Content uploaded by Andrew Adamczyk
Author content
All content in this area was uploaded by Andrew Adamczyk on Jan 19, 2016
Content may be subject to copyright.
1
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
Aldehyde Emissions from Flexible Molded Foam
JENNIFER E. A. AL-RASHID
Air Products and Chemicals, Inc
7201 Hamilton Blvd
Allentown, PA 18195
TORSTEN PANITZSCH
Air Products GmbH
Schnackenburgallee 41a
22525 Hamburg, Germany
JACKSON SU
Air Products and Chemicals
China Investment Co., Ltd
No 88, Lane 887, Zu Chong Zhi Road,
Shanghai 201203, P.R.C
GAURI SANKAR LAL
Air Products and Chemicals, Inc
7201 Hamilton Blvd
Allentown, PA 18195
ANDREW J. ADAMCZYK
Air Products and Chemicals, Inc
7201 Hamilton Blvd
Allentown, PA 18195
ALLEN R. ARNOLD
Air Products and Chemicals, Inc
7201 Hamilton Blvd
Allentown, PA 18195
ABSTRACT
Automotive OEM requirements for indoor air quality have made volatile organic compounds (VOC) and aldehyde emissions
increasingly important to the polyurethane industry. As more OEM’s implement aldehyde emission requirements Air
Products’ is actively developing offerings to enable our customers to meet the large range of requirements in different regions.
This paper will detail efforts to develop an aldehyde scavenger to lower aldehyde emissions from molded polyurethane foam.
There are multiple methods used to measure aldehyde emissions and it is challenging to get reproducible results. Air Products
has developed techniques to obtain reproducible emission results. Experiments were run to identify the different sources of
aldehydes in foam emissions and are detailed in the paper. Different chemistry families have been looked at for their ability
to react with formaldehyde, acetaldehyde, and acrolein and both experimental and computational results will be shared as well
as the physical properties and emissions of three new aldehyde scavenger prototypes.
INTRODUCTION
In industrialized nations the average person spends more than one hour per day in vehicles and around 90% of their time
indoors [1, 2]. Concerns about indoor air quality and aldehydes began in Japan in the 1960’s when people were affected by
sick house syndrome caused by high indoor levels of formaldehyde from the urea-formaldehyde binder in particle boards. As
a result, guideline values for formaldehyde and individual volatile organic compounds (VOC) were announced [3]. In
Germany MAK values, (“maximale Arbeitsplatz-Konzentration”: maximum workplace concentration) the maximum
concentration of a chemical in the workplace which does not have known adverse effect on the health when the person is
repeatedly exposed, are published annually [8]. Formaldehyde comes to the attention of the general public when publicized
events occur such as in 2006 when high levels of formaldehyde were found inside mobile homes in the US and in 1992 when
high levels of formaldehyde emissions were found in Germany from furniture coatings [4].
German car manufacturers have set VOC specifications using VDA (Verband der Automobilindustrie) and chamber test
methods to measure emissions from materials in the interior of the automobile. The Japanese Automobile Manufacturers
Association (JAMA) set voluntary VOC specifications for the cabin of new models of passenger cars manufactured and sold
beginning in fiscal year 2007. The indoor concentration guidelines were established for 13 VOCs by the Ministry of Health,
Labor, and Welfare. Each carmaker must continuously strive to reduce VOC concentrations in the interior cabin [5]. The
Chinese government introduced regulations beginning in 2007 with the VOC test method HJ-T400-2007 for Cabin of
2
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
Vehicles. Beginning March 2012, GB/T 27630-2011 “Guidelines for air quality assessment of passenger car” which is based
on the Chinese “Indoor air quality standards” GB/T 18883-2002 was formally implemented [6, 7].
Based on these industry wide guidelines, this paper focuses on understanding the sources of three aldehydes, formaldehyde,
acetaldehyde, and acrolein in polyurethane foam currently of highest concern in the automotive industry. Three new offerings
will be discussed that have been found to lower aldehyde emissions from polyurethane foam. Table 1 shows a few
characteristics of the aldehydes.
DEFINITIONS
TLV: Threshold Limit Value: a level to which it is believed a worker can be exposed day after day for a working lifetime
without adverse health effects.
TWA: Time Weighted Average: average exposure on the basis of an 8h/day, 40h/week work schedule.
STEL: Short Term Exposure Limit: spot exposure for a duration of 15 minutes that cannot be repeated more than 4 times per
day with at least 60 minutes between exposure periods.
Table 1. Aldehyde Information.
Formaldehyde Acetaldehyde Acrolein
TLV 0.30 ppm 25 ppm 0.1 ppm
TWA 0.75 ppm 360 mg/m
3
0.1 ppm
STEL 2 ppm Not Determined 0.8 mg/m
3
RAW MATERIALS
Data presented in this paper is a summary of hand mixed produced foams. Standard MDI and TDI formulations were used.
The following list of raw materials was used:
Tertiary Amine Catalysts
DABCO®NE1070 Second generation of novel non-emissive (NE) gelling catalyst binding chemically into the polyurethane
foam matrix
DABCO®NE300 Non-emissive (NE) strong blowing catalyst binding chemically into the polyurethane foam matrix
Silicone Surfactants
DABCO®DC2525 Low emissions MDI silicone surfactant from Air Products and Chemicals, Inc.
DABCO®DC6070 Low emissions TDI silicone surfactant from Air Products and Chemicals, Inc.
Cross linker
DABCO®DEOA-LF Diethanolamine cross linker from Air Products and Chemicals, Inc.
Polyols
Polyol A Conventional polyol with an OH# of 28
Polyol B Copolymer polyol with an OH# of 22 and a solids level of 41%
Polyol C Cell opening polyol having an OH# of 35 and an EO content of 75%
Isocyanates
TDI A 80/20 2,4/2,6 isomer blend of toluene diisocyanate
3
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
MDI A modified isomer blend of diphenylmethane diisocyanate (32.5 % NCO)
METHODS
Hand-Mix Foam Procedure
Hand mix evaluations were carried out by mixing together all the components except isocyanate. Pre-blends were kept at
23 °C. TDI (an 80/20 2,4/2,6-isomer blend of toluene diisocyanate) or MDI (modified isomer blend of diphenylmethane
diisocyanate) were added in an amount corresponding to the isocyanate index. Further mixing with a mechanical stirrer at
5,000 rpm was performed after isocyanate addition. The foaming mixture was transferred to a paper bucket and allowed to
free rise while data collection was performed. Hand mix flexible molded foam parts were prepared by pouring the reactive
mixture into a heated aluminum mold at 70 °C. The mold had internal dimensions of 12 in x 12 in x 4 in. Solvent based release
agent was sprayed into the mold prior to every pouring to allow easy removal of the finished part from the mold. Pouring of
the reactive mixture was typically done at the center of the mold to allow for a uniform distribution of the polymerizing mass
into the mold. Demolding of foam was typically done after 4 minutes to allow proper polymer cure. Foam cure assessment
was done following the guidelines described in the literature [14]. The foam was allowed to sit in a controlled temperature
and humidity room for seven days prior to testing.
Formulations
Foam was prepared with the TDI or MDI formulations below in Tables 2 and 3. Emissions were measured with the three
test methods described above.
Table 2. 41kg/m
3
TDI Foam Formulation
Material pphp
Polyol A 85
Polyol B 15
Polyol C 1.5
Dabco®DC6070 1
Dabco®DEOA-LF 1.13
Added water 2.18
Dabco®NE1070 0.8
Dabco®NE300 0.2
TDI index 100
Table 3. 41kg/m
3
MDI Foam Formulation
Material pphp
Polyol A 100
Polyol C 1.5
Dabco®DC2525 1
Dabco®DEOA-LF 0.9
Added water 2.83
Dabco®NE1070 1.1
Dabco®NE300 0.2
MDI-100 index
EMISSION METHODS
Various emission methods are used to measure aldehydes depending on the OEM and region of the world. Three methods
were examined for the work in this paper and a brief description of each follows.
Bottle Method
4
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
GMW15635 is an aldehyde emission technique that uses heated water in a polyethylene bottle to extract the aldehydes from
the polyurethane foam. In the GM method, a foam sample is hung in a polyethylene bottle above water and heated for three
hours at 60
o
C. The water is then analyzed for aldehyde content using HPLC. The results are expressed as µg/g. The method’s
detection limit for formaldehyde, acetaldehyde, and acrolein is 0.35 µg/g [9].
Bag Methods
Tedlar bag emission methods range in size from 3L to 2000L. A sample is placed inside the bag which is then sealed, filled
with nitrogen or air, and then heated [10]. After heating, the gas is sampled, and aldehydes are collected on a dinitrophenyl
hydrazine (DNPH) column. They are then quantified using HPLC. The results are expressed as emissions per volume of air
sampled (µg/m
3
). An example of a bag method is NES M0402 for Nissan. The detection limit for formaldehyde, acetaldehyde,
and acrolein is 10 µg/m
3
[11].
Chamber Methods
Emission techniques using a stainless steel chamber and include the 1m
3
chamber used by BMW, method number GS 97014-
3, and the micro chamber method used in the spray foam industry (ISO 12219-3). In the BMW chamber test 1 kg of foam is
heated to 65
o
C with air flowing over it. The VOCs and aldehydes are collected, quantified, and expressed as emissions per
mass of foam per volume of the sample chamber (µg/m
3
/kg) [12]. The Microchamber / Thermal Extractor (µ-CTE, Markes
International) comprises small cylindrical chambers each with an interior volume of approximately 45 mL and typical sample
size of 45 mm. By reducing the chamber volume, it was intended to reduce typical emission test time and costs while
generating meaningful emissions data that correlate with data from conventional emissions test chambers [13].
Examples of Emission Techniques
Table 4 shows representative aldehyde emission specifications from different OEM’s. The test methods all measure
aldehydes but the chemistry caused by the measurement conditions are different making the results difficult to compare. The
units on the emission measurements are also all different. A polyurethane supplier to the automotive industry would have to
run multiple test methods for the different OEM’s which is costly, time consuming, and confusing.
Table 4. Examples of Aldehyde Emission Specifications.
OEM
(Method) Type
HH
O
Formaldehyde
HO
Acetaldehyde
HO
Acrolein
BMW GS 97014-3
Chamber
<30 µg/m
3
/kg
<30 µg/m
3
/kg
Non-detect
Nissan NES M0402
Tedlar Bag <150
µ
g/m
3
<60 µ
g/m
3
Ford M99P2222
Tedlar Bag <60
µ
g/m
3
sum of aldehydes <120
µ
g/m
3
GM GMW15635 Bottle <8
µ
g/g <8
µ
g/g <8
µ
g/g
RESULTS
Establishing a Reproducible Test Method to Measure Aldehyde Emissions
Molded foam was prepared with the TDI formulation in Table 2 and aldehyde emissions were measured with four different
methods. Table 5 shows the results with four emission methods. The results vary greatly depending if the emission method
was a static bag method, a dynamic chamber method, or an equilibrium based bottle method. To determine the reproducibility
of three test methods, six aldehyde measurements were made with foam prepared with the TDI formulation shown in Table
2. Table 5 shows the average and standard deviation of the three aldehyde test method results [10]. Poor reproducibility with
5
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
large standard deviation was seen with both the Tedlar bag method and 1m
3
chamber. Surprisingly, the micro chamber was
unable to detect any aldehydes even when high air flow rates and up to 24 hour sample collection times were used.
Table 5. Aldehyde Emissions TDI Foam Formulation.
Method (units)
(Standard Deviation) Type
HH
O
Formaldehyde
HO
Acetaldehyde
HO
Acrolein
1
GS97014-3 (µg/m
3
/kg)
Chamber
127
(19.1)
17
(12) 121
(136)
2
NES M0402
(µ
g/m
3
)
Bag method 99
(28.8)
15
(9.1) 0
3
ISO12219-3
(µ
g/m
3
)
Micro chamber 0 0 0
4
GMW15635 (µg/g)
Bag method
2.0
(0)
0.3
(0.1) <0.1
Sources of variability in the foam preparation were identified and eliminated, if possible, such as the density gradient in
molded foam, age and source of raw materials, potential striation in the air sampled from the Tedlar bag, and break through
on the DNPH column used to collect the aldehydes from the air sampled. The micro chamber method was also looked at more
closely. There are a number of papers on the use of the micro chamber for VOC and aldehyde emissions. Suggestions from
the papers such as the air flow rate and proximity of the sample collection tubes were worked through [8, 11, 12]. The key to
getting aldehyde emission measurements with the micro chamber was a measurement condition not previously reported.
Ensuring the sample was not touching the walls of the chamber allowed the air, which enters and exits the chamber through
the top, to circulate around the foam. Emission data with the optimized foam preparation and measurement technique using
the micro chamber is shown below in Table 6. Good detection limits and excellent reproducibility were seen. The optimized
sampling techniques gave similar improvements with the Tedlar bag and 1m
3
chamber methods.
Table 6. Aldehyde Emissions TDI Foam Formulation using Micro Chamber method.
ISO 12219-3
(µg/m
3
)
average 80.8
standard deviation 4.9
%RSD 6.2
Detection limits 9
Variables Impacting Aldehyde Emissions
To determine the impact of foam age on aldehyde emissions, foam was prepared with the MDI formulation shown in Table
3 and then allowed to age in a controlled temperature and humidity room. Table 7 shows the aldehyde emission results over
45 days. Before the foam is 30 days old the aldehyde levels do not change. After 30 days the formaldehyde and acetaldehyde
levels double perhaps because the antioxidants used in the polyols are consumed. At 45 days the aldehyde levels increased
further.
6
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
Table 7. Aldehyde Emissions over time using Micro Chamber method.
(
µ
g/m
3
)
Time
(days)
HH
O
Formaldehyde
HO
Acetaldehyde
HO
Acrolein
1
3
80
<48
ND
2 7
84 <48 ND
3 30
172 57 ND
4 45 191 77 ND
To understand the impact of aldehydes in the raw materials on foam emissions the total aldehyde content was measured on
the wet chemicals used in the MDI formulation in Table 3 and then the aldehyde emissions were measured on the foam
prepared with the same raw materials. The wet chemicals used to make the foam contained a total of 26.7 ppm of aldehydes
(line 1 in Table 8). Molded foam was made and allowed to cure in a controlled temperature humidity room for three days.
Emissions were measured using a tedlar bag and found to be 114 µg/m
3
, a four-fold increase in aldehydes.
To further explore the impact of raw material aldehyde content on emissions, formaldehyde was spiked into the polyol pre-
mix at 34,000 ppm and 70,000 ppm, line 2 and 3 in Table 8. The aldehyde emissions increased accordingly from the control
in both spiking examples indicating the level of aldehydes in the raw materials is an important factor in emissions. Although,
at the low levels the aldehydes are seen in polyurethane foam the raw material aldehyde content is not the main contributor to
emissions.
Polyurethane foam is porous and it was not known if foam could absorb aldehydes present in the air causing different
emission results in different regions of the world with different air quality. Typical background levels of formaldehyde in our
testing is 20 ug/m3. A 1,000 fold excess, 20,000 ppm, of formaldehyde in the form of formalin was added to a Tedlar bag
with a polyurethane foam sample in it. The high level of formaldehyde was used to ensure the emission response, if any, was
substantial. Care was taken to ensure the liquid formalin did not touch the piece of polyurethane foam. After two weeks of
exposure, aldehyde emissions were measured from the foam sample and the results are shown in line 4 of Table 8. The
aldehyde emissions were over 20,000 ppm indicating the foam environment impacts the aldehyde emissions. It is not known
if the same effect would occur at lower levels of ambient formaldehyde more closely resembling indoor air levels and the
result should not be extrapolated beyond the one data point.
Table 8. Total Aldehyde Emissions.
Environment Emissions
(µg/m
3
)
1 MDI Formulation (26.7 ppm aldehyde content in raw material) 114
2 34,000 ppm HCHO spike 28,345
3 70,000 ppm HCHO spike 74,220
4 20,000 ppm HCHO in sealed bag 22,320
Environmental factors influencing aldehyde emissions were looked at next. The impact of heat, UV light, and air were
measured using a three factor, two level, Design of Experiments (DOE). Each measurement was done once and a repeat foam
was run. The DOE is shown inTable 9. Foam aged with UV light or darkness was sealed in a tedlar or Mylar bag respectively,
and the bags were filled with either nitrogen or air. Heat aging was done in an oven.
7
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
Table 9. DOE Total Aldehyde Emissions.
Run X1: Irradiation X2: Atmosphere X3: Temperature
1 UV light Air 22
o
C
2 UV light Air 70
o
C
3 UV light Nitrogen 22
o
C
4 UV light Nitrogen 70
o
C
5 Dark Air 22
o
C
6 Dark Air 70
o
C
7 Dark Nitrogen 22
o
C
8 Dark Nitrogen 70
o
C
9 UV light Air 22
o
C
Emission measurements were done for formaldehyde, acetaldehyde, and acrolein and results for acetaldehyde emission are
shown below in Figure 1. UV light had the largest impact on aldehyde emissions. Foam stored in the dark had 150 µg/m
3
acetaldehyde content versus foam exposed to sunlight had over 1900 µg/m
3
. Temperature and air had the approximately the
same impact on acetaldehyde levels. Consistent with an oxidation mechanism for aldehyde formation, lower temperature and
an inert atmosphere caused lower acetaldehyde emissions. The large impact of light on acetaldehyde emissions may be due to
challenges in controlling the temperature of the light treated foam and the results with light and heat may be confounded.
Figure 1. DOE Response Graph
8
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
ADDITIVES TO LOWER ALDEHYDE EMISSIONS
Screening Methods
After understanding what variables impact aldehyde emissions and establishing a reproducible test method to measure
emissions, additives were looked at to lower aldehyde emissions. The initial screening was done using computational modeling
to determine if the tool could be a predictive technique to identify a solution to reduce aldehyde emissions from polyurethane
foam. Aldehyde emission testing is very time intensive and expensive process. Having a computational method to quickly
evaluate many chemical families is a valuable tool. Reactions were proposed between three additives belonging to different
families of chemistry, and formaldehyde and acrolein to understand from a fundamental perspective the effect of different
functional groups on aldehyde emissions. The thermodynamic feasibility of the proposed reactions were studied in order to
explore if the ability to reduce aldehyde emissions was related to the stability of the intermediate formed between the additive
and the aldehyde.
Computational and Experimental Aldehyde Emission Results of three Chemistry Families
Quantum chemical calculations and conventional statistical thermodynamics were performed using Accelrys Materials
Studio 7.0 software [14] to calculate reaction energetics at 0 K and Gibbs free energy of reaction in the gas phase. All
electronic energies were calculated using Density Functional Theory (DFT) by way of the BLYP/DNP/AE level of theory
with 4.0 A orbital cutoff. Partition functions based on the harmonic oscillator and rigid rotor approximations were used to
calculate thermodynamic properties as a function of temperature. Gas-phase calculations were performed as a proxy to the
real foam system to assess the qualitative ranking of reaction favorability and minimize demand on computational resources.
Reaction energetics at 0 K were calculated for relative reaction exothermicity / endothermicity and Gibbs free energies of
reaction were calculated for reaction favorability ranking. The computational modeling did not take into account the complex
environment in polyurethane chemistry and simply studied the reactive center in the presence of various functional groups
with formaldehyde and acetaldehyde. The computational results are then compared to the experimental results in Tables 10 -
12.
Using the MDI formulation detailed above in Table 3, additives from three families of chemistry were individually added
to the polyol, surfactant, and catalyst pre-mix and allowed to mix for 2 hours. Molded foam was prepared according to the
procedure detailed above. The foam was allowed to cure in a controlled temperature and humidity room for seven days then
tested according to NES M0402 (tedlar bag).
Table 10 shows the molecular modeling results of Additive 1 with formaldehyde and acetaldehyde. The gas-phase
calculations show mild exothermicity for both reactions, and the Gibbs free energy of reaction indicates the reaction with
formaldehyde is slightly more favorable than with acetaldehyde. The experimental foam results are shown below the
computational modeling results. The experimental results show a decrease in both formaldehyde and acetaldehyde
concentrations with an absolute reduction of formaldehyde by 130 µg/g versus 78 µg/g for acetaldehyde which is consistent
with the ∆G calculations despite the use of the simplified gas-phase model.
Table 10. Thermodynamic Calculations of Additive 1.
∆E
(kcal/mol) ∆G
(350K, kcal/mol)
Formaldehyde -3.9 -2.7
Acetaldehyde -1.5 -0.4
10L Tedlar Bag Method Results
Foam Sample Formaldehyde
(µg/g) Acetaldehyde
(µg/g)
Control 202 78
Additive 1 72 Non-detect
The computational and experimental results for Additive 2 with formaldehyde and acetaldehyde are shown in Table 11. The
reaction energetics show the reaction with formaldehyde is exothermic while the reaction with acetaldehyde is heat neutral.
The moderately positive ∆G for both reactions indicates low reaction favorability in the absolute sense, however, the
qualitative ranking ∆G matches the ranking of the experiments. The experimental results show that Additive 2 decreases
9
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
formaldehyde concentration very well but actually increases acetaldehyde concentration indicating a secondary reaction must
be occurring which was not considered in the computational modeling. The discrepancy between experimental and calculated
results is most likely due to the simplified model not taking into account the complex foam environment.
Table 11. Thermodynamic Calculations of Additive 2.
∆E
(kcal/mol) ∆G
(@ 350K, kcal/mol)
Formaldehyde -5.2 +13.7
Acetaldehyde 0.8 +19.7
10L Tedlar Bag Method Results
Foam Sample Formaldehyde
(µg/g) Acetaldehyde
(µg/g)
Control 202 78
Additive 2 Non-detect 168
The reaction of Additive 3 with formaldehyde and acetaldehyde was studied both experimentally and computationally. The
results are shown in Table 12. Gas-phase computational results for both reactions show moderate exothermicity with negative
reaction energetics. Overall the reactions are predicted to be unfavorable as both Gibbs free energies of reaction are positive;
however, the qualitative ranking of the Gibbs free energies of reaction match experimental rankings. Experimentally, the
formaldehyde was moderately reduced and acetaldehyde was unchanged which agrees with the computational results.
Table 12. Thermodynamic Calculations of Additive 3 with Aldehydes.
∆E
(kcal/mol) ∆G
(@ 350K, kcal/mol)
Formaldehyde -11.3 +5.8
Acetaldehyde -3.0 +14.9
10L Tedlar Bag Method Results
Foam Sample Formaldehyde
(µg/g) Acetaldehyde
(µg/g)
Control 68 43
Additive 3 20 49
Additives that Optimize a Reduction in Aldehyde Emissions
Based on the results above, families of chemistry that gave the most promising results were further studied. Additives with
functional groups that reduce formaldehyde and acetaldehyde were investigated further to optimize the reduction in aldehyde
emissions. The MDI formulation in Table 3 was used and the polyol, surfactant, and catalyst pre-mix was allowed to mix for
2 hours. The additives were further optimized to ensure the polyurethane physical properties were not negatively impacted
and that they are non-emissive by VDA 278.
The aldehyde emission results were grouped into three categories, additives that reduce formaldehyde but not acetaldehyde,
additives that reduce formaldehyde but increase acetaldehyde, and additives that decrease formaldehyde and acetaldehyde.
Acrolein was non-detect for all the experiments below. Results for additives that reduce formaldehyde but not acetaldehyde
are shown below in Figure 2. In agreement with the computational results above, formaldehyde was more readily reduced
than acetaldehyde giving a large number of additives that reduced only formaldehyde, leaving the acetaldehyde emissions
unchanged.
10
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
Figure 2. Formaldehyde and Acetaldehyde Emissions with Families of Chemistry that only Reduce Formaldehyde.
The second group of additives reduced formaldehyde but increased acetaldehyde emissions. Additive A reduced
formaldehyde while doubling the acetaldehyde emission. Additives O and P, caused acetaldehyde emissions to increase four
to five times higher than the control while reducing formaldehyde to non-detect levels (Figure 3).
0
10
20
30
40
50
60
70
80
90
100
Control B F G H I J
Reduces Formaldehyde Only (µg/m3)
Formaldehyde Acetaldehyde
Air Product’s Aldehyde
Emission Reduction
Target,
<30 µg/m
3
Decrease in Formaldehyde
emissions only. No change in
Acetaldehyde emissions
11
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
Figure 3. Families of Chemistry that reduce Formaldehyde but Increase Acetaldehyde Emissions.
The third group of additives significantly reduced formaldehyde and acetaldehyde. Due to the large range of aldehyde
emission specifications and test methods that exist for the different OEMs, an internal goal of reducing formaldehyde and
acetaldehyde below 30 µg/m
3
was set. Additives C, D, and E shown in Figure 4 achieved the goal. Experimental prototypes
were further developed to ensure good physical properties, wide process latitudes, and no contribution to foam emissions.
0
50
100
150
200
250
Control A O P
Reduces Formaldehyde and Increases Acetaldehyde Emissions
(µg/m3)
Formaldehyde Acetaldehyde
Up to a 5 fold
increase
in Acetaldehyde
Emission caused by Additives A, O, and P
Air Product’s Aldehyde
Emission Reduction
Target,
<30 µg/m
3
12
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
Figure 4. Different Families of Chemistry that Significantly Reduce Formaldehyde and Acetaldehyde Emissions.
ALDEHYDE SCAVENGER PROTOTYPE FOAM PROPERTIES
Individual scavenger additives were further optimized and three Aldehyde Scavenger Prototypes developed by examining
foam physical properties and emissions by VDA 278. Physical Properties were evaluated on the three Aldehyde Scavenger
Prototypes over a range of use levels using the MDI formulation shown in Table 3. Table 13 shows the tensile, tear, and
elongation with increasing amounts of Aldehyde Scavenger Prototype 1. The tensile and tear strength were unchanged as more
of Prototype 1 was used but the elongation was slightly lower versus the control. Both the ambient and humid aged
compression sets increased at the highest use level of Prototype 1. Aldehyde Scavenger Prototype 1 did not contribute to the
VOC or FOG emissions as measured by VDA 278 and at the highest use level reduced the total aldehyde emissions by 75%.
Air Product’s Aldehyde
Emission Reduction
Target,
<30 µg/m
3
13
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
Table 13. 41 kg/m
3
MDI Foam Physical Properties and Emissions Using Aldehyde Scavenger Prototype 1.
Control
Increasing Use Level of Aldehyde Scavenger Prototype 1
Tensile Strength (kPa)
(standard deviation) 86.02
(7.14) 77.84
(4.41) 74.38
(15.70) 81.24
(6.20)
Tear Strength (N/m)
(standard deviation) 200.8
(27.2) 200.6
(4.93) 205.2
(15.5) 218
(19.0)
Tensile Elongation (%)
(standard deviation) 66.15
(8.55) 52.92
(9.11) 59.33
(2.90) 58.38
(2.97)
Ambient Compression Set
(%)
(standard deviation)
8.86
(0.5) 8.03
(0.68) 9.13
(0.69) 10.18
(0.91)
Humid Aged Compression
Set (%)
(standard deviation)
19.26
(1.22) 20.03
(1.40) 21.85
(0.86) 23.32
(0.68)
Reduction in Total
Aldehyde Emissions Not Applicable 52% 61% 75%
Contribution of Aldehyde
Scavenger to Emissions by
VDA 278 (µg/g) Not Applicable 0 0 0
Physical Properties of molded foam made with Aldehyde Scavenger Prototype 2 are shown in Table 14. The tensile
improved slightly as more of Prototype 2 was used while the tear strength and elongation were unchanged from the control
with different amounts of Prototype 2. The ambient and humid aged compression sets were also unchanged with the addition
of Prototype 2. Aldehyde Scavenger Prototype 2 did not contribute to the VOC or FOG emissions as measured by VDA 278
and at the highest use level reduced the total aldehyde emissions by 56%.
14
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
Table 14. 41 kg/m
3
MDI Foam Physical Properties and Emissions Using Aldehyde Scavenger Prototype 2.
Control
Increasing Use Level of Aldehyde Scavenger Prototype 2
Tensile Strength (kPa)
(standard deviation) 86.02
(7.14) 79.93
(7.25) 83.40
(6.62) 95.20
(7.64)
Tear Strength (N/m)
(standard deviation) 200.8
(27.2) 189.5
(15.4) 205.6
(12.7) 187.2
(18.9)
Tensile Elongation (%)
(standard deviation) 66.15
(8.55) 56.90
(5.02) 61.05
(5.14) 67.53
(4.37)
Ambient Compression Set
(%)
(standard deviation)
8.86
(0.50) 8.77
(0.80) 7.72
(0.33) 8.15
(0.80)
Humid Aged Compression
Set (%)
(standard deviation)
19.26
(1.22) 20.03
(1.40) 19.78
(0.93) 19.42
(1.14)
Reduction in Total
Aldehyde Emissions Not Applicable 43% 53% 56%
Contribution of Aldehyde
Scavenger to Emissions by
VDA 278 (µg/g) Not Applicable 0 0 0
Physical Properties of Aldehyde Scavenger Prototype 3 in molded foam are shown in Table 15. Tensile and tear strength
are unchanged over a range of use levels for Prototype 3. The elongation increased versus the control as more of Prototype 3
was used. The ambient and humid aged compression sets were unchanged with the addition of Prototype 3. Aldehyde
Scavenger Prototype 3 did not contribute to the VOC or FOG emissions as measured by VDA 278 and at the highest use level
reduced the total aldehyde emissions by 77%.
15
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
Table 15. 41 kg/m
3
MDI Foam Physical Properties and Emissions Using Aldehyde Scavenger Prototype 3.
Control
Increasing Use Level of Aldehyde Scavenger Prototype 3
Tensile Strength (kPa) 86.02
(7.14) 85.39
(5.82) 85.24
(4.93) 81.64
(4.46)
Tear Strength (N/m) 200.8
(27.2) 214
(13.97) 211
(12.39) 199.9
(19.00)
Tensile Elongation (%) 66.15
(8.55) 79.0
(2.30) 71.86
(7.09) 80.78
(4.14)
Ambient Compression Set
(%) 8.86
(0.50) 7.75
(0.33) 7.78
(0.50) 7.42
(0.47)
Humid Aged Compression
Set (%) 19.26
(1.22) 21.79
(0.57) 22.17
(0.79) 21.55
(1.05)
Reduction in Total
Aldehyde Emissions Not Applicable 55% 62% 77%
Contribution of Aldehyde
Scavenger to Emissions by
VDA 278 (µg/g) Not Applicable 0 0 0
CONCLUSION
Polyurethane foam sold into the automotive industry increasingly needs to pass stringent VOC and aldehyde emission tests.
Air Products is committed to offering products that enable customers to meet the most stringent requirements. The aldehyde
content of the raw materials contributed to foam emissions to some extent but the post treatment and age of the foam were
found to have the largest impact. Consistent with an oxidative mechanism, aldehyde emissions increased when polyurethane
foam was exposed to light, heat, and air.
Computational modeling proved to be a good screening tool for identifying chemistry families to lower aldehyde emissions.
The tool proved valuable to quickly screen many chemical families quickly and inexpensively.
Combining the learnings about aldehyde emission sources as well as the reactivity of different chemistry families allowed
three new Aldehyde Scavenger Prototypes to be developed. The Prototypes not only reduce aldehyde emissions up to 77%
but they are non-emissive by VDA 278 and maintain the excellent physical properties of the polyurethane. Currently, the
Aldehyde Scavenger Prototypes are being scaled up for line trial evaluations with the hope of bringing a solution to the
industry based the results.
ACKNOWLEDGEMENTS
The authors of this paper would like to extend thanks to the personnel in the Polyurethane Additives Group and Performance
Materials Division at Air Products. Special thanks are given to Juan Burdeniuc, Stephan Wendel, Emmanuelle Le Grand,
Agnes Derecskei, and Jane Kniss from Air Products as well as Jenan Elias, Todd McEvoy, and Dawn Enstrom from Intertek.
The information in this paper is provided for discussion purposes only and does not make any representation or warranty
with respect to the information. Air Products disclaims all warranties, conditions, or representations of any kind, express or
implied. No patent should be practiced without authorization from the patent owner.
Dabco is a trademarks of Air Products and Chemicals, Inc.
16
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
REFERENCES
1. Muller, D.; Klingelhofer, D.; Uibel, S.; Groneber, A. D. J. Occupational Medicine and Toxicology 2011, 6, 33.
2. Bruinen de Bruin, Y.; Koistinen, K.; Kephalopoulos, S.; Geiss, O.; Tirendi, S.; Kotzias, D. Environ Sci Pollut Res
2008, 15, 417.
3. Harada, K.; Hasegawa, A.; Wei, C-N.; Minamoto, K.; Noguchi, Y.; Hara, K.; Matsushita, O.; Noda, K.; Ueda, A.
Journal of Health Science, 2010, 56, 488.
4. Salthammer, T.; Mentese, S.; Marutzky, R. Chem. Rev. 2010, 110, 2536.
5. Makhlouf, K. “Nissan’s Action for Cabin VOC.”
6. GB/T 27630-2011 “Guidelines for air quality assessment of passenger car”
7. “Indoor air quality standards” GB/T 18883-2002
8. Commission for the Investigation of Health Hazards of Chemical Compounds in the Work Area. List of MAK and
BAT Values for 2012. 2012, 48, 1.
9. GMW 15635 and GMW 3059. Material Specifications, Restricted and Reportable Substances, GM Worldwide
Engineering Standards, 2014.
10. ISO 12219-3, “Interior air of road vehicles – Part 3: Screening method for the determination of the emissions of
volatile organic compounds from vehicle interior parts and materials – Micro-scale chamber method.” First edition
2012-06-15.
11. Nissan Engineering Standard, NES M0402, VOC Test Method of Cabin Parts, 2012.
12. BMW Group Standard, GS 97014-3, version 2001-04. “Emissions measurement with air exchange in a testing
chamber.”
13. Schripp, T.; Nachtwey, B.; Toelke, J.; Salthammer, T.; Uhde, E.; Wensing, M.; Bahadir, M. Anal Bioanal Chem,
2007, 387, 1907.
14. Herrington, R.; Hock, K. “Dow Polyurethanes Flexible Foams” 1997, 11.16.
15. Accelrys Software Inc., Materials Studio Modeling Environment, Release 7.0, San Diego: Accelrys Software Inc.,
2014.
17
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
BIOGRAPHIES
Jennifer E. A. Al-Rashid
Jennifer Al-Rashid received her B.S. in Chemistry from Alma College in 2003 and her Ph.D. in
Organic Chemistry from the University of Wisconsin, Madison in 2008. Jenny then joined Air Products
and has worked in Product R&D and Process R&D for Performance Materials and Electronic
Materials. Jenny currently has responsibilities for application development and technical service
around catalysts and surfactants for the flexible molded and CASE market segments.
Torsten Panitzch
Torsten Panitzsch received his diploma in Chemistry in 1994 and a Ph.D. in Organic Chemistry in
1998 from the University of Kiel, Germany. He joined Air Products and Chemicals GmbH in 1999.
Torsten worked several years as an application chemist responsible for rigid and shoe sole application
development and technical service. Currently he is responsible for application development and
technical service for catalysts and surfactants with emphasis on the flexible molded market segment.
Jackson Su
Jackson Su received his Master degree in Physical chemistry from Chongqing University in 1999.
He joined BASF China as an Application Development / Technical Service Chemist for
polyurethane flexible molded, shoe sole and case applications in 2000. He joined Air Products as
an Application Development / Technical Service Chemist supporting the flexible foam market in
2006. He is now the Asia flexible molded chemist.
Gauri S. Lal
Gauri Sankar Lal received a PhD in Organic Chemistry in 1985 from the University of New Brunswick, Canada. After two
years as a post-doctoral researcher at Glaxo Smith Kline Pharmaceutical and two years as an Assistant Professor of Organic
Chemistry at Drexel University, Philadelphia he joined Air Products in 1989. At Air Products Sankar worked on the
development of new organic fluorine containing compounds for medicinal and electronic applications. In addition he
conducted research on the synthesis of novel bio-degradable surfactants, epoxy curing agents and polyurethane catalysts.
Andrew J. Adamczyk
Andrew J. Adamczyk is a chemical engineer working in the Computational Modeling Center at Air
Products and Chemicals, Inc. He received his Ph.D. from Northwestern University and B.S. from
the Illinois Institute of Technology. Before Air Products, Andrew spent time abroad as a research
fellow at the Laboratory for Chemical Technology at Ghent University in Belgium. Andrew was
also a postdoctoral fellow at the University of South California (with Nobel Laureate Arieh
Warshel) and the Massachusetts Institute of Technology (MIT).
18
©2015 Air Products and Chemicals, Inc. ©2015 American Chemistry Council
Allen R. Arnold Al Arnold is currently working as a Research Technician for Performance Materials Division supporting
applications development and technical service for both flexible slabstock and the flexible molded market
segment. He received his degree in Medical Laboratory Technology from NCACC, Northampton,
Pennsylvania in 1981. He joined Air Products in 1990.
