Integrated Environmental Assessment and Management — Volume 00, Number 00—pp.

1–12
Received: 10 October 2018 | Returned for Revision: 8 January 2019 | Accepted: 20 June 2019 1

Environmental Policy & Regulation

Is It Safe to Paint Your Wall White? A Case Study on Titanium

Dioxide Classification
E Elina Kähkönen*†
†Aalto University, Design Factory, Espoo, Finland

ABSTRACT
Titanium dioxide (TiO2) is in the process of being classified as a suspected carcinogenic substance (Carc 2). The present
case study probes the outcomes of this potential classification in terms of the reduction of hazardous exposure to TiO2 due
to its classification. Furthermore, the case study examines the elements that are causing ambiguity during the classification
process. This study was conducted by walking through the process from the present exposure to TiO2 to the hazard
assessment associated with TiO2 exposure, to the regulatory classification process, and to practical outcomes affecting TiO2
usage. Finally, the impact of the classification on exposure, which was originally considered potentially hazardous, is eval-
uated. The case study shows that TiO2 classification as a carcinogen will not directly reduce respiratory exposure to TiO2,
which was the original reason for the classification. Instead, the classification will lead to restrictions on recycling. Moreover,
the classification will have an impact on certain solid artifacts and liquid mixtures for which hazardous exposure was not
detected. Altogether, the present case raises questions concerning hazard communications associated with the Carc 2
classification; treatment of poorly soluble low toxicity (PSLT) particles and nanoparticles in the Registration, Evaluation,
Authorisation and Restriction of Chemicals (REACH) and the Classification, Labelling and Packaging (CLP) classifications; and
use of human exposure studies for the purposes of chemical regulations. Based on the present study, the following rec-
ommendations are made: the final decision on the TiO2 classification should be reconsidered together with those of other
PSLT particles and take into account extensive developments in the field of nanoscience. Furthermore, the European
Chemicals Agency (ECHA) should develop state‐of‐the‐art guidance on how to use the available human exposure data.
Finally, the authorities that are in charge of European Union chemicals management are advised to further develop the
regulatory network to utilize the information generated in REACH processes as efficiently as possible and to verify that the
connections between the regulations result in the intended outcome. Integr Environ Assess Manag 2019;00:1–12. © 2019
SETAC

Keywords: REACH regulation CLP regulation Titanium dioxide Hazard assessment Hazard communication

INTRODUCTION from nanoscale to microscale, and TiO2 is commonly placed

In the near future, consumers buying white paint may see
a “Health hazard” warning label (Figure 1) and the text
“Suspected of causing cancer via inhalation” on the paint
can, which was formerly marked with an ecolabel. The
reason behind this change lies in a new harmonized classi-
fication of titanium dioxide (TiO2)—the white pigment in
paint and ink—as a carcinogenic substance. Globally, TiO2
is, by far, the most widely used pigment, with an average
annual production of more than 30 million metric tons from
1916–2011 (Jovanović 2014). Titanium dioxide is used in a
wide range of applications, from foodstuff to cosmetics and
technochemical products. Titanium dioxide is an inorganic
substance with differing crystal structures and surface
chemistries. Additionally, the particle size of TiO2 ranges
* Address correspondence to elina.kahkonen@aalto.fi
Published 9 July 2019 on wileyonlinelibrary.com/journal/ieam.
in the group of “poorly soluble low toxicity particles” (PSLT
particles) (ECHA 2017a). Furthermore, the nanoparticles of
TiO2 belong to a keenly studied group of substances. A
previous study of nanoparticles focused on hazard assess-
ment and the development of hazard assessment methods
suitable for nanoparticles (Drobne 2018). Development in
this field is keenly followed by the European Chemicals
Agency (ECHA) Nanomaterial Working Group, which seeks
to find common ground with regard to hazard evaluation
and regulatory implementation for nanomaterials (ECHA
2015a). The warning label on a paint can and the text con-
stitute one, but not the only, result of this classification. In-
deed, the emerging classification establishes new
restrictions for the use of TiO2 from raw material to product,
recycling, and as a waste. Evidently, TiO2 is not the only
substance that has been given a carcinogenicity classi-
fication. However, this case provides an interesting and
topical insight for the implementation of major European

Integr Environ Assess Manag 2019:1–12 DOI: 10.1002/ieam.4186 © 2019 SETAC

2 Integr Environ Assess Manag 000, 2019—EE Kähkönen
Figure 1. Health hazard‐warning label.
Union (EU) chemical regulations: Regulation 1907/2006
concerning the Registration, Evaluation, Authorisation and
Restriction of Chemicals (REACH) and Regulation 1272/2008
on Classification, Labelling and Packaging (CLP) of sub-
stances and mixtures, which aim at “ensuring high level of
protection of human health and the environment” (EC 2006,
2008a). Notably, CLP is an adaptation of the Globally
Harmonized System (GHS) of classification and labeling of
chemicals in the European Union (ECHA 2019a). The GHS
has been developed within the United Nations to unify
globally the communication on substances and mixtures.
Specifically, the pictograms for hazard labeling and hazard
classification are globally consistent (ECHA 2019a). The
prevalence of these pictograms is increasing, along with a
continually increasing number of substances in Annex 4 of
the CLP on harmonized classifications for substances (EC
2008a). In addition to the official warnings, consumers have
access to ubiquitous digital information. Riley (2014) pres-
ents both increasing prevalence and digital information as
potential reasons for diminishing attention to warnings.
Furthermore, Riley (2014) is concerned about reducing the
ability for source criticism among people. Together, these
factors do not necessarily lead to an optimal situation re-
garding safe choices in the use of chemical products, which
would result in the desired reduction in unwanted exposure
to chemical substances. The present case study on the TiO2
classification examines the process behind the simplified
warning label on consumer products and aims to 1) pinpoint
the ambiguous process steps that may lead to differing
conclusions regarding hazard assessment and communica-
tion, and 2) assess the outcomes of the entire process in
terms of improved health protection related to products
containing TiO2.
Briefly, the goal of improved health protection through
hazard assessment and communication may be framed as
follows: When a product contains a hazardous substance,
the user will be aware of the hazard and take adequate
Integr Environ Assess Manag 2019:1–12 wileyonlinelibrary.com/journal/ieam
protection for the safe use of the product. This seemingly
simple, unambiguously agreed‐upon goal is, after all, far
from simple. First, the definition of a hazardous substance
calls for thorough scientific testing, and second, the com-
munication calls for a defined process for bringing the re-
sults of hazard assessment into relevant warnings and clear
instructions. In practice, REACH and the associated Regu-
lation on Test Methods (EC 2008b) focus on the first point,
whereas CLP focuses on the second point. As an indication
of its complexity, REACH with its Test Methods Regulation
and CLP have in total nearly 5000 pages (EC 2006, 2008a,
2008b).
The REACH regulation lays a common ground for as-
sessing hazards related to substances that are placed on the
EU market (as such, or as mixtures or in articles). The tools in
the hazard assessment comprise physicochemical, ecotox-
icological, and toxicological tests for the substances (EC
2008b). The present study focuses on the toxicological and,
more specifically, on the carcinogenicity studies; a brief
overview on these test procedures is given here.
Carcinogenicity assessment in brief
Carcinogenicity studies comprise toxicological assess-
ments in which human, animal, or bacteria cells, or animals,
are exposed to the substance being tested. The responses
in terms of neoplasm development, chromosomal changes,
micronucleus formation, and DNA and gene mutations, for
example, indicate carcinogenicity (EC 2008b). The mode of
action in cancer evolvement in test animals is considered
highly important in assessing interspecies correlations. The
mode of action assessment is built on the assessment of
biochemical, cellular, and molecular key events (OECD
2012). In addition, available data on human exposure, such
as epidemiological studies on occupational exposure, are
used (EC 2008a).
Communication on carcinogenic substances and products
The CLP regulation sets rules for translating the hazard
assessment results into instructions for safe use and warn-
ings attached to a product that contains a hazardous sub-
stance. In the case of carcinogenicity, CLP translates the
toxicological test results into 3 human carcinogenicity cat-
egories: Carc 1A: Known to have carcinogenic potential for
humans; Carc 1B: Presumed to have carcinogenic potential
for humans; and Carc 2: Suspected human carcinogens
(EC 2008a).
The carcinogenicity classification is made on the basis of
strength of evidence, that is, according to demonstrated
causality between the exposure to a substance and the
development of cancer (sufficient evidence) or a positive
association between the exposure and the development of
cancer (limited evidence). Further, additional consid-
erations, such as tumor type, multisite responses, responses
in one or both sexes, responses in a single or several spe-
cies, and mode of action and its relevance for humans, are
made (EC 2008a).
© 2019 SETAC
A Case Study on TiO2 Classification—Integr Environ Assess Manag 000, 2019
Carc 1A and 1B trigger hazard statement H350: May cause
cancer, and Carc 2 triggers the hazard statement H351:
Suspected of causing cancer. All categories trigger the
“Health hazard” warning label (Figure 1) when the defined
concentration of a classified substance is exceeded. In the
cases of Carc 1 (A and B), the concentration limit is 0.1 weight
% (wt%) of the Carc 1A or 1B classified substance in a mix-
ture, and in the case of Carc 2 the limit is 1 wt% (EC 2008a).
Harmonized classification process
Substances are mainly self‐classified by companies that
are placing the substances on the EU market. To achieve a
uniform EU‐wide classification for a substance, the Member
State Competent Authority (MSCA), a manufacturer, im-
porter, or downstream user of a substance can initiate a
harmonized classification process. A simplified process from
its initiation to the final inclusion in the Annex 6 of the CLP
regulation proceeds as follows. First, the ECHA receives a
notice of intention and a dossier for a harmonized classi-
fication dossier preparation from an entitled party. The
ECHA publishes these intentions and dossiers on its website
for public consultation and comment. Second, public con-
sultation, during which interested parties can comment on
the harmonized classification dossier, lasts for 60 d. Third,
ECHA’s Committee for Risk Assessment (RAC), which is
composed of 54 members, develops the RAC opinion
based on the dossier and its comments jointly with the
parties concerned. The RAC opinion on the harmonized
classification proposal is adopted within 18 mo from re-
ceiving the dossier on the classification proposal. The
opinion is published on the ECHA web page. Finally, ECHA
sends the RAC opinion to the European Commission, which
makes the decision on the inclusion of the substance in
Annex 6 of the CLP regulation (ECHA 2018) (Figure 2).
Previous studies related to TiO2 classification
Previous reviews and comments on the specific case of
TiO2 are published by the parties involved in the process,
Figure 2. Harmonized classification process. ECHA = European Chemicals Agency; RAC = ECHA’s Committee for Risk Assessment.
Integr Environ Assess Manag 2019:1–12 DOI: 10.1002/ieam.4186
3
namely, the authorities (ANSES 2016; ECHA 2017a) and
industry (CEPE 2017; TDMA 2018). The other previous re-
search on REACH and CLP focuses largely on various test
methods and their applicability in a regulatory context (e.g.,
Cappelli et al. 2015; Alépée et al. 2017; Sobanska et al.
2017) and in the specific impacts of REACH and CLP on
products and applications that are not directly within the
scope of these regulations, such as wastes and recycling
(Bodar et al. 2018), as well as food materials (Geueke and
Muncke 2017). Moreover, references to bending regulatory
situation for substance groups regarding their assessment
according to REACH, for example, nanomaterials and plas-
tics, are discussed (Walser and Studer 2014; Raucher et al.
2016; Steensgaard et al. 2017; Drobne 2018). However,
none of these studies display a crosscutting approach
similar to that of the present study.
CASE STUDY
The present study probes the TiO2 classification process,
its impacts on TiO2 exposure, and the consequent reduction
of the risk related to TiO2 exposure. The research questions
are as follows:
1) How will the TiO2 classification process improve safety in
the production and use of TiO2?
2) Which steps in the TiO2 classification process are am-
biguous with multiple interpretations?
The present study approaches the research questions by
examining the TiO2 classification process in 5 steps
(Figure 3).
1) The first step focuses on the current exposure to TiO2
that may lead to health impacts. Here, the core attempt
is to define the starting point from which stems the
further process for safety improvements. The data here
are mainly based on the publication of the International
Agency for Research of Cancer (IARC) on TiO2
(IARC 2010).
© 2019 SETAC
4 Integr Environ Assess Manag 000, 2019—EE Kähkönen
Figure 3. Case study approach.
2) The second step continues with the studies on occupa-
tional exposure and animal test on toxicity of TiO2. The
focus here is in summarizing different toxicity studies and
the differing results that are achieved from the studies.
Here, the data are collected from the review of IARC
(2010) on the carcinogenicity of TiO2 (and carbon black
and talcum); the French Agency for Food, Environmental
and Occupational Health and Safety (ANSES) proposal
on harmonized classification of TiO2 (ANSES 2016); and
the European Food Safety Authority (EFSA) assessment
on TiO2 (EFSA 2016).
3) The third step reviews the decisions and dialogue during
the 4‐step classification process. In this step the analysis
focuses on the different interpretations of the test results
by different parties in the process. Here, the primary
sources of information are the RAC opinion (ECHA 2017)
and ANSES (2016).
4) The fourth step evaluates the impacts of the carcinogen
classifications in TiO2 use as a raw material, in products,
and as waste. Here, the analysis focuses on detecting
connections between different guidance and CLP classi-
fications. During this stage the regulations, directives,
and other guidance documents that refer to hazard
classification are reviewed.
5) The fifth step assesses how the exposure to TiO2 will
change with the reclassification. The focus here is on the
realization of how the use of TiO2 will change after
the classification and how this change in use will impact
the exposure.
CASE STUDY APPROACH
Current TiO2 exposure
The use of TiO2 extends from the paint, paper, and
plastics industries to those producing foodstuffs, pharma-
ceuticals, and cosmetics. Titanium dioxide is used in a wide
range of products, such as paper, ink, white prints, chewing
gum, pharmaceuticals, and sunscreens.
The routes for TiO2 exposure are oral, respiratory, and
dermal contact. Respiratory exposure is the most discussed
topic, and it is the highest in TiO2 production. Notably, the
exposures have decreased from about 14 mg/m3 in the
1970s to about 3 mg/m3 in the 2000s (Fryzek et al. 2003).
The current occupational exposure standards of 25
Integr Environ Assess Manag 2019:1–12 wileyonlinelibrary.com/journal/ieam
countries for TiO2 exposure range from 3 to 10 mg/m3 (8‐h
average) of TiO2 in the air (IARC 2010). Respiratory exposure
to TiO2 also takes place during painting. However, the ex-
posure to paints that contain TiO2 may not lead to exposure
to TiO2 particles, given that the TiO2 in paint is a component
of a liquid mixture instead of dry particles (Boffetta
et al. 2001).
The highest dermal exposure to TiO2 is achieved when
using sunscreens, which may contain up to 25% of TiO2
(EFSA 2016). Furthermore, accidental splashes of paint or
touching almost any white surface, painted or paper, may
lead to dermal exposure to TiO2. However, the dermal TiO2
exposure in the latter cases remains negligible compared to
that of sunscreen administration. Oral exposure may occur
when eating white‐colored candies, pharmaceuticals, or
when using toothpaste. According to EFSA (2016), the es-
timated maximum oral intake of TiO2 as a food additive is
32.4 mg/kg (body weight) per day, that is, 2.2 g/d for a 70‐kg
person.
Health impacts related to TiO2 exposure
A multitude of toxicological and epidemiological assess-
ments on TiO2 have been conducted in previous studies.
For instance, EFSA, IARC, and ANSES have compiled large
reviews that focus on the health impact of TiO2 use and
production. IARC (2010) reviewed approximately 60 studies,
which focus directly on the health impacts of TiO2 exposure,
whereas ANSES (2016) and EFSA (2016) reviewed about 80
and about 90 similar (partly the same) studies, respectively.
These reviews present findings on human exposure studies,
animal studies, and studies conducted with cells and genetic
material. These studies aim at determining the health im-
pacts related to oral, dermal, and respiratory exposure to
TiO2. All the mentioned reviews concluded that TiO2 shows
no carcinogenicity in oral or dermal use. The main concern
and debate is around respiratory exposure to TiO2.
The IARC (2010) reviewed 5 epidemiological studies on
TiO2 impacts in occupational respiratory exposure. In total,
more than 20 000 cases of exposure were screened in 5
studies on at least 1‐mo occupational exposure to TiO2.
None of the studies suggested a statistically significant in-
crease in any type of cancers. In addition to these, 6 case
studies on 1 to 13 individuals who were exposed to TiO2
were reported. The studies disclosed, for instance, TiO2
findings in blood and lungs (IARC 2010). Both ANSES and
© 2019 SETAC
A Case Study on TiO2 Classification—Integr Environ Assess Manag 000, 2019
IARC (2010) conclude that there is no evidence for carci-
nogenicity of TiO2 based on human exposure. The major
critique of the human exposure studies, case studies, or
epidemiological studies was on uncertainties related to the
background information (other exposure, including tobacco
smoking) and to the exposure levels and time. For instance,
the relevance of 1 mo of exposure for carcinogenicity
studies may well be questioned.
Reviews of IARC (2010) and ANSES (2016) compiled res-
piratory exposure studies on approximately 1000 test animals.
The inhaled TiO2 quantities varied from 5 to 250 mg/m3, and
the exposure times ranged from 3 to 24 mo. Different daily,
weekly, and monthly exposure cycles were used. A total of
5 inhalation studies on TiO2 carcinogenicity were included in
the IARC review. Three of the studies suggested no carcino-
genicity, and 2 studies found that TiO2 was carcinogenic. To
complement the inhalation exposure data, 27 animal studies
have been conducted with various rodents (rats, mice,
and hamsters in the order of prevalence) by intratracheal
instillation (TiO2 is led via a tube directly into the lungs) to
learn about the retention by and clearance of the lungs. The
results with female rats indicated carcinogenicity (IARC 2010).
To determine the mutagenicity and genotoxicity potential
of TiO2, in vitro studies were conducted on human, animal,
and bacteria cells, and on DNA plasmids. While IARC (2010)
summarized the results of 36 in vitro studies to be incon-
clusive, ANSES (2016) found also indications of a potential
direct genotoxic mode of action, which means that substance
is able to interact directly with DNA or other genetic material.
Classification of TiO2 according to CLP
Before the present process, 88% of the companies that
have registered TiO2 according to REACH have self‐classi-
fied TiO2 as nonclassified. The rest of the classifications in-
clude, for example, Carc 2 and 1B, Acute toxicity 4, Skin
irritation 2, and Mutagenicity 2 (Table 1) (ECHA 2019b).
The harmonized classification process started when ECHA
received an intention (17 March 2015) and a dossier for
harmonized classification (27 May 2016) from ANSES (ECHA
2019b). The proposed harmonized classification was Carc
1B, presumed to have carcinogenic potential for humans. A
public consultation was started on the ECHA website.
1) During the 60‐d public consultation period, 514 com-
ments were received: 5 from MSCAs, 38 from individuals,
and 471 from organizations. Four of the MSCAs agreed
with the proposed classification whereas 1 MSCA con-
sidered that due to uncertainties related to material
characterization, high concentrations, and relevance to
humans arising from species differences TiO2 would in-
stead meet the criteria of Carc 2: Suspected human
carcinogen. Industry opinions were in favor of sustaining
the nonclassified status (ECHA 2017).
2) The RAC examined both the dossier and the comments,
consequently forming an opinion regarding the 3 proposed
classifications: Carc 1B, Carc 2, and no classification. The
Integr Environ Assess Manag 2019:1–12 DOI: 10.1002/ieam.4186
5
RAC published its opinion on TiO2 harmonized classification
(14 September 2017) (ECHA 2017).
All parties agreed on the overall substance charac-
terization; TiO2 particles are of different sizes from
nanoscale to microscale, with fibrous or granular shapes
having various kinds of coating. The parties agreed on
giving different TiO2 qualities a uniform classification. In
general, TiO2 particles were considered to resemble
other PSLTs. All parties agreed that there is no evidence
for TiO2 carcinogenicity via oral or dermal admin-
istration. Moreover, it was agreed that human evidence
on respiratory exposure did not indicate carcinogenicity
in humans. In general, all parties agree that ex-
trapolation from animal exposure results to human re-
sponse is controversial due to interspecies differences
but also due to restricted animal lifespans in chronic
exposure testing (ANSES 2016; ECHA 2017). The major
differences in interpretations regarding the respiratory
exposure among ANSES, commentary opinions, and
final RAC opinions are listed in items a through c below.
It should be noted that “comments” in the points a
through c refers to extensive comments from industry.
The ECHA received contradictory comments from more
than 200 named companies and more than 90 named
associations around the world and representing TiO2
producers and users (ECHA 2017).
a) Substance properties: Comments questioned
whether the adverse impacts of TiO2 are related to
its chemistry or rather to TiO2 properties as insoluble
particles with similar particle toxicity regardless of
the chemical properties. The RAC agrees on the
particle toxicity aspect and that TiO2 properties are
equal to other PSLT particles. However, RAC con-
cludes that, according to CLP regulations, the toxicity
assessment results define the classification regardless
of the property that causes toxicity. This conclusion
applies equally to any other PSLT particles.
b) Human exposure: The parties considered differently
the significance of human evidence. Although such
evidence was seen as strong evidence supporting
nonclassification (as a human carcinogen), ANSES,
MSCAs, and RAC questioned the reliability of the
studies regarding the exact measurement of TiO2
exposure and the lack of background data on the
other exposures (e.g., tobacco smoking). Specifically,
neither RAC nor ANSES considered that the negative
results from human exposure are strong enough to
rule out the positive results on rat carcinogenicity.
c) Animal studies: In the case of the animal studies, the
parties had differing opinions regarding the dose,
response, and correlation between the animal carci-
nogenicity results and human carcinogenicity.
Major arguments regarding the dose included
whether the high overdose (up to 250 mg/m3 ) in
rat studies resulted in relevant data for human
carcinogenicity assessment. High dose results
were considered a key finding by ANSES, whereas
© 2019 SETAC
 6

Table 1. TiO2 self‐classifications as published in ECHA’s CLP inventorya

Hazard class and Prevalence in self‐

category Hazard statement Hazard statement as written in generalized form Self‐classification (nr) classification (%)

Not classified 3909 88

Integr Environ Assess Manag 2019:1–12

Carc 2 H351 Suspected of causing cancer (state route of exposure if it is conclusively proven that 170 4
no other routes of exposure cause the hazard)

Acute Tox 4 H332 81 2

STOT SE 3 H335 May cause respiratory irritation 81 2

Eye Irrit 2 H319 Causes serious eye irritation 74 2

STOT RE 1 H372 Causes damage to organs (state all organs affected, if known) through prolonged or 71 2
repeated exposure (state route of exposure if it is conclusively proven that no
other routes of exposure cause the hazard)

Skin Irrit 2 H315 Causes skin irritation 12 0.3

STOT SE 2 H371 May cause damage to organs (state all organs affected, if known) (state route of 11 0.2
exposure if it is conclusively proven that no other routes of exposure cause the
hazard)

wileyonlinelibrary.com/journal/ieam
Carc 1B H350 May cause cancer (state route of exposure if it is conclusively proven that no other 8 0.2
routes of exposure cause the hazard)

Muta 2 H341 Suspected of causing genetic defects (state route of exposure if it is conclusively 1 0.02
proven that no other routes of exposure cause the hazard)

Resp Sens 1B H334 May cause allergy or asthma symptoms or breathing difficulties if inhaled 1 0.02

Acute Tox = acute toxicity; Carc 1B = presumed to have carcinogenic potential for humans; Carc 2 = suspected carcinogenic substance; CLP = Classification, Labelling and Packaging regulation; ECHA = European
Chemicals Agency; Eye Irrit = eye irritation; Muta = mutation; Resp Sens = respiratory sensitivity; Skin Irrit = skin irritation; STOT RE = specific target organ toxicity–repeat exposure; STOT SE = specific target organ
toxicity–single exposure.
a
ECHA 2019b.

© 2019 SETAC
Integr Environ Assess Manag 000, 2019—EE Kähkönen
A Case Study on TiO2 Classification—Integr Environ Assess Manag 000, 2019
comments from the industry claimed them to be
unrealistic compared to the maximal occupational
exposures. Moreover, the intratracheal method
of administering the dose was questioned in the
industry comments, whereas ANSES considered
the results as supporting evidence for carcinoge-
nicity. Instead, RAC argued that the evidence of
extremely high doses may not have a major role in
defining the classification.
Animal carcinogenicity of TiO 2 was tested in
mice, rats, and hamsters. The positive findings
were all in female rats. Here, ANSES considered
the sole test on mice as inconclusive because the
control group of mice had a high number of neo-
plasms. Accordingly, the negative result in the
case of hamsters was concluded to be due to an
effective lung clearance system and antioxidant
mechanisms, which differ between humans and
rats (ANSES 2016, p 61). The RAC concluded TiO 2
to be a carcinogen to rats but did not conclude
that there is evidence for TiO 2 carcinogenicity in 2
or more species. Moreover, RAC mentions TiO2 to
be a relatively weak rat carcinogen. The views on
the mode of action of TiO2 carcinogenicity and its
relevance to humans were different. Although
ANSES interpreted direct genotoxicity via particle
accumulation in the nucleus to be possible, RAC
instead referred to studies, for example, research
by EFSA, and weighed oxidative stress to be the
likely mode of action instead of direct genotox-
icity. The RAC concluded that the mode of action
in rats may be different from humans and that the
human evidence is not conclusive enough to prove
that rat carcinogenicity does not correlate with
human carcinogenicity.
3) The RAC concluded the classification as Carc 2: Sus-
pected human carcinogen (and not Carc 1B) based on
TiO2 showing no other than respiratory effects, no con-
clusive data on species other than rats (and only in fe-
males), indications that other rodents and humans are
less sensitive to TiO2, and the consistent evidence of
epidemiological studies that do not support a correlation
between TiO2 exposure and lung cancer. However, TiO2
was found to be a rat carcinogen in 2 independent
studies and RAC did not consider there to be enough
evidence in human data to exclude the possibility of the
rat‐like effect in humans (ECHA 2017).
Impacts of classification of TiO2 as Carc 2
Raw material. In the production of TiO2 and its inclusion in
various products, the occupational safety standards are
defined for the individual substances, and they do not have
a direct link to the hazard classification.
Purchasing of raw materials is strongly tied to the com-
panies’ own ecodesign guidance, which may contain spe-
cific lists, for example, Volvo’s black, gray, and white lists
(Volvo 2018), or more generic advice to avoid hazardous
Integr Environ Assess Manag 2019:1–12 DOI: 10.1002/ieam.4186
7
substances. The latter is likely to include carcinogens. For
instance, EU guidance for green public procurement (EC
2018) provides detailed instructions for specifying green
products. These criteria for paint, textile, paper, and furni-
ture products include prohibition of substances with any
cancer classification (EC 2018).
Product. Carc 2 classification has a direct impact on the
product information. Mixtures containing at least 1% of TiO2
become equally classified (as a suspected human carci-
nogen). As a consequence of the classification, products
that fall into the regime of the CLP regulation, such as
paints, will have a warning label attached (Figure 1) and will
include statement H351: Suspected of causing cancer via
inhalation (CLP, Annex 1, Table 3.6.3). Food, cosmetic, and
pharmaceutical products are exempted from the CLP reg-
ulation because they have their own safety regulations and
safety assessment requirements (EC 2008a). Product‐
specific restrictions for toys ban the use of any class of car-
cinogenic substances (EC 2009b). Other articles, which are
in the scope of REACH or other product‐specific regu-
lations, such as The Restriction of the use of certain Haz-
ardous Substances in electrical and electronic equipment
(RoHS; EC 2011), do not have a direct link to CLP classi-
fications (Molander and Rudén 2011).
Commonly applied ecodesign guidance for product: eco-
labels, ban substances with a classification as carcinogenic of
any category, for example, the EU ecolabel criteria for textiles,
footwear, paints and varnishes, and furniture (EU 2018).
Waste. Mixtures that contain ≥1% of substances with Carc 2
classification become hazardous waste (EC 2008c). This waste
is to be separated from waste streams (EC 2008c, Point 28)
and labeled accordingly by industry (EC 2008c, Point 34).
Moreover, the producers and the other actors dealing with
hazardous waste must keep records of quantities, nature of
origin, and other such information related to hazardous waste.
Further, the Waste Framework Directive sets a ban on all kinds
of mixing of hazardous waste with other material (EC 2008c,
Article 18). This ban will have an impact on the recyclability of
the products containing TiO2 ≥ 1%. Titanium dioxide is
present in varying contents in wastes, such as printed material
and painted building waste. Unlike the Waste Framework Di-
rective (EC 2008c), the Water Framework Directive (EC 2000),
and the Sewage Sludge Directive (EEC 1986) do not refer to
substance classifications (Molander et al. 2012).
Reduced TiO2 exposure?
This section considers changes in exposures (respiratory,
oral, and dermal) to TiO2 due to the new classification. Ex-
posures are related to different uses in the present situation
and in the case in which Carc 2 classification is applied. The
section adopts the same approach as in the previous section
Impacts of classification of TiO2 as Carc 2 and proceeds
from exposure to TiO2 as a raw material to TiO2 in different
products and, finally, TiO2 exposure in waste treatment and
recycling.
© 2019 SETAC
8 Integr Environ Assess Manag 000, 2019—EE Kähkönen

The new classification does not directly reduce respiratory
exposure to TiO2 in raw material production and in other
production processes in which TiO2 is used. This ob-
servation results from the fact that occupational regulations
are not tied to CLP classifications (IARC 2010). Similarly, oral
and dermal exposure to TiO2 in food, pharmaceuticals, and
cosmetics will not change due to the new classification,
given that the regulations for these products do not refer to
CLP (EFSA 2016). Instead, the TiO2 classification will reduce
dermal and oral exposure from toys, ecolabeled products,
and recycled materials (EC 2009a, 2008c; EU 2018). How-
ever, none of these exposures (especially when related to
solid artifacts) were considered hazardous (ECHA 2017). In
the case of TiO2‐containing paints and other products, the
warning labels will be attached to products. Here, too, this
kind of respiratory (and dermal) exposure to liquid paint was
not considered hazardous (Boffetta 2001). In waste treat-
ment, various kinds of exposures take place. Control of
these exposures again under occupational regulations, re-
gardless of the waste type (normal waste or hazardous
waste).
In conclusion, the direct reduction in the exposure due to
the new classification will be achieved in dermal and oral
exposure in toys, recycled materials, and ecolabeled prod-
ucts. In regard to products for which the “Health hazard”
warning label is required, the impact on exposure is am-
biguous. In none of these cases was the hazard associated
with the TiO2 exposure considered significant nor was it the
original cause for the classification. The regulatory con-
sequences are summarized in Table 2.
DISCUSSION AND CONCLUSIONS
Impacts of Carc 2 classification in hazardous exposure
According to the present analysis, the proposed Carc 2
classification for TiO2 will not directly reduce exposure re-
lated to health hazards, which was the main concern and the
origin of the classification proposal. The original cause for
the classification proposal stemmed from possible carcino-
genicity associated with respiratory exposure to pure TiO2
particles. Instead, the direct consequences impact TiO2 in
mixtures and solid artifacts in which the hazard was not
claimed (Boffetta 2001). Evidently, we may speculate on
how much the classification will reduce the total production
of TiO2 and, hence, lead to reduced exposure in production.
However, due to the high price of TiO2, there has been a
long‐standing and continual attempt to reduce its use. The
attempts at price reduction are mainly discussed in trade
magazines, for example, those of the coating industry
(Bussell and Verheijen 2011) but are also referred to in sci-
entific articles (Narayan and Kothapalli 2000; Ruzala et al.
2015). Despite all the effort, TiO2 still accounts, by far, for
the largest (and growing) global pigment volume (Jovanović
2014). Consequently, there are no easy replacements for
TiO2 or an evident route toward reducing TiO2 production
and consequently reducing exposure in this way.

Table 2. Expected changes in TiO2 hazardous exposure due to classification as suspected carcinogen

Exposure No classification Carc 2

Raw material

Respiratory exposure: particles in production Exposure restricted by Exposure restricted by

occupational regulationsa occupational regulationsa

Product

Oral exposure: Additive in food, medical, and cosmetics Use controlled by food, medical, Use controlled by food, medical,
products and cosmetics regulationsb and cosmetics regulationsb

Dermal exposure: Additive or pigment in cosmetics, Use controlled by food, medical, Use controlled by food, medical,
medical, and food products, and pigment in toys, and cosmetics regulationsb and cosmetics regulationsb
paper, paint, ecolabel, and recycled materials
Use banned in toysc, ecolabel
productsd, and recycled
materialse

Respiratory exposure: Additive or pigment in cosmetics Use controlled by cosmetics Use controlled by cosmetics
and paint spray regulationsb regulationsb

“Health hazard” warning label

attached in productsc

Waste

Dermal, respiratory: Additive, pigment, and residual in Exposure controlled by Exposure controlled by
waste treatment and recycling occupational regulationsa occupational regulationsa
a
IARC 2010.
b
EFSA 2016.
c
EC 2009b.
d
EU 2018.
e
EC 2008c.

Integr Environ Assess Manag 2019:1–12 wileyonlinelibrary.com/journal/ieam © 2019 SETAC

A Case Study on TiO2 Classification—Integr Environ Assess Manag 000, 2019
Impacts of Carc 2 classification in waste treatment and
recycling
Although not directly reducing the health hazard asso-
ciated with TiO2 exposure, the classification will directly
reduce the recycling of materials that contain or may contain
TiO2. For instance, paper is currently recycled regardless of
the printing ink pigmentation used in it. Instead of recycling,
the material containing TiO2 will end as hazardous waste.
The obstacles in recycling are clearly in conflict with material
efficiency and circular economy requirements. Here, TiO2
showcases a clash of regulations similar to that of bromi-
nated fire retardants in plastics recycling. Plastics that con-
tain brominated fire retardants, for example, are not to be
recycled in order to avoid exposure to hazardous sub-
stances. There are no evident solutions to the dilemma of
balancing the goals of waste reduction and a circular
economy and the goal of reducing hazardous substances in
society (Janssen et al. 2016). What is specific to the case of
TiO2 is that the circulated solid (e.g., printed material) or
liquid (e.g., paint), that contain TiO2, will not lead to the
exposure that was suspected for the carcinogenicity, that is,
the respiratory exposure to TiO2 particles. Consequently, in
the case of TiO2 the goals of a circular economy are missed
without reducing the prevalence of cancer.
Classification of mixtures with PSLT particles
The TiO2 classification challenges CLP in regard to
dealing with the hazards associated with liquid mixtures that
contain PSLT particles. The special focus here is on nano-
particles, which are undergoing extensive research today.
Strikingly, the volume of the research in the field of nano-
toxicology has reached the point at which it has become
unmanageable by regulators. Drobne (2018) highlights the
importance of dialogue among all the stakeholders. Spe-
cifically, scientists in the field were entitled to provide
comprehensive, transparent, reproducible, and updateable
summaries regarding hazards associated with TiO2.
In practice, respiratory exposure of PSLT particles may
lead to hazard findings, which (as in the present case) lead to
a classification requirement for a liquid mixture containing
the particle. However, in a liquid mixture, for example, in
paint, the possibility for respiratory exposure associated with
health hazards remains minimal according to Boffetta et al.
(2001). Accordingly, it appears questionable whether the
severe “Health hazard” warning regarding respiratory ex-
posure is in its appropriate place when it strongly warns
against a hazard related to the respiratory exposure to dry
particles when using a liquid product, such as the afore-
mentioned paint. A similar debate will undoubtedly take
place on the ANSES proposal for another commonly used
PSLT substance; carbon black will be in a similar classification
process (ECHA 2016b). Hence, it is necessary to lay down
rules for the classification of PSLT substances in mixtures to
avoid warnings and restrictions in those applications in which
they are misleading. This recommendation is in accordance
with Drobne (2018), who recommends further dialogue
Integr Environ Assess Manag 2019:1–12 DOI: 10.1002/ieam.4186
9
concerning nanotoxicology before making the final decision
on classification of TiO2. More particularly, Drobne (2018)
emphasized the role of scientists in regulatory decision
making. In practice, focused scoping papers were mentioned
as one way in which scientists may support the process.
Differing opinions during classification process
Regulatory processes include dialogue among authorities,
industry, and nongovernmental organizations (NGOs), and
differing perspectives are a basic element in these dis-
cussions. The debated topic varies from case to case. For
instance, in the case of glyphosate, the weighing of evi-
dence from animal studies resulted in different conclusions
by RAC and IARC. Here, IARC proposed Carc 2 classi-
fication, whereas RAC concluded there was no adequate
evidence for carcinogenicity (ECHA 2016a). Another recent
debate concerns the restriction on bisphenol A (BPA) use in
thermal paper. Here, the main emphasis in the recent de-
bates focused on the socioeconomic consequences and
(non)availability of substitutes (ECHA 2015b; Pivnenko et al.
2018), whereas in the case of the identification of bis(pen-
tabromophenyl) ether (DecaBDE) as a substance of very high
concern (SVHC) the focus is on the test methods for bio-
degradability and bioaccumulation and on the definitions
for persistent, bioaccumulative, toxic (PBT) and very persis-
tent, very bioaccumulative substances (vPvB) (ECHA 2012).
In the present case of TiO2 classification, ANSES and RAC
arrived at different conclusions on the basis of genotoxicity
assessments. The challenge in interpretation of in vitro
genotoxicity results was recently elaborated by Charles et al.
(2018). The study sought an explanation of why 36 studies
did not lead to a clear picture on TiO2 genotoxicity. The
study suggested further studies on TiO2 genotoxicity but
also further development of in vitro assessment. The as-
sessment methods in the study included comet assays, ox-
idative stress assays, and micronucleus assays. Regardless of
the method used, the researchers were advised to take into
account possible interference between the studied nano-
particles and an assay when choosing a method or a com-
bination of methods. The recommendations for developing
nanoparticle testing included detailed instructions, such as
taking the UV light into account, using different durations,
and paying attention to the choice of cell line. Furthermore,
general instruction to develop mechanistic studies for as-
sessing the genotoxic mode of action was given.
In the present study, it became clear where the views of
the authorities and industry differ in the hazard assessment
of TiO2, that is, weighing the results from human exposure
studies and the results from animal studies. Neither of the
strategies offers bulletproof results. On one hand, the set-
ting in human exposure studies was not perfect, but com-
prised variables, such as duration and level of exposure, and
as exposure to other substances (e.g., tobacco smoke). On
the other hand, the correlation between animal tests with
the short exposure time and high quantities to longtime
exposure to low quantities in humans is not straightforward
either. These debates are a rule rather than an exception in
© 2019 SETAC
10
the regime of chemical regulation. However, in the case of
TiO2 the extensiveness of the human exposure studies
(more than 20 000 humans) and the consensus on the out-
comes (not showing carcinogenicity) highlights the differ-
ence in weighing evidence. One aspect here is the second
aim of REACH: “the promotion of alternative methods for
assessment of hazards of substances.” The case of TiO2
shows that the traditional animal tests are likely to outweigh
other evidence. Notably, this resolution complies with the
carcinogenicity classification rules that are laid down in CLP
(EC 2008a). Consequently, the animal tests sustain their
status as the gold standard of hazard assessment despite
argumentation against their reliability. In an esteemed
study, the interspecies correlation between rats, mice, and
rabbit is 40% to 63%, that is, in the range of tossing a coin,
and there is no reason to expect higher correlation between
rodents and humans (Hartung 2009). However the in vitro
methods are not currently feasible direct substitutes for
animal tests, especially in higher tier testing such as carci-
nogenicity testing (ECHA 2017b). The challenges are in the
development of new methods and in standardization itself,
and there are major challenges in achieving applicability of
methods alternative to animal tests in regulatory decision
making (OECD 2019). Altogether, the push toward new
toxicological methods is also accepted by regulators, and
there is a strong aim at promoting alternative, nonanimal
test methods. For instance, the Joint Research Centre of the
European Commission has provided a state‐of‐the‐art re-
view of alternative methods to ECHA (Worth et al. 2014).
Similarly, ECHA provides guidance on using nonstandard
test methods (Brautigam 2016). However, none of the
guidance specifies criteria for the reliability of exposure or
epidemiological studies in the human population. The
human exposure studies are conducted, and the information
from them should be utilized when available, according to
REACH (Annex 1). The present case study delivers a dis-
couraging message on the utility of these studies. Con-
sequently, the author recommends the production of “state
of the art” instruction and criteria on how to use exposure
studies in a regulatory context.
Gaps in the regulatory network
The present study clearly visualizes connections and gaps
between the different instruments in the EU chemical man-
agement. The basic division may be drawn between the
regulations that refer to CLP classification, namely, Toy
Safety Directive (EC 2009a), REACH (restrictions) (EC 2006),
Waste Framework Directive (EC 2008c), and those that do
not have a link to CLP, namely, RoHS (EC 2011), Water
Framework Directive (EC 2000), Sewage Sludge Directive
(EEC 1986), occupational regulations (EC 2017), cosmetics
regulations (EC 2009b), food regulations (EC 2002), and
pharmaceutical regulations (Molander and Rudén 2011;
Molander et al. 2012; Sobek et al. 2013). This kind of sit-
uation calls for further development. The goals of the de-
velopment should be in getting the most out of the
resources invested in the REACH processes (Molander and
Integr Environ Assess Manag 2019:1–12 wileyonlinelibrary.com/journal/ieam
Integr Environ Assess Manag 000, 2019—EE Kähkönen
Rudén 2011), avoiding illogical treatment of substances in
different products (Sobek et al. 2013), and avoiding poten-
tial unintended outcomes of combined regulations, such as
reducing the application of regulatory risk where risk did not
exist and, simultaneously, not reaching the area where the
risk was suspected, as indicated in the present study.
Hazard communication attached with Carc 2 classification
One interesting question rising from the TiO2 case is re-
lated to Carc 2 classification itself. Although Carc 1 classi-
fication causes a clear‐cut ban on the use of a substance
with this classification in consumer mixtures, these mixtures
may contain Carc 2–classified substances. As discussed, the
mixtures will obtain the most severe “Health hazard” warning
labels. Even as the label indicates a severe warning, the true
meaning of the “Health hazard” label is more perplexing.
What is the intended guidance for consumers? Is this
product not safe to use due to carcinogenicity? Then, it is
more logical to ban the product. Is it safe to use this product
despite the suspected carcinogenicity? Then, the warning
label does not support this message. Are you to use this on
your own responsibility with caution? In this case, in order to
be able to make informed choices, the consumer should
have access to information that states the findings and un-
certainties related to the suspected carcinogenicity. Ac-
cordingly, ECHA puts effort into the dissemination of the
substance information in the “Information on Chemicals”
database (ECHA 2019b). However, in the current form, the
strictly regulated discussion is not easily accessible even to
those who want to know more. For instance, the recent
study by Ingre‐Khan (2018) concluded that scrutiny of the
substance information in ECHA’s database remains chal-
lenging to the public audience. The requirement of acces-
sible communication is a challenge for the scientific
community, and there are no quick fixes to it. However,
ECHA and other scientific communities must put effort into
producing understandable messages for varying audiences.
The EHCA is the key actor here and should further develop
the readability of substance information in its database.
In summary, as a response to research question 1, TiO2
classification as Carc 2 will not directly reduce the assumed
and debated hazard related to respiratory exposure. The
sources of the ambiguity during the process (research
question 2) are pointed to weighing the evidence of human
and animal studies and to contradicting interpretations of
genotoxicity studies. The present study revealed an unin-
tended outcome of the classification process: the negative
impact on material recycling without reduction of hazardous
substances in society. Moreover, the case visualized more
general topics of interest: treatment of PSLT particles in
CLP, lack of guidance for using human exposure data, and
lack of effort in communicating chemical hazards to non-
professionals. The author’s recommendations are as follows:
1) The final classification of TiO2 classification should be
postponed and dealt with together with other PSLT
particles. A common guideline is needed for assessing
© 2019 SETAC
A Case Study on TiO2 Classification—Integr Environ Assess Manag 000, 2019
and communicating hazards associated with dry particles
in liquid mixtures. Here, the outcomes of nanotoxicology
development should be taken into account as advised by
Drobne (2018).
2) Guidance on using human exposure data should be
compiled to encourage the adaptation and utilization of
the existing data.
Acknowledgment—Jaana Suviniitty was helpful with the
language‐related matters of this article.
Data Availability Statement—Data were retrieved from
publicly available web pages (authorities, NGOs) and from
scientific articles.
REFERENCES
Alépée N, Grandidier M‐H, Cotovio J. 2017. Assessment of the Human Epi-
dermis SkinEthic™ RHE model for in vitro skin corrosion testing of
chemicals. In: Eskens C, van Vliet E, Maibach HI, editors. Alternatives for
dermal toxicity testing. Cham (CH): Springer. p 143–157.
[ANSES] French Agency for Food, Environmental and Occupational Health
and Safety. 2016. CHL report – Proposal for harmonised classification and
labelling based on Regulation No 1272/2008 (CLP Regulation), Annex VI,
Part 2‐ Substance name: Titanium dioxide. Maisons‐Alfort (FR). 159 p.
Bodar C, Spijker J, Lijzen J, Waaijers‐van der Loop S, Luit R, Heugens E,
Janssen M, Wassenaar P, Traas T. 2018. Risk management of hazardous
substances in a circular economy. J Environ Manage 212:108–114.
Boffetta P, Gaborieau V, Nadon L, Parent MF, Weiderpass E, Siemiatycki J.
2001. Exposure to titanium dioxide and risk of lung cancer in a pop-
ulation‐based study from Montreal. Scand J Work Env Hea 27:227–232.
Brautigam T. 2016. Alternatives to animal testing—What's new in 2016?
ECHA Newsletter 1. [accessed 2019 Sep 6]. https://newsletter.echa.
europa.eu/home/‐/newsletter/entry/1_16_alternatives‐to‐animal‐testing‐
whats‐new‐in‐2016
Bussell S, Verheijen M. 2011. Cutting the cost of opacity. Eur Coat J 1(1):114–117.
Cappelli CI, Benfenati E, Cester J. 2015. Evaluation of QSAR models for
predicting the partition coefficient (log P) of chemicals under the REACH
regulation. Environ Res 143(A): 26–32.
[CEPE] European Council of the Paint, Printing Ink and Artists’ Colours Industry.
2017. Paint, coatings and printing inks remain safe to use. Brussels (BE). 1 p.
[accessed 2018 Sep 3]. http://www.cepe.org/wp‐content/uploads/2017/10/
01‐TiO2‐Media‐statement‐for‐customers‐170609_Final.pdf
Charles S, Jomini S, Fessard V, Bigorgne‐Vizade E, Rousselle C, Michel C.
2018. Assessment of the in vitro genotoxicity of TiO2 nanoparticles in a
regulatory context. Nanotoxicology 12(4):357–374. https://doi.org/10.
1080/17435390.2018.1451567
Drobne D. 2018. Spotlighting CLH report for TiO2: Nano‐safety perspective.
Chem Eng J 340(15):192–195. https://doi.org/10.1016/j.cej.2018.01.007
[EC] European Commission. 2000. Directive 2000/60/EC of the European
Parliament and of the Council of 23 October 2000 establishing a frame-
work for Community action in the field of water policy. Official Journal of
the European Union 327:1–72.
[EC] European Commission. 2002. Regulation No 178/2002 of the European
Parliament and of the Council of 28 January 2002 laying down the general
principles and requirements of food law, establishing the European Food
Safety Authority and laying down procedures in matters of food safety.
Official Journal of the European Communities 31:1–24.
[EC] European Commission. 2006. Regulation No 1907/2006 of the European
Parliament and of the Council of 18 December 2006 concerning the Reg-
istration, Evaluation, Authorisation and Restriction of Chemicals (REACH).
Consolidated version. Helsinki (FI): ECHA. 512 p.
[EC] European Commission. 2008a. Regulation No 1272/2008 of the Euro-
pean Parliament and of the Council of 16 December 2008 on Classi-
fication, Labelling and Packaging of substances. Consolidated version.
Helsinki (FI): ECHA. 1355 p.
Integr Environ Assess Manag 2019:1–12 DOI: 10.1002/ieam.4186
11
[EC] European Commission. 2008b. Commission Regulation No 440/2008 of
30 May 2008 laying down test methods pursuant to Regulation (EC) No
1907/2006 of the European Parliament and of the Council on the Regis-
tration, Evaluation, Authorisation and Restriction of Chemicals (REACH).
Official Journal of European Union L142:1–739.
[EC] European Commission. 2008c. Directive 2008/98/EC of the European
Parliament and of the Council of 19 November 2008 on waste and re-
pealing certain directives. Official Journal of the European Union
L312:3–30.
[EC] European Commission. 2009a. Directive 2009/48/EC of the European
Parliament and of the Council of 18 June 2009 on the safety of toys.
Official Journal of the European Union L170:1–37.
[EC] European Commission. 2009b. Regulation No 1223/2009 of the
European Parliament and of the Council of 30 November 2009 on
cosmetic products. Official Journal of the European Communities
L342:59–209.
[EC] European Commission. 2011. Directive 2011/65/EU of the European
Parliament and of the Council of 8 June 2011 on the restriction of the use
of certain hazardous substances in electrical and electronic equipment.
Official Journal of the European Union 174:88–110.
[EC] European Commission. 2017. Commission staff working document ex‐
post evaluation of the European Union occupatinal safety and health di-
rectives (refit evaluation) accompanying the document communication
from the commission to the European Parliament, the Council, the Euro-
pean Economic and Social Committee and the Committee of the regions
safer and healthier work for all—Modernisation of the EU Occupational
Safety and Health Legislation and Policy. Brussels (BE). 269 p.
[EC] European Commission. 2018. EU GPP [Green Public Procurement] cri-
teria. Luxemburg (LU). 80 p. [accessed 2018 Jul 4]. http://ec.europa.eu/
environment/gpp/eu_gpp_criteria_en.htm
[ECHA] European Chemicals Agency. 2012. Comments on an Annex Xv
Dossier for identification of a substance as SVHC and responses to these
comments. Helsinki (FI). 10 p. [accessed 2019 Apr 11]. https://echa.
europa.eu/management‐of‐pbt‐vpvb‐substances
[ECHA] European Chemicals Agency. 2015a. Update of the workplan on
nanomaterials. Luxemburg (LU). 2 p. [accessed 2019 Feb 2]. https://echa.
europa.eu/documents/10162/21844190/mb_41_2015_workplan_nanoma
terials_incl_annexes_en.pdf/20da85dc‐6ca0‐46f1‐bfd5‐50b92d015c2a
[ECHA] European Chemicals Agency. 2015b. Response to comments on the
SEAC draft opinion on the Annex XV dossier proposing restriction on
Bisphenol A; 4,4'‐isopropylidenediphenol. Helsinki (FI). 28 p.
[ECHA] European Chemicals Agency. 2016a. CLH report: Proposal for
harmonised classification and labelling: Based on Regulation (EC) No
1272/2008 (CLP Regulation), Annex VI, Part 2, Substance Name: N‐
(phosphonomethyl)glycine; Glyphosate (ISO). Helsinki (FI). 169 p.
[ECHA] European Chemicals Agency. 2016b. Justification document for the
selection of a CoRAP substance [carbon black]. 13 p. [accessed 2019 Sep
9]. https://echa.europa.eu/documents/10162/6d692fdb‐df2a‐48f2‐8d74‐
539bddd49c53
[ECHA] European Chemicals Agency. 2017a. Committee for Risk Assessment,
RAC opinion proposing harmonised classification and labelling at EU
level of titanium dioxide. Helsinki (FI). 52 p.
[ECHA] European Chemicals Agency. 2017b. The use of alternatives to
testing on animals for the REACH regulation. Helsinki (FI). 103 p.
[ECHA] European Chemicals Agency. 2018. Harmonized classification and
labelling. Helsinki (FI). [accessed 2018 Jul 4]. https://echa.europa.eu/
regulations/clp/harmonised‐classification‐and‐labelling
[ECHA] European Chemicals Agency. 2019a. Introductory Guidance on the CLP
Regulation. Helsinki (FI). 93 p. [accessed 2019 Sep 6]. https://echa.europa.
eu/documents/10162/23036412/clp_introductory_en.pdf/b65a97b4‐8ef7‐
4599‐b122‐7575f6956027
[ECHA] European Chemicals Agency. 2019b. Summary of classification and
labelling (TiO2). Helsinki (FI). [accessed 2019 Feb 5]. https://echa.
europa.eu/information‐on‐chemicals/cl‐inventory‐database/‐/discli/
details/100661
[EEC] European Economic Community. 1986. Council Directive 86/278/EEC
of 12 June 1986 on the protection of the environment, and in particular of
© 2019 SETAC
12
the soil, when sewage sludge is used in agriculture. Official Journal of the
European Communities 181:6–12.
[EFSA] European Food Safety Authority. 2016. Re‐evaluation of titanium di-
oxide (E 171) as a food additive ‐ EFSA Panel on Food Additives and
Nutrient Sources added to Food (ANS). EFSA J 14:1–83.
[EU] European Union. 2018. The EU Ecolabel product catalogue ‐ Criterias for
your product group. [accessed 2018 Sep 15]. http://ec.europa.eu/ecat/
Fryzek JP, Chadda B, Marano D et al. 2003. A cohort mortality study among
titanium dioxide manufacturing workers in the United States. J Occup
Environ Med 45:400–409.
Geueke B, Muncke J. 2017. Substances of very high concern in food contact
materials: Migration and regulatory background. Packag Technol Sci
31(3).
Hartung T. 2009. Toxicology for the twenty‐first century. Nature 460:208–212.
Hassas BV, Karakaş F, Celik MS. 2015. Effect of precipitated calcium carbonate
additions on waterborne paints at different pigment volume concen-
trations. Prog Org Coat 83:64–70.
[IARC] International Agency for Research on Cancer. 2010. Carbon black, ti-
tanium dioxide, and talc. IARC Monog Eval Carc 93:193–276.
Ingre‐Khan E. 2018. Transparency within REACH? Regulatory risk assessment
of industrial chemicals [Doctoral dissertation]. Stockholm (SE): Stockholm
University Universitetsservice. 58 p.
Jovanović B. 2014. Critical review of public health regulations of titanium di-
oxide, a human food additive. Integr Environ Assess Manag 11(1):10–20.
Molander L, Breitholtz M, Andersson PL, Rybacka A. 2012. Are chemicals in
articles an obstacle for reaching environmental goals? Missing links in EU
chemical management. Sci Total Environ 435–436:280–289.
Molander L, Rudén C. 2011. Narrow‐and‐sharp or broad‐and‐blunt ‐ Regu-
lations of hazardous chemicals in consumer products in the European
Union. Regul Toxicol Pharmacol 62(3):523–531.
Narayan R, Kothapalli RV. 2000. Use of calcined clay as part replacement of
titanium dioxide in latex paint formulations. J Appl Polym Sci 77(5):
1029–1036.
[OECD] Organisation for Economic Co‐operation and Development. 2012.
Guidance Document 116 on the conduct and design of chronic toxicity
Integr Environ Assess Manag 2019:1–12 wileyonlinelibrary.com/journal/ieam
Integr Environ Assess Manag 000, 2019—EE Kähkönen
and carcinogenicity studies, supporting Test Guidelines 451, 452 and
453. 2nd ed. Series on Testing and Assessment, No. 116. Paris (FR).
[OECD] Organisation for Economic Co‐operation and Development. 2019.
Integrated approaches to testing and assessment (IATA). Paris (FR). [ac-
cessed 2019 Apr 4]. http://www.oecd.org/chemicalsafety/risk‐assessment/
iata‐integrated‐approaches‐to‐testing‐and‐assessment.htm
Pivnenko K, Lanerb D, Astrupa TF. 2018. Dynamics of bisphenol A (BPA) and
bisphenol S (BPS) in the European paper cycle: Need for concern? Resour
Conserv Recycl 133:278–287.
Riley D. 2014. Mental models in warnings message design: A review and two
case studies. Safety Sci 61:11–20.
Ruszala MJA, Rowson NA, Grover LM, Choudhery RA. 2015. Low carbon
footprint TiO2 substitutes in paint: A review. Int J Chem Eng Appl 6(5):
331–340.
Sobanska M, Scholz S, Nyman A‐M, Cesnaitis R, Alonso SG, Klüver N, Kühne
R, Tyle H, de Knecht J, Dang Z et al. 2017. Applicability of the fish embryo
acute toxicity (FET) test (OECD 236) in the regulatory context of Regis-
tration, Evaluation, Authorisation, and Restriction of Chemicals (REACH).
Environ Toxicol Chem 37(3):657–670.
Sobek A, Bejgarn S, Rudén C, Molander L. 2013. In the shadow of the Cos-
metic Directive ‐ Inconsistencies in EU environmental hazard classification
requirements for UV‐filters. Sci Total Environ 461–462C:706–711.
Steensgaard IM, Syberg K, Rista S, Hartmann NB, Boldrina A, Hansen SF.
2017. From macro‐ to microplastics ‐ Analysis of EU regulation along the
life cycle of plastic bags. Environ Pollut 224:289–299.
[TDMA] Titan Dioxide Manufacturer’s Association. 2018. Titanium dioxide is
safe. [accessed 2018 Jul 6]. https://tdma.info/titanium‐dioxide‐is‐safe/
Volvo. 2018. Chemical substances which should not be present in processes or
products within the Volvo Group Volvo’s grey list. 16 p. [accessed 2018
Jul 6]. https://webstd.volvo.com/webstd/docs/100‐0003
Worth A, Barroso J, Bremer S, Burton J, Casati S, Coecke S, Corvi R, Desprez
B, Dumont C, Gouliarmou V et al. 2014. Alternative methods for regulatory
toxicology—A state‐of‐the‐art review. JRC Science and policy reports.
RC91361. Luxembourg (LU): Publications Office of the European Union.
470 p.
© 2019 SETAC