Designation: F3159 − 15´1

Standard Safety Specification for

Liquid Laundry Packets1
This standard is issued under the fixed designation F3159; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (´) indicates an editorial change since the last revision or reapproval.

ε1 NOTE—Editorially corrected 5.3.3.2, 6.1, and 6.2 in December 2015.

INTRODUCTION

In November 2012 the U.S. Consumer Product Safety Commission (CPSC) issued a Safety Alert
to inform parents and caregivers that Liquid Laundry Detergent Packets need to be kept away from
children as those who are exposed to packet contents are at risk of serious injury and even death due
to the highly concentrated nature of the product. Children who have accidently ingested Liquid
Laundry Detergent Packets have experienced a range of injuries including loss of consciousness,
respiratory distress, vomiting, coughing, choking and drowsiness, and in cases where there has been
contact with the eyes, painful irritation of the eyes and corneal burns have occurred. In addition, death
has been reported to occur following ingestion of Liquid Laundry Detergent Packets, including in one
child.

1. Scope 2. Referenced Documents

1.1 This specification provides requirements for household 2.1 ASTM Standards:2
Liquid Laundry Detergent Packet safety to help reduce unin- D3475 Classification of Child-Resistant Packages
tentional exposures to the contents of the packets, especially to D4332 Practice for Conditioning Containers, Packages, or
children. Packaging Components for Testing
D4359 Test Method for Determining Whether a Material Is
1.2 This standard applies exclusively to household Liquid
a Liquid or a Solid
Laundry Detergent Packets. “Liquid Laundry Detergent Pack-
2.2 ANSI Standard:3
ets” are single-use laundry detergent products that contain a
ANSI Z535.4 Safety Color Code—Environmental Facility
liquid detergent enclosed in a water soluble outer layer (“pouch
Safety Signs—Criteria for Safety Symbols—Product
film”). This includes laundry detergent packets in soluble film
Safety Sign and Labels and Accident Prevention Tags
that contain liquid only (that is, all liquid), as well as those that
contain both liquid and non-liquid components. 3. Terminology
1.3 This standard does not purport to address all of the 3.1 Definitions:
safety concerns, if any, associated with Liquid Laundry Deter- 3.1.1 liquid, n—a substance or mixture which: (1) at 50°C
gent Packet use. It is the responsibility of the user of this has a vapour pressure of not more than 300 kPa (3 bar), (2)
standard to establish appropriate safety and health practices which is not completely gaseous at 20°C and at a standard
and determine the applicability of regulatory limitations prior pressure of 101.3 kPa, and (3) which has a melting point or
to use. It is the responsibility of the user of the product to follow initial melting point of 20°C or less at a standard pressure of
the warning statements and use the product appropriately. 101.3 kPa.

2
For referenced ASTM standards, visit the ASTM website, www.astm.org, or
1
This specification is under the jurisdiction of ASTM Committee F15 on contact ASTM Customer Service at [email protected]. For Annual Book of ASTM
Consumer Products and is the direct responsibility of Subcommittee F15.71 on Standards volume information, refer to the standard’s Document Summary page on
Liquid Laundry Packets. the ASTM website.
3
Current edition approved Sept. 15, 2015. Published October 2015. DOI: Available from American National Standards Institute (ANSI), 25 W. 43rd St.,
10.1520/F3159-15E01. 4th Floor, New York, NY 10036, http://www.ansi.org.

Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States

1
 F3159 − 15´1
3.1.1.1 Discussion—A viscous substance or mixture for 5.2.1 Comply with this standard through the full life cycle
which a specific melting point cannot be determined shall be of product package.
subjected to the Test Method D4359-90 test; or to the test for 5.2.2 The package must meet the option standard for which
determining fluidity (penetrometer test) prescribed in section the package is designed while also accounting for any other
2.3.4 of Annex A of the European Agreement concerning the way the package could conceivably be opened for example by
International Carriage of Dangerous Goods by Road (ADR). twisting, pulling, or use of singular force.
3.1.2 liquid laundry detergent packets, n—individual pack- 5.3 Options for Packaging:
ets that contain liquid laundry detergent and are intended to
5.3.1 A package that meets the performance requirements of
dissolve when used as intended.
16 CFR Part 1700, section 1700.15 and testing requirements of
3.1.3 pouch film, n—the water-soluble outer layer of a 16 CFR Part 1700, section 1700.20.
Liquid Laundry Detergent Packet that contains laundry deter- 5.3.2 An individually-wrapped package that contains no
gent or other liquid ingredients, or both, and is designed to more than one packet and incorporates either:
dissolve when used as intended.
5.3.2.1 A hidden tab or notch or other means of opening that
4. Liquid Laundry Detergent Packet Requirements is only exposed after the package has been folded or manipu-
lated in an instructed manner, or
4.1 The Liquid Laundry Detergent Packet must meet the
5.3.2.2 A feature described in Classification D3475-14,
requirements set forth in European Commission Regulation
Type IV Non-reclosable packaging-flexible and Type V Unit
(EU) No 1272/2008, Annex II, Part 3, Section 3.3.3, and
non-reclosable packaging-rigid.
Sections 4 (Aversive Agent in the Soluble Film) and 5 (Capsule
Integrity) of the accompanying AISE Liquid Laundry Deter- 5.3.3 A package that requires manipulative skill or dexterity
gent Capsules Guidelines on CLP Implementation4, including to open, including, but not limited to:
as may be amended. For reference, EU No 1272/2008, Annex 5.3.3.1 A package with two or more closure mechanisms
II, Part 3, Section 3.3.3, is attached as Annex A5, and Sections that are interdependent, so that the package cannot be fully
4 and 5 of the AISE Liquid Laundry Detergent Capsules opened without releasing at least two of the closure mecha-
Guidelines on CLP Implementation are attached as Annex A7. nisms.
4.2 For the avoidance of doubt, European Commission 5.3.3.2 A double-action release mechanism, defined as ei-
Regulation (EU) No 1272/2008, Annex II, Part 3, Section 3.3.3 ther:
(i) does not specify any particular manner in which the soluble (1) A mechanism requiring two consecutive motions, the
packaging containing the agent must have the aversive agent first of which must be maintained (and which may include the
added to the soluble packaging. A company may choose, by act of physically holding or stabilizing the package) while the
way of non-limiting examples, to introduce the aversive agent second is carried out in order to fully open the package, or
to the soluble packaging by admixing it into a slurry that is (2) Two separate and independent motions that must be
subsequently formed or cast into a film and/or by coating it activated or occur simultaneously to fully open the package.
onto a previously formed film. It is up to each company to NOTE 1—A simple zipper closure (pull to open or simple slider) would
select the aversive agent and technology they deem appropriate not meet the requirements of either 5.3.3.2(1) or (2).
for their products and effective for meeting the criteria of being 5.3.3.3 A release mechanism or system of mechanisms
safe and eliciting oral repulsive behavior within a maximum which requires two independent release mechanisms to be
time of 6 s as provided by and in accordance with the EU performed consecutively in order to fully open the package.
Regulation.
5.3.3.4 The package must be designed so that the user is
5. Packaging Requirements able to close the package and re-engage the release mecha-
nism(s) in a manner that requires only one re-engagement
5.1 Liquid Laundry Detergent Packets shall be contained in action on the part of the user.
opaque packaging or packaging that employs any equivalent
measure intended to mask the visibility of the individual Liquid 5.3.4 A package closure that meets the requirements set
Laundry Detergent Packets (“Package” or “Packaging”). The forth in European Commission Regulation (EU) No 1272/
package must not be labeled with graphics that make the 2008, Annex II, Part 3, Section 3.3.2 (iv), and Section 3.3
opaque package appear transparent or translucent. (Outer Packaging: Closures) of the accompanying AISE Liquid
Laundry Detergent Capsules Guidelines on CLP
5.2 Packaging described in this Voluntary standard is pack- Implementation, as may be amended. For reference, EU No
aging that is designed or constructed to be difficult for children 1272/2008, Annex II, Part 3, Section 3.3.2 (iv), is attached as
to access Liquid Laundry Detergent Packets. To comply with Annex A4, and Section 3.3 of the AISE Liquid Laundry
this standard, the package shall have the characteristics of at Detergent Capsules Guidelines on CLP Implementation is
least one of the following six options outlined below in 5.3.1. attached as Annex A6.
In addition to meeting with at least one of the six options, a 5.3.5 A package that requires the intellectual skill or cogni-
package must also: tive ability of a child at least 6 years of age to open, that meets
all of the criteria set forth in 5.3.5.1 – 5.3.5.3, or a reasonable
4
A.I.S.E., Liquid Laundry Detergent Capsules Guidelines On CLP equivalent of 5.3.5.1 – 5.3.5.3, and meets the requirements of
Implementation, Version 1.0, 27 February 2015. 5.3.5.4:

2
 F3159 − 15´1
5.3.5.1 Identification of a non-obvious opening method that 6.1.2 Distinctively separated from any other wording or
requires cognitive understanding of a manipulative concept. graphics, in a “quiet zone” (that is, placed on a single-color,
5.3.5.2 The mechanical means that secures the package is contrasting background); and
obscured from view and not readily apparent when handling 6.1.3 Located on the product in a prominent location so they
the package. are visible to the consumer.
5.3.5.3 Requires manipulation of a visual or tactile feature 6.2 The following statements shall appear on the front
in a way that is non-obvious unless the user understands the panel/principal display panel of the package:
manipulative concept. WARNING:
5.3.5.4 The package must be designed so that the user is Harmful if put in mouth or swallowed. Eye irritant.
able to close the package and re-engage the release mecha- Packets can burst if children put them in mouth or play with
nism(s) in a manner that requires only one re-engagement them.
action on the part of the user. See warning on [back/side] label.
5.3.6 A package that, in order to be opened, requires either: Keep out of reach of children.
(1) An opening force greater than that which a child is
6.2.1 See Annex A1 for an example of an FHSA-compliant
capable of generating while not being greater than a senior
layout.
adult is capable of generating, or
(2) Hand anthropometric characteristics greater than those 6.3 The Safety Alert Symbol and text of the precautionary
of an average-sized child. statements shall be laid out on the back or side panel of the
5.3.6.1 The ages and distribution of children to be tested for secondary container as set forth in ANSI Z535-4 (2011)
purposes of establishing the opening force strength or hand Figures 3 through 12 and Figures B26 through B28 or in a
anthropometric characteristics must meet the requirements set substantially similar format, except that neither the borders nor
forth in 16 CFR 1700.20(a)(2)(i). boxes depicted in those figures are required. The precautionary
5.3.6.2 The opening force strength or hand anthropometric statements shall include the following:
characteristics shall be set at the 95th percentile for children, 6.3.1 The Safety Alert Symbol, as found in ANSI Z535-4,
and the 5th percentile for senior adults. Fig. 1 (D) or (E).
5.3.6.3 Research must demonstrate that the opening force 6.3.2 Annex A2 includes additional safety symbols to be
strength or hand anthropometric characteristics do not exceed used with the warning statements. Each package must feature
those of senior adults between the ages of 50 and 70 years to at least one pair (that is, one “keep out of reach of children”
access the package, as tested pursuant to the requirements set symbol and one “keep contents out of eyes” symbol).
forth in 16 CFR 1700.20(a)(3). 6.4 The following statements shall appear on the back or
5.3.6.4 Research that establishes opening strength or hand side panel of the secondary container:
anthropometric characteristics as set forth above must be WARNING:
conducted by an independent third party, and the results Concentrated detergent packets can burst if children put
published in a peer-reviewed journal. them in mouth or play with them. The liquid inside is
5.3.6.5 A manufacturer must demonstrate through testing harmful if put in mouth, swallowed, or in eyes.
that the opening force required or hand anthropometric char- Keep packets out of reach of children.
acteristics for its packaging design are within the limits of the • Store container where children cannot reach or climb to it,
data. out of sight and in a secure place.
• Keep container fully closed.
6. Labels • Never leave any packets out of container.
6.1 Each package shall be labeled with warning statements. • DO NOT let children handle packets, even if supervised.
The warning statements shall be: Avoid breaking packets.
6.1.1 In contrasting color(s), permanent, conspicuous, and • Do not handle packets with wet or moist hands. Do not
in sans serif style font; expose packets to moisture.

FIG. 1 Safety Alert Symbol

3
 F3159 − 15´1
• Do not cut or puncture packets. If packets stick together, 6.6 Each individually-wrapped packet that is contained in a
do not try to separate them. larger outer package that contains multiple individually-
Call poison control center immediately if detergent gets in wrapped packets shall include at least one pair (that is, one
mouth or eye or on skin. Immediately and thoroughly rinse eye “keep out of reach of children” symbol and one “keep contents
or skin with water for 15 min. out of eyes” symbol) of the additional safety symbols from
6.5 All individually-wrapped sample packages shall contain Annex A2.
no more than one packet. Each individually-wrapped sample 6.7 The language listed above and icons shown in Annex A2
package shall include the following statements:
may be modified as necessary to ensure compliance with local
WARNING:
regulatory requirements, or for translation purposes. Additional
Concentrated detergent packets can burst if children put
warnings or cautionary statements or, if appropriate, alternate
them in mouth or play with them. The liquid inside is
harmful if put in mouth, swallowed, or in eyes. first aid instructions may also be included on the label,
Keep packet out of reach of children. depending on the formula, packaging used and other consid-
• Store where children cannot reach or climb to it, out of erations. Furthermore, and for the avoidance of doubt, other
sight and in a secure place. words, such as “pac” or “pack” or a trademarked name for the
• DO NOT let children handle packet, even if supervised. product, may be substituted for “packets” in these statements in
• Use packet immediately after opening. order to allow for consistent terminology on each product’s
Avoid breaking packet. package.
• Do not handle packet with wet or moist hands. Do not
expose packet to moisture. 7. Keywords
• Do not cut or puncture packet. 7.1 child deterrent; container; detergent; ingestion; laundry
Call poison control center immediately if detergent gets in packet
mouth or eye or on skin. Immediately and thoroughly rinse eye
or skin with water for at least 15 min.

ANNEXES

(Mandatory Information)

A1. FHSA-COMPLIANT PRINCIPAL DISPLAY PANEL (FRONT PANEL)

WARNING: HARMFUL IF PUT IN MOUTH OR SWAL- Keep out of reach of children.

LOWED. EYE IRRITANT. Packets can burst if children put
them in mouth or play with them. See warning on [back/side]
label.

A2. ICON AND ALERT SYMBOL EXAMPLES

A2.1 See Fig. A2.1.

4
 F3159 − 15´1

NOTE 1—If words are not included within the prohibition surround shape, use standard prohibition circle (rather than the version that is widened to
accommodate words, as seen in the first example).
FIG. A2.1 Icon and Alert Symbol Examples

5
 F3159 − 15´1

A3. OTHER INFORMATION SOURCES RELATING TO THE SAFETY OF LIQUID LAUNDRY DETERGENT PACKETS

ACCC, http://www.productsafety.gov.au/content/index. AISE (Europe), http://www.aise.eu/go.php?pid=44122

phtml/itemId/999447 & http://www.productsafety.gov.au/ &topics=1
content/index.phtml/itemId/998653/fromItemId/999447 CPSC (USA), http://www.cpsc.gov//PageFiles/132488/
Accord, http://www.accord.asn.au/public_information_ 390%20Laundry%20Packets.pdf
submission/children_and_safe_storage
ACI (USA), http://www.cleaninginstitute.org/clean_living/
singleload_liquid_laundry_packets.aspx

A4. EUROPEAN COMMISSION REGULATION (EU) NO 1272/2008, ANNEX II, PART 3, SECTION 3.3.2(iv)

A4.1 Without prejudice to the requirements of section 3.1, children to open it; (b) maintains its functionality under
be fitted with a closure that: (a) impedes the ability of young conditions of repeated opening and closing for the entire life
children to open the packaging by requiring coordinated action span of the outer packaging.
of both hands with a strength that makes it difficult for young

A5. EUROPEAN COMMISSION REGULATION (EU) NO 1272/2008, ANNEX II, PART 3, SECTION 3.3.3

A5.1 The soluble packaging shall: A5.1.2 Retain its liquid content for at least 30 s when the
A5.1.1 Contain an aversive agent in a concentration which soluble packaging is placed in water at 20°C;
is safe, and which elicits oral repulsive behaviour within a A5.1.3 Resist mechanical compressive strength of at least
maximum time of 6 s, in case of accidental oral exposure; 300 N under standard test conditions.

A6. A.I.S.E. LIQUID LAUNDRY DETERGENT CAPSULES GUIDELINES ON CLP IMPLEMENTATION (SECTION 3.3)

A6.1 Closures standing is that the closure requirements for Soluble Packaging
A6.1.1 The closure of the LLDC outer packaging must meet are different from child-resistant fastenings in section 3.1 and
two main requirements that need to be balanced: apply independently, without conflict. So section 3.1 of Annex
A6.1.1.1 impede young children from opening the packag- II continues to apply for certain mixture classifications and, in
ing and addition, the new section 3.3 applies to LLDCs regardless of
A6.1.1.2 for adults, allow easy regular opening and reclos- their classification.
ing after use. A6.1.6 No performance standards exist today for ‘child-
impeding closures’ that are not fully ‘child-resistant’ (in the
A6.1.2 These functionalities must be maintained during the
meaning of the ISO 8371 standard). Our industry is committed
packaging life span.
to work on the development of a performance standard to
A6.1.3 In addition, the pack (that is, the ‘outer packaging’ in assess the ‘child-impeding’ function of packaging, taking into
the Soluble Packaging Regulation) should be self-standing and account that ‘coordinated action of both hands’ is required.
should remain so throughout the life span of the pack.
A6.1.7 It should be noted that it would require at least two
A6.1.4 With regard to closure design, the Soluble Packaging years to shelf new packaging designs in all markets.
Regulation refers qualitatively to two elements: ‘requiring A6.1.8 In the meantime, A.I.S.E. suggests the following:
coordinated action of both hands’ and ‘a strength’ for opening. A6.1.8.1 ‘coordinated action of both hands’ for opening: in
A6.1.5 These requirements apply ‘without prejudice to the the lack of clear design description in the legal text, it is up to
requirements of section 3.1 [of Annex II to CLP]’ which each company to assess the design against compliance with this
prescribe child-resistant fastenings for specific mixture classi- general requirement. It builds on the fact that the key differ-
fications (such as skin corrosive products). A.I.S.E.’s under- entiator between adults and children is mental capacity, logic

6
 F3159 − 15´1
and dexterity. Coordination may include the required use of A6.1.8.3 ‘easily reclosable’: the outer packaging closure
hands to secure a pack to enable the opening of a closure must be able to be closed by adults in a single action, such as
system (for example, stand-up pouches). but not limited to, one clip to be pushed, a gentle pressure on
A6.1.8.2 ‘with a strength’ for opening: is to be seen in the the lid to lock, one zipper to be activated.
context of the target age group, namely children below the age A6.1.8.4 ‘maintains its functionality under conditions of
of 6 years. No strength value is specified in the legal text but repeated opening and closing for the entire life span’: the
it should be sufficient so that the closure cannot be opened closure system must meet the above criteria on opening and
unintentionally (for example, by simply touching the outer
reclosing for the designed life of the packaging, which corre-
packaging). Again, it should be borne in mind that the key
sponds to at least the number of capsules/unit doses in the outer
differentiator between adults and children is dexterity and logic
packaging.
rather than strength.

A7. A.I.S.E. LIQUID LAUNDRY DETERGENT CAPSULES GUIDELINES ON CLP IMPLEMENTATION (SECTIONS 4 AND 5)

A7.1 Aversive Agent in the Soluble Film nella et al. (2005)5. Consequently, it is reasonable to assume
that the observed oral rejection times with young adults are
A7.1.1 According to the Soluble Packaging Regulation, the
similar to or higher than what may be expected with young
soluble packaging (that is, the capsule wall) must contain an
children.
aversive agent in a concentration which is safe, and which
elicits oral repulsive behavior within a maximum time of 6 s, A7.1.8 It is important to note that it is up to each company
in case of accidental oral exposure. to demonstrate effectiveness of the aversive agent chosen to
their own situation (soluble film/agents) at design stage. This is
A7.1.2 This measure is intended to further reduce the because:
chance of ingestion of the liquid content in case a child left A7.1.8.1 different aversive agents may lead to different
unattended has managed to gain access to a capsule and places human responses and
it in his/her mouth. A7.1.8.2 the effective concentration of aversive agent may
A7.1.3 A.I.S.E. has developed and evaluated a protocol to be affected by the polymer chemical composition, presence of
determine effective levels of aversive agent contained in other chemicals in the film, etc.
soluble packaging, that is, in the soluble film. The resulting A7.1.9 It is also up to each company to select the aversive
study protocol is provided in Annex A8. agent they deem appropriate for their products, taking into
account that some limitations of use related to Intellectual
A7.1.4 The objectives of this work were:
Property may apply to certain aversive agents, films or tech-
A7.1.4.1 to develop a method for measuring the oral rejec- nologies.
tion time, as a function of the level of aversive agent in the
film; A7.1.10 It is advised to foresee a safety margin so that the
effectiveness of the aversive agent is maintained during the
A7.1.4.2 to prove the concept of effectiveness testing (at whole life cycle of the product.
different concentrations of aversive agent), in other words to
establish a ‘benchmark test’. A7.1.11 Companies will need to document the levels of
aversive agent used and the rationale, and keep such records
A7.1.5 One grade of film and one particular aversive agent for 10 years (in line with the general REACH and CLP record
were selected for the study. keeping deadlines).
A7.1.6 The A.I.S.E. study has shown that, for the particular A7.1.12 Further, the Soluble Packaging Regulation requires
aversive agent and film selected, it was possible to determine a the effective concentration of aversive agent to be safe.
level of aversive agent sufficient to elicit a median oral A.I.S.E. recommends to determine that the concentration
rejection in less than 6 s. Above this concentration, the chosen is safe in case of ingestion of the amount of film
‘dose-response’ curve was flat, that is, higher levels of aversive contained in one capsule, by means of a human health
agent were not found to lead to lower rejection times. A toxicological risk assessment, based on the highest level of
summary of the study findings is provided in Annex A9. aversive agent contained in the soluble packaging at any time
of the product life cycle and adapted to the target age group
A7.1.7 For ethical reasons, the study was run on young
(young children, including babies). The safety date sheet of the
adults instead of children. This is a conservative approach,
aversive agent is a useful source of toxicological data but may
because a child’s palate is much more sensitive than that of not be sufficient to run a full risk assessment.
adults. Infants have around 30 000 taste buds spread through-
out their mouths. By the time adulthood is reached, only about 5
Julie A. Mennella, M. Yanina Pepino, and Danielle R. Reed. Genetic and
a third of these remain, mostly on the tongue. The decreasing environmental determinants of bitter perception and sweet preferences. Pediatrics,
sensitivity to bitterness with age was demonstrated by Men- 2005, 115 (2), e216–e222.

7
 F3159 − 15´1
A7.1.13 Environmental safety should also be documented. A7.2.4 The capsules will be tested at least 24 h after
It should be reminded that the REACH Registration is the main production after having been conditioned in an environment
mechanism to assess environmental safety of substances and with a standard temperature and relative humditiy. More details
demonstrate the use is safe (unless a particular substance does are provided in the test protocols (Annex A11 and Annex A12).
not need to be registered by law). Annex A10 provides an A7.2.5 Liquid Containment Function:
example of a screening environmental risk assessment for one A7.2.5.1 The Soluble Packaging Regulation requires the
particular aversive agent (denatonium benzoate) showing that, soluble packaging to retain its liquid content for at least 30 s
even under conservative assumptions, the addition of this when the capsule is in contact with water. Some of the testing
bittering agent in unit dose soluble films is of no concern from parameters are set by the Regulation (water, temperature).
an environmental perspective. A7.2.5.2 To A.I.S.E.’s knowledge, no standard method ex-
A7.2 Capsule Integrity ists for such type of test.
A7.2.5.3 Building on the experience from its members,
A7.2.1 Two specific requirements apply under the Soluble A.I.S.E. has developed a containment function test protocol,
Packaging Regulation in relation to capsule integrity: mechani- which is provided in Annex A11 to this document.
cal resistance and liquid containment.
A7.2.6 Mechanical Integrity:
A7.2.2 Both the mechanical and the containment function A7.2.6.1 The Soluble Packaging Regulation requires the
tests are understood as ‘design tests’. They serve a safety soluble packaging to resist a mechanical compression strength
purpose in the qualification of products/validation of processes. of 300 N under standard test conditions.
They are not considered as quality control tests since it is A7.2.6.2 A.I.S.E. recommends running a dynamometric
impossible in practice to test every single capsule. test: the purpose of such compression test is to assess the
A7.2.3 These tests should be performed on an appropriate, mechanical integrity of a capsule submitted to a compressive
representative number of capsules at design stage and should strength.
be repeated, at the minimum, at every substantial design A7.2.6.3 Building on the experience from its members,
change in product, film specification, formulation or manufac- A.I.S.E. has developed a test protocol, which is provided in
turing process. Annex A12 to this document.

A8. STUDY PROTOCOL: ASSESSMENT OF THE EFFECTIVENESS OF AN AVERSIVE AGENT IN SOLUBLE FILM FOR
LIQUID LAUNDRY DETERGENT CAPSULES

A8.1 Objective A8.2.4 Each panelist, unaware of what to expect, will be

given a sheet of the soluble film containing a given level of
A8.1.1 The objective of this test is to determine the effec-
aversive agent, and will be asked to lick the film to experience
tiveness of a given aversive agent contained in a given soluble
the taste. It will then be recorded whether the panelist rejects
packaging film. The dose-response relationship of the level of
the film and if so, after how much time. panelists are only
aversive agent with the observed oral rejection time is inves-
allowed to participate once, to avoid any bias due to prior
tigated. From this, the level of aversive agent that is expected
experience with a bad tasting sample.
to lead to a rejection time below 6 s is determined.
A8.2.5 A concentration series will be tested, in two rounds.
A8.2 General Study Description First, in a screening round, a broad range of levels of aversive
A8.2.1 The response of test panelists to tasting water- agent in film shall be assessed, as well as an untreated blank.
soluble film with different levels of aversive agent is to be Subsequently, based on the screening round results, suitable
observed. From this, a dose-response relationship is to be aversive agent test levels shall be defined for a definitive
established that links the deterring effect (rejection of the film) testing round, aiming to refine the dose-response relationship
with the level of the aversive agent. for those levels leading to a rejection time close to the target of
maximum 6 s.
A8.2.2 The test panel shall consist of young adults, as a
proxy for the target audience for the safety measures on liquid A8.3 Test Material
laundry detergent capsules (that is, young children). There are
A8.3.1 The test material is a combination of one specific
reliable indications that, especially for bitter taste, children are
water-soluble film type with one specific aversive agent, at
usually more sensitive than adults.
different concentration levels. Both the water-soluble film and
A8.2.3 The test product is the water-soluble film containing the aversive agent tested shall be identified in the study report
(different levels of) the aversive agent. The film shall be used and/or in the study sponsor’s confidential study placement
in isolation for tasting: actual detergent capsules shall not be documentation. The results of the study are specific to the
used, to ensure the safety of the panelists. type/grade of water-soluble film and the type/grade of aversive

8
 F3159 − 15´1
agent used. Consequently, results cannot be extrapolated to study has been designed to eliminate as much of the adults
substantially different combinations of film and aversive ‘over thinking’ to the test as possible, attempting to gage a
agent6. ‘true’ reaction time.
A8.3.2 Preparation of Water-Soluble Film Treated with A8.4.3 The test panel shall consist of the following indi-
Aversive Agent: viduals:
A8.3.2.1 Water-soluble films with different levels of the A8.4.3.1 young adults, in the age group of 18-25 years old
aversive agent shall be prepared: A8.4.3.2 equal mix male/female
(1) Screening test: untreated (blank) – 10ppm – 100ppm – A8.4.3.3 exclusion criteria:
1000ppm – 10000ppm (*)(**) (1) smokers shall be excluded.
(*) a toxicological safety assessment shall be conducted (2) panelists with prior experience on tests of aversive
prior to the study. if toxicological concerns exist with the agents in this context shall be excluded.
highest screening levels, an alternative concentration series A8.4.4 Each panelist shall participate to only one single
with lower levels should be used. tasting session, to avoid a biased response driven by prior
(**) a range with a different upper level may be used if experience.
pre-existing information suggests this is more appropriate.
(2) Final test: 6 levels (no blank) to be determined based on A8.5 Test Design and Instructions
the outcome of the screening test.
A8.5.1 The test shall be conducted in two rounds:
A8.3.2.2 Accuracy of the aversive agent’s levels in the film, A8.5.1.1 a screening round in which a wide range of levels
and homogeneity of its distribution, shall be ensured by the of the aversive agent is assessed;
producer of the treated film.
A8.5.1.2 a final round in which the dose-response relation
A8.3.3 Preparation of the Film Sheets For Taste Testing: close to the rejection time target is refined.
A8.3.3.1 The treated water-soluble films shall be cut into A8.5.2 In the screening round, there shall be 4 test concen-
strips of 3cm by 10cm. For each test concentration, at least 12 trations in addition to a blank (untreated film). In the final
replicates shall be prepared. The strips of film shall then be round, there shall be 6 test concentrations, and no blank. There
placed in individual bags, to ensure contamination is not an shall be at least 12 replicates for each concentration. Hence, in
additional variable for the study. total, there will be at least 132 tasting sessions (5×12=60 for
A8.3.3.2 Subsequently, for each concentration, the strips the screening round, and 6×12=72 for the final round). If
shall be split into two equal batches—one batch for male deemed necessary based on the results of the screening round,
panelists, one batch for female panelists. The sets of test a higher number of replicates may be used for the final round.
specimens for female and male panelists shall be kept separate
and identified as such. A8.5.3 The levels of aversive agent for the screening round
A8.3.3.3 These test specimens shall be individually labelled are predetermined. The levels for the final round are to be
using a coding system that links the specimen to its aversive defined based on the screening results. Consequently, the final
agent level. The coding shall not disclose the level of aversive round can only be organized several weeks after the screening
agent neither to the panelists, nor to the persons directly round, to allow for processing of the screening data, and for
handing the test specimens to the panelists. This is to avoid any preparation and shipment of the film and test specimens.
bias, by applying a double-blind approach. For the same A8.5.4 For the actual testing, the test specimens shall be
reason, preparation, packing and labelling of the film strips provided to the person conducting the study in two batches:
shall be done by different persons than those conducting the one for female panelists, and one for male panelists. As outline
study with the panelists. above, each of these batches shall contain an equal number of
replicates for each test concentration. Consequently, every test
A8.4 Test Panel concentration shall be tested with an equal number of males
A8.4.1 A test panel with as many participants as there are and females, to avoid any potential bias driven by the panelists’
test specimens (that is, in total 10 test concentrations + one gender.
blank, with minimum 12 replicates each, hence a total of at A8.5.5 For every tasting session (one panelist, one level of
least 132 panelists) is required to conduct this study for one the aversive agent) the following method shall be followed:
film/aversive agent combination. A8.5.5.1 The test shall be conducted such that participating
A8.4.2 For ethical considerations, the test panel shall not panelists cannot see the reaction of others in the test, and
consist of young children, but instead, as a proxy, young adults cannot talk to others who have just completed the test. The
shall be used. It should be noted that this is expected to lead to panelists shall not be informed about the presence of an
some difference in the results, as adults tend to ‘think’ about the aversive agent in the sample. The persons providing the test
bad taste that is happening rather than react and spit it out. The samples to the panelists shall not be informed about the level
of the aversive agent in the sample.
A8.5.5.2 The panelist shall drink a defined small amount
6
(50 ml) of still water.
It is also up to each company to select the aversive agent they deem appropriate
for their products, taking into account that some limitations of use related to A8.5.5.3 The following exact instructions shall be given to
Intellectual Property may apply to certain aversive agents, films or technologies. the panelist: “This is a taste test and it is what we call

9
 F3159 − 15´1
‘double-blind’, meaning I do not know what taste you are driven by genetic differences. People who lack sensitivity in
going to receive. It could be anything from a neutral non-taste the receptors that are targeted by a specific aversive agent, will
to something pleasant or unpleasant, it could be salty, sweet, experience the aversive taste to a limited extent, if at all
acidic7 etc. If, when you are licking it, you think the taste is (irrespective of the concentration of the aversive agent). For
neutral or pleasant, I want you to continue licking it until I tell example, in Sibert & Frude (1991)8 in a test where children
you to stop. If, when you are licking it, you discern that the were given orange juice spiked with a common aversive agent
taste is something unpleasant, I want you to stop licking it (denatonium benzoate) at a level known to be effective, over 15
immediately. You are going to take the film that is in the bag % of the test subjects showed no evident response.
and hold it in your hands and lick it like so...” and then the A8.6.3 The aim of the study is to determine the appropriate
panelist will be shown how to hold and lick the film. aversive level that leads to oral rejection within 6 s of the initial
A8.5.5.4 At the moment of contact of the film strip with the exposure. For non-sensitive subjects, this rejection time cannot
tongue/mouth, a timer shall be started, and no further instruc- be achieved, irrespective of the level of aversive agent used.
tions shall be provided to the panelist. Each panelist’s reaction Consequently, data from non-sensitive subjects should be
may be filmed for future reference. ignored when determining the appropriate level. Hence, the
A8.5.5.5 It shall be recorded whether the panelist rejected median of the different replicates at a given level shall be used
the test specimen prior to the strip’s dissolution in half, and if as the relevant metric for comparison with the 6-s target.
so, exactly after how many seconds the rejection occurred.
A8.5.5.6 Participants shall be given something to eat or A8.6.4 By means of suitable statistical methods (to be
drink to remove the bad taste. What is to be offered will depend determined case-by-case, depending, for example, on the
on the aversive agent under study. For example, for bittering distribution shape and amount of scatter of the data) it shall be
agents, strong dark chocolate is known to effectively remove determined which levels of aversive agent have led to a median
the bitter taste. In addition, flavoured lip balm shall be offered rejection time below the target of 6 s, with at least 90 %
in case the bad flavour has travelled to the lips of the panelists. confidence. If feasible (depending on the quality of the data), a
A8.5.5.7 Exclude the panelist from any further participation mathematical dose-response relationship shall also be
to this test or similar tests in the future. developed, that allows to determine rejection time as a function
of the aversive agent level. Furthermore, it shall be determined
A8.6 Analysis and Reporting of Results up to which level of aversive agent the observed rejection time
A8.6.1 All raw data collected during the study shall be is not significantly different from the blank; and as of which
reported, except for the identities of the panelists (that are to level of aversive agent the observed rejection time no longer
remain confidential to the testing laboratory). Note that these decreased.
identities shall be archived by the testing laboratory for further A8.6.5 The final outcome of the study is the determination
reference, to avoid their participation in other similar studies in of the lowest aversive agent level that is expected to lead to a
the future. median rejection time (either observed as tested; or calculated
A8.6.2 Among the panelists, it is expected that there will be if a mathematical dose-response relationship could be devel-
a natural variability in taste receptor sensitivity, primarily oped) below 6 s, with at least 90 % confidence.

7
The actual description of the taste of the aversive agent under study shall not
8
be used here. For example if a bittering agent is used, the word ’bitter’ shall not be Sibert J. R. Frude N. (1991). Bittering agents in the prevention of accidental
mentioned; if the aversive agent has an acidic taste, the word ’acidic’ shall not be poisoning: children’s reactions to Denatonium Benzoate (Bitrex). Archives of
used, etc. Emergency Medicine, 1991, 8, 1-7.

A9. SUMMARY OF INTERTEK STUDY FINDINGS: “ASSESSMENT OF THE EFFECTIVENESS OF AN AVERSIVE AGENT IN
SOLUBLE FILM FOR LIQUID LAUNDRY DETERGENT CAPSULES”

A9.1 Executive Summary is typically less or not sensitive to a given aversive agent due
A9.1.1 The proposed test method to assess the effectiveness to natural variability (genetic predisposition), which leads to
of an aversive agent in soluble film for liquid laundry detergent skewed distributions and scattered observational data. This
capsules was found to be practically feasible, and to allow implies that a sufficient number of replicates (at least 12 but
defining an aversive agent’s effective level in the context of ideally more) is required per tested level of aversive agent, to
Commission Regulation (EU) No 1297/2014. ensure robustness of the results.
A9.1.2 It is recommended to use the median oral rejection A9.1.3 For one specific grade of PVA film, treated with the
time as the appropriate metric to assess compliance with the bittering agent denatonium benzoate, a dose-response relation-
requirements. Non-parametric statistical methods are needed, ship was observed with a decreasing rejection time up to 220
because the rejection times between panelists are not normally ppm. The rejection time remained the same when the aversive
distributed. Specifically, a certain percentage of the population agent’s level was further increased beyond this level. The

10
 F3159 − 15´1
median rejection time for levels>= 220 ppm was 2.7 s, and was noted that follow-up to assess possible non-sensitivity of
demonstrated to be significantly less than 6 s with >95 % panelists with long rejection times was not found to add
confidence. substantial value. Instead, appropriate statistical methods
should be used that implicitly take into account the ‘biological
A9.2 Background outliers’.
A9.2.1 Commissioned by A.I.S.E., Intertek carried out a A9.3.5 As a consequence of the non-normality, the use of
study to measure the reaction time of young adults when mean rejection time is not relevant. Instead, the median should
coming into oral contact with soluble film treated with an be used, as this is independent of the distribution shape at its
aversive agent. The response of the test panelists to tasting extremes. Using Sign Analysis (a non-parametric method) it
water-soluble film with different levels of aversive agent was can be assessed whether the observed median is significantly
observed. From this, a dose-response relationship was estab- below the required threshold of 6 s, with a given level of
lished that links the deterring effect (rejection of the film) with statistical confidence (for example, 90 or 95 %).
the level of the aversive agent.
A9.3.6 Another consequence of the non-normality is that a
A9.2.2 The objectives of this study were twofold: sufficiently high number of replicates is required for each
A9.2.2.1 the development of a method for measuring the tested level. 12 replicates per level, as applied in the final round
oral rejection time by young adults, as a function of the level of this study, is judged to be a minimum. But a larger number
of aversive agent present in water-soluble film of detergent of replicates is to be preferred, to increase statistical robust-
capsules; and ness.
A9.2.2.2 proof of the concept with one specific commonly-
used aversive agent and one specific film. A9.4 Proof of concept for a specific PVA film containing
Denatonium Benzoate
A9.2.3 For the method development, a pilot study was
conducted with internal Intertek employees. Next, through a A9.4.1 As a proof of concept, the method was applied to
screening study with 50 panelists (5 tested levels, 10 replicates determine the required effective level of one specific aversive
each), it was determined what are the appropriate levels of the agent (a bittering agent: denatonium benzoate) selected based
aversive agent to be tested in more detail. A final study was on its commonality and one specific polyvinyl alcohol (PVA)
then conducted with 72 panelists (6 tested levels, 12 replicates film grade (Monosol M8630).
each). A follow-up study with orange juice that was spiked A9.4.2 No reduction of the oral rejection time versus
with the aversive agent, was conducted afterwards with 10 untreated film was seen up to 10 ppm of denatonium benzoate
panelists, to assess whether these may have been non-sensitive in the film. At 50 ppm, a clearly lower rejection time was
to the aversive agent. All these studies were conducted at the observed, and this further decreased at 110 ppm and again at
Intertek facility in Oak Brook, Illinois in the US. 220 ppm, where a median value of less than 3 s was reached.
Higher levels did not cause the median rejection time to drop
A9.3 Method Development further. The dose-response relationship is shown in the below
A9.3.1 Overall, it can be concluded that the developed test chart. Please note that for the ppm levels a logarithmic scale
protocol is practically feasible and that it can be used to was used. The data shown are from the final study except the
determine the effective level of an aversive agent leading to data points in red (screening study). (See Fig. A9.1.)
rejection within a defined time period. NOTE A9.1—The median of 9.8 s observed in the screening round for
A9.3.2 The most suitable method of delivery was found to 1000 ppm is judged to be on artifact caused by the too limited number of
replicates. When the observed rejection times for 1000 ppm (screening
be a sheet of film (3×10cm), to be licked by the panelists until
round) and those for the very similar level of 960 ppm (final round) are
discerning that the taste is something unpleasant. Clear word- grouped, the median is 3 s.
ing was developed to have unambiguous instructions for the
A9.4.3 The observed dose-response relationship is statisti-
panelists. This was well understood (with only 1 exception out
cally supported by the Mann-Whitney test. This shows that the
of 132 panelists).
rejection times at the higher levels were not significantly
A9.3.3 To rule out any difference due to different taste different from those at 220 ppm (that is, flat dose-response
sensitivities between males and females, both genders should beyond 220 ppm). Further, this test shows that all treatment
be equally represented and each gender group should receive levels in the final study led to significantly lower rejection
the same distribution of aversive agent levels tested. times than the 0 ppm blank, and that the rejection time at 220
A9.3.4 The observed rejection times (especially for those ppm was significantly less than at 110 ppm. Finally the test
aversive agent levels that lead to a substantial repulsive effect), shows that the rejection time at 10 ppm (screening round) was
were found to not follow a normal distribution. A majority (75 not less than for the blank.
to 80 %) was clustered around a short rejection time, while the A9.4.4 The median oral rejection time for denatonium
remainder was very scattered. This is directly driven by the benzoate levels in film ≥220 ppm (in the final study) was on
biology: genetically, a certain part of the population has less (or average 2.7 s. Sign analysis shows that for each of these levels,
no) effective receptors for the specific aversive taste. As such, the median was significantly below 6 s, with a confidence level
it can be anticipated that similar distribution shapes may be of >95 %. The 75th percentile of the observed rejection times
found with other aversive agents and/or other soluble films was also below 6 s for all levels ≥220 ppm (in the final study),
than the ones used for the method development. It should be however, statistical significance could not be demonstrated.

11
 F3159 − 15´1

FIG. A9.1 Final Study Data

A9.4.5 It can be concluded that, for the specific film grade film is adequate to meet the requirements of Commission
that was tested, a denatonium benzoate level of 220 ppm in the Regulation (EU) No 1297/2014.

A10. SCREENING ENVIRONMENTAL RISK ASSESSMENT FOR DENATONIUM BENZOATE (EXAMPLE OF AVERSIVE
AGENT)

A10.1 Substance Identification the EU ecolabel). The assessment factor to extrapolate from an
acute EC50 to the ecosystem safe level is a factor 1000. Hence,
A10.1.1 Denatonium benzoate is a salt of the quaternary
the PNEC = 13 µg/L.
ammonium cation denatonium with the inert anion benzoate:
A10.2.2 Biodegradability:
A10.2.2.1 The active cation denatonium was not found to be
either biodegraded or adsorbed to sludge in a Semi-Continuous
Activated Sludge (SCAS) study (Corby et al., 1993). As a
SCAS test simulates fate in actual sewage treatment plants, it
CAS 3734-33-6 is fair to assume that denatonium benzoate will not be removed
Molecular formula C28H34N2O3 in sewage treatment.
Molar mass 446.581
A10.2.2.2 Furthermore (cf. CPSC, 1992), in an OECD
A10.2 Environmental Properties 301D test, no chemical deterioration of Denatonium benzoate
was observed. In the Zahn-Wellens test (OECD 302B), a 36 %
A10.2.1 Ecotoxicity: breakdown was found after 28 days. A carbon dioxide produc-
A10.2.1.1 In the European Classification & Labelling noti- tion test showed that denatoinum benzoate is poorly metabo-
fication process (ECHA, 2015), denatonium benzoate was lized (4.5 % after 28 days).
notified as Aquatic Chronic 3 (H412) by most notifiers. A10.2.2.3 In the EU Ecolabel DID LIST, denatonium ben-
A10.2.1.2 In the EU Ecolabel DID LIST (European zoate is assumed to be not removed in sewage treatment
Commission, 2014), denatonium benzoate is included (ingre- (DF=1).
dient nr. 2604). As the relevant acute LC50, a value of 13 mg/L A10.2.3 Bioaccumulation—Denatonium benzoate is highly
is mentioned. Chronic data are absent. water soluble (45 g/L) and has a low octanol/water partitioning
A10.2.1.3 The following ecotoxicologal data are reported in coefficient (Kow = 0. 91) (cf. Health Canada, 2011).
several safety data sheets (from multiple suppliers) of denato- Consequently, there is no risk for bioaccumulation.
nium benzoate, and/or in regulatory reviews (for example, US
CPSC 1992; Health Canada, 2011): A10.3 Environmental Risk Assessment
(1) Fish: 96h LC50 Rainbow Trout: >1000 mg/L A10.3.1 Tonnage Estimate:
(2) Invertebrates: 96h LC50 Shrimp (salt water): 400 mg/L A10.3.1.1 As it has not yet been registered under REACH,
(3) Invertebrates: 48hr EC50 Daphnia magna: 13 mg/L the tonnage of denatonium benzoate across the EU is in the
(4) No effects on bacteria up to 150 mg/L order of <100 ton per year per legal entity. For the purpose of
A10.2.1.4 The Predicted No-Effect Concentration (PNEC) this screening assessment one could conservatively assume a
can be derived from the lowest acute data point, in this case for total of 100 ton per year, which is equivalent to 200 mg per
the water flea Daphnia magna (which is also the value used for capita per year in the EU (with 500 million people).

12
 F3159 − 15´1
A10.3.1.2 The incremental consumption of denatonium A10.3.2.4 The PEC/PNEC ratio for denatonium benzoate is
benzoate in the context of liquid laundry detergent capsules, 0.55 / 13 = 0.04. In other words, the calculated safety margin
can be estimated as follows: is by a factor >20. It should be noted that this is based on a
(1) One laundry capsule of 5cm × 5cm with a film conservative tonnage estimate of total consumption, which is
thickness of 100 µm has 5 × 5 × 0.01 × 2 sides = 0.5 cm3 of also nearly two orders of magnitude higher than the expected
soluble film as outer packaging. With a density of 1.3 this use of this substance for laundry capsules.
corresponds to 0.65 g of film per capsule.
(2) When 200 [respectively 1000] ppm is used as aversive A10.4 Conclusion
agent contained in the film, this leads to the presence of 130
A10.4.1 Using conservative assumptions, especially regard-
[resp. 650] µg of denatonium benzoate per capsule.
ing tonnage, this screening assessment shows no concerns with
(3) In the United Kingdom, which is to date the most
mature market for laundry capsules, on average about 20 the environmental safety of denatonium benzoate. The incre-
capsules are sold per year per inhabitant (total market: 1150 mental use of denatonium benzoate as aversive agent in
million capsules; population of 64 million). laundry detergent capsules is minimal compared to the as-
(4) This corresponds to 2.6 [resp. 13] mg of denatonium sumed total tonnage, and is not anticipated to negatively
benzoate per capita per year. impact this conclusion.
A10.3.1.3 Consequently, an assessment of the assumed
A10.5 References
current tonnage of 100 ton/year in the EU (= 200 mg/cap.year)
Corby, J., Doi, J., Conville, J., Murphy, S. et al., “Biode-
covers any potential increase due to the introduction of
denatonium benzoate in the soluble film of laundry capsules at gradability of a Denatonium Bitterant,” SAE Technical Paper
the envisaged levels. 930587, 1993, doi:10.4271/930587.
European Chemicals Agency http://echa.europa.eu. Ac-
A10.3.2 Risk Assessment: cessed 7.1.2015.
A10.3.2.1 The average water use per person per year in the European Commission (2014). Detergents Ingredients
EU is 100-200 L per capita per day (EEA web site).
Database, version 2014.1 http://ec.europa.eu/environment/
Conservatively, a water use of 100 L/day is assumed.
ecolabel/documents/did list/didlist part a en. pdf
A10.3.2.2 Assuming 100 ton/year in the EU, the concentra-
European Environment Agency http://www.eea.europa.eu/
tion of denatonium benzoate in household waste water is 200
mg/cap. year divided by 365 days/year × 100 L/cap.day = 5.5 themes/households Accessed 7.1.2015.
µg/L. As denatonium benzoate is not removed in sewage Health Canada (2011). Proposed Re-evaluation Decision
treatment plants, this is also the predicted concentration for PRVD2011-15 Denatonium Benzoate 08 November 2011.
treated effluent. Finally, the Predicted Environmental Concen- http://www.hc-sc.gc.ca/cps-spc/altformats/pdf/pubs/pest/
tration (PEC) in river water, taking into account a standard decisions/rvd2012-06/rvd2012-06-eng.pdf
dilution factor of 10, is 0.55 µg/L. US Consumer Product Safety Commission 1992. Study of
A10.3.2.3 The aquatic PNEC for denatonium benzoate is 13 Aversive Agents https://www.cpsc.gov//PageFiles/96066/
µg/L (derived from Daphnia magna acute data with an assess- aversive.pdf
ment factor of 1000).

A11. CONTAINMENT FUNCTION TEST PROTOCOL

A11.1 This test is a design test. It serves a safety purpose for A11.4 Test Method
the qualification of products and the production process.
A11.4.1 A beaker of sufficient capacity is filled with at least
A11.2 This test shall be performed on an appropriate, 1 L of demineralised water.
representative number of capsules and repeated, at the
A11.4.2 Once the temperature has stabilised at 20°C, one
minimum, at every substantial design change in product, film
pre-conditioned capsule is gently introduced into the beaker
specification, formulation or manufacturing process.
until it is entirely submerged by water. The capsule shall be
A11.3 Sample Conditioning Prior to Testing surrounded by water on all sides.
A11.3.1 Capsules shall be tested after having been condi- A11.4.3 In case the density is such that the capsule either
tioned at 23 6 1°C/50 6 2 % Relative Humidity for at least floats or sinks, the capsule shall be placed inside a device that
24 h in the original outer packaging opened to the conditioning prevents floating or sinking, such as a metal cage, a netting bag
atmosphere. or a similar device that allows visual observation.
A11.3.2 These conditions are in line with Practice D4332.

13
 F3159 − 15´1

FIG. A11.1 Stages of Containment Loss

A11.5 Recording Containment Loss not releasing their content within this period of time are
A11.5.1 A timer shall be started as soon as the capsule is recorded as successful, capsules that release content before 30
submerged by water. s are counted as failures. By means of binomial statistics,9 it
can be determined how many failures are allowed as a function
A11.5.2 The dissolution of the capsule shall be observed of the total number of samples tested, to achieve an overall 85
visually as a function of time, with the following event % success rate with 90 % confidence:
recorded: “Liquid Content Release,” which corresponds to the
Number of failures allowed
first visual evidence of liquid leaving the capsule. To success- Number of samples tested (85 % success rate
fully pass the criteria, the time recorded once liquid content at 90 % confidence)
released is observed should be at least 30 s. < 15 No valid test possible
15-24 0
A11.5.3 The following pictures (see Fig. A11.1) visually 25-33 1
34-42 2
illustrate the observable stages of containment loss (these 43-51 3
experiments were not strictly conducted according to the above 52-59 4
protocol and visual observations may differ depending on the 60-67 5
68-76 6
capsule design, colour size or shape. These pictures are only 77-84 7
for illustrative purpose). 85-92 8
A11.5.3.1 Prior to product release with fully closed contain- 93-100 9
ment. NOTE A11.1—The above table can only be used in the context of design
A11.5.3.2 The moment in time when first release of prod- testing, to ensure with 90 % statistical confidence that 85 % of the samples
uct occurs is noted. will meet the criteria. The table is not applicable for other purposes such
A11.5.3.3 Further progress of product release and air es- as inspections, for which different statistical criteria need to be applied.
capes from capsule. A11.7.1.2 Determining the Content Release Time by De-
structive Testing—The test is conducted until the liquid content
A11.6 Criteria for Passing the Test
starts to be released, and this time is recorded for each capsule
A11.6.1 In line with general principles of testing of safety- in the test. From these data, a statistical distribution is
related features (such as ISO 8317), the test will be successful constructed. Using appropriate statistical methods, that depend
when at least 85 % of the capsules tested do not release their on the shape of the observed distribution, it can be determined
content within minimum 30 s, with a 90 % confidence level. whether 85 % of the population of samples will have a content
release time greater than 30 s, with 90 % confidence. The
A11.7 Experimental Design information can also be used to optimise the number of
A11.7.1 A.I.S.E. recommends applying one of the two samples required for future testing.
following methods to determine whether the content release
time is at least 30 s, for 85 % of the capsule, with 90 %
confidence. 9
Gilliam, D., Leigh, S., Rukhin, A., and Strawderman, W., “Pass-Fail Testing:
A11.7.1.1 Attribute Test, Non-destructive—The test is con- Statistical Requirements and Interpretations,” J. Res. Natl. lnst. Stand. Technol. 114,
ducted for exactly 30 s for each capsule that is tested. Capsules 195-199 (2009).

14
 F3159 − 15´1

A12. DYNAMOMETRIC TEST PROTOCOL

A12.1 This test is a design test. It serves a safety purpose for A12.4.5 The type of instrument used for such testing is
the qualification of products and the production process. shown in Fig. A12.1.10

A12.2 This test shall be performed on an appropriate, A12.5 Criteria for Passing the Test
representative number of capsules and repeated, at the A12.5.1 In line with general principles of testing of safety-
minimum, at every substantial design change in product, film related features (such as ISO 8317), the test will be successful
specification, formulation or manufacturing process. when at least 85 % of the capsules tested resist a mechanical
compression of 300N, with a 90 % confidence level.
A12.3 Sample Conditioning Prior to Testing
A12.6 Experimental Design
A12.3.1 Capsules will be tested after having been condi-
tioned at 23 6 1°C/50 6 2 % Relative Humidity for at least A12.6.1 A.I.S.E. recommends applying one the two meth-
24 h in the original outer packaging opened to the conditioning ods to determine whether 85 % of the capsules resist against a
atmosphere. mechanical compression of 300 N, with 90 % confidence.
A12.6.1.1 Attribute Test, Non-destructive—The test is con-
A12.3.2 These conditions are in line with Practice D4332. ducted until a compression strength of exactly 300 N is reached
for each capsule that is tested. Capsules resisting this compres-
A12.4 Test Method sion are recorded as successful, capsules that burst before 300
A12.4.1 One capsule is submitted to an increasing compres- N is reached are counted as failures. By means of binomial
sion force at a rate of 200 to 250 mm/min (typically in the statistics,9 it can be determined how many failures are allowed
range of operation of standard equipment) until 300 N is as a function of the total number of samples tested, to achieve
reached or until it releases its content, under standard test an overall 85 % success rate with 90 % confidence:
conditions. Number of failures allowed
Number of samples tested (85 % success rate
at 90 % confidence)
A12.4.2 ‘Standard test conditions’ in this context refers to < 15 No valid test possible
test conditions which are similar to the conditioning atmo- 15-24 0
sphere of capsules (see previous paragraph). Therefore, cap- 25-33 1
34-42 2
sules shall be tested shortly after having been sampled from the 43-51 3
conditioning atmosphere. 52-59 4
60-67 5
A12.4.3 The instrument is made of two flat plates of a 68-76 6
77-84 7
surface larger than the surface area of the capsule. 85-92 8
93-100 9
A12.4.4 The capsule is to be placed in a plastic bag to avoid
spillage and positioned between the two plates that apply the NOTE A12.1—The above table can only be used in the context of design
testing, to ensure with 90 % statistical confidence that 85 % of the samples
force, resting on its largest surface area. will meet the criteria. The table is not applicable for other purposes such
as inspections, for which different statistical criteria need to be applied.
A12.6.1.2 Determining the maximal compression strength
by destructive testing—For each capsule, the mechanical com-
pression strength shall be gradually increased at the rate
mentioned above until the capsule breaks. The strength value
applied at this point of breakage shall be recorded. The
distribution curve (number of samples versus strength applied
at the break point) shall be constructed. Using appropriate
statistical methods that depend on the shape of the observed
distribution, it can be determined whether 85 % of the
population of samples will have a compression resistance
greater than 300 N, with 90 % confidence. The information can
also be used to optimise the number of samples required for
future testing.

10
The sole source of supply of the apparatus (Instron model 5566) known to the
committee at this time is Instron, 825 University Ave, Norwood, MA, 02062-2643,
www.instron.com. If you are aware of alternative suppliers, please provide this
information to ASTM International Headquarters. Your comments will receive
careful consideration at a meeting of the responsible technical committee,1 which
FIG. A12.1 Testing Instrument you may attend.

15
 F3159 − 15´1
ASTM International takes no position respecting the validity of any patent rights asserted in connection with any item mentioned
in this standard. Users of this standard are expressly advised that determination of the validity of any such patent rights, and the risk
of infringement of such rights, are entirely their own responsibility.

This standard is subject to revision at any time by the responsible technical committee and must be reviewed every five years and
if not revised, either reapproved or withdrawn. Your comments are invited either for revision of this standard or for additional standards
and should be addressed to ASTM International Headquarters. Your comments will receive careful consideration at a meeting of the
responsible technical committee, which you may attend. If you feel that your comments have not received a fair hearing you should
make your views known to the ASTM Committee on Standards, at the address shown below.

This standard is copyrighted by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959,
United States. Individual reprints (single or multiple copies) of this standard may be obtained by contacting ASTM at the above
address or at 610-832-9585 (phone), 610-832-9555 (fax), or [email protected] (e-mail); or through the ASTM website
(www.astm.org). Permission rights to photocopy the standard may also be secured from the Copyright Clearance Center, 222
Rosewood Drive, Danvers, MA 01923, Tel: (978) 646-2600; http://www.copyright.com/

16