolomics: state of the field
and future prospects, Trac-Trends Anal. Chem. 78 (2016) 23–35.
[33] M. Aznar, A. Rodriguez-Lafuente, P. Alfaro, C. Nerin, UPLC-Q-TOF-MS analysis of
non-volatile migrants from new active packaging materials, Anal. Bioanal. Chem.
404 (6–7) (2012) 1945–1957.
[34] P. Vera, E. Canellas, C. Nerin, Identification of non-volatile compounds and their
migration from hot melt adhesives used in food packaging materials characterized
by ultra-performance liquid chromatography coupled to quadrupole time-of-flight
mass spectrometry, Anal. Bioanal. Chem. 405 (14) (2013) 4747–4754.
[35] E. Canellas, P. Vera, C. Nerin, UPLC-ESI-Q-TOF-MSE and GC-MS identification and
quantification of non-intentionally added substances coming from biodegradable
food packaging, Anal. Bioanal. Chem. 407 (22) (2015) 6781–6790.
[36] E. Canellas, P. Vera, C. Nerin, Multiple headspace-solid phase microextraction for
the determination of migrants coming from a self-stick label in fresh sausage, Food
Chem. 197 (2016) 24–29.
[37] E. Canellas, P. Vera, C. Nerin, Risk assessment derived from migrants identified in
762
View publication stats
Talanta 188 (2018) 750–762
several adhesives commonly used in food contact materials, Food Chem. Toxicol. 75
(2015) 79–87.
[38] UNE-EN 14338:2004. Papel y cartón para contacto alimentario. Condiciones para la
determinación de la migración en papel y cartón utilizando óxido de polifenileno
modificado (MPPO) como simulante., UNE-EN 14338:2004. Papel y cartón para
contacto alimentario. Condiciones para la determinación de la migración en papel y
cartón utilizando óxido de polifenileno modificado (MPPO) como simulante.
[39] Threshold of toxicological concern (TTC), ILSI Europe concise monograph series
(2005) threshold of toxicological concern (TTC), ILSI Eur. Concise Monogr. Ser.
(2005).
[40] Risk Assessment of non-listed substances (NLS) and not-intentionally added sub-
stances (NIAS) under article19 of Plastic Europe Risk Assessment of non-listed
substances (NLS) and not-intentionally added substances (NIAS) under article19 of
Plastic Europe
[41] C. Nerin, P. Alfaro, M. Aznar, C. Domeno, The challenge of identifying non-in-
tentionally added substances from food packaging materials: a review, Anal. Chim.
Acta 775 (2013) 14–24.
[42] K. Bhunia, S.S. Sablani, J. Tang, B. Rasco, Migration of chemical compounds from
packaging polymers during microwave, conventional heat treatment, and storage,
Compr. Rev. Food Sci. Food Saf. 12 (5) (2013) 523–545.
[43] R. Rodil, J.B. Quintana, G. Basaglia, M.C. Pietrogrande, R. Cela, Determination of
synthetic phenolic antioxidants and their metabolites in water samples by down-
scaled solid-phase extraction, silylation and gas chromatography-mass spectro-
metry, J. Chromatogr. A 1217 (41) (2010) 6428–6435.
[44] R.Z. Liu, T. Ruan, S.J. Song, Y.F. Lin, G.B. Jiang, Determination of synthetic phe-
nolic antioxidants and relative metabolites in sewage treatment plant and recipient
river by high performance liquid chromatography-electrospray tandem mass spec-
trometry, J. Chromatogr. A 1381 (2015) 13–21.
[45] N. Gerlach, M. Mentel, T. Kohler, B. Tuchscherer, B. Garbe, J. Ulker, H. Tronnier,
U. Heinrich, M. Farwick, Effect of the multifunctional cosmetic ingredient sphin-
ganine on hair loss in males and females with diffuse hair reduction, Clin. Cosmet.
Investig. Dermatol. 9 (2016) 191–203.
[46] S.M. Choi, B.M. Lee, Safety and risk assessment of ceramide 3 in cosmetic products,
Food Chem. Toxicol. 84 (2015) 8–17.
[47] F.C. Fernandes, R. Gadioli, E. Yassitepe, M.A. De Paoli, Polyamide-6 composites
reinforced with cellulose fibers and fabricated by extrusion: effect of fiber bleaching
on mechanical properties and stability, Polym. Compos. 38 (2) (2017) 299–308.
[48] M.A. Farajzadeh, M.R. Vardast, M. Bahram, Optimization of Dispersive liquid-liquid
microextraction of Irganox 1010 and Irgafos 168 from polyolefins before liquid
chromatographic analysis, Chromatographia 69 (5–6) (2009) 409–419.