Sei sulla pagina 1di 23

12/19/2019 Gasoline Particulate Filters

DieselNet Technology Guide

DieselNet.com. Copyright © Ecopoint Inc. Revision 2019.05

Gasoline Particulate Filters


W. Addy Majewski

Abstract: Gasoline particulate filters (GPF) have been introduced to reduce particle number emissions from GDI vehicles. The filters utilize wall-flow substrates first
developed for diesel particulate filters. The GPF regenerates passively, but an active regeneration assist is needed to prevent filter plugging during low temperature duty
cycles. Ash has an impact on GPF performance and—if the GPF is coated with a three-way catalyst—can be a source of catalyst poisoning.

Introduction
Materials and Configuration
Emissions and Engine Performance
GPF Operation and Control

1. Introduction

Gasoline particulate filters (GPF) are an emission aftertreatment technology based on diesel particulate filters (DPF), developed to control particulate emissions
from gasoline direct injection (GDI) engines. The technology is also referred to as petrol particulate filter (PPF) and, in some German literature, as Otto particle
filter (Ottopartikelfilter in German), abbreviated OPF.

The population of GDI vehicles has been increasing, driven by CO2 and/or fuel economy requirements. In 2016, an estimated 2/3 of new gasoline cars in
Europe were GDI [JM 2017]. The proportion of GDI vehicles has been also rapidly increasing in North America—within nine years after its first significant use
in the market, GDI penetration has climbed to 48.5% of new light vehicle sales in the United States [Davis 2017]. Emissions from the growing GDI vehicle fleet
are a public health concern and a potential major source of ambient particle pollution in highly populated urban areas.

GPFs are expected to be used primarily in the European Union and in China, to meet the particle number (PN) emission standards for gasoline passenger cars
and light commercial vehicles adopted in both jurisdictions. The Euro 6 regulations set PN (as well as PM) limits for GDI vehicles that are equivalent to those
https://www.dieselnet.com/tech/gasoline_particulate_filters.php 1/23
12/19/2019 Gasoline Particulate Filters

for diesels. The European PN standards, both effective for new types of GDI cars from September 2017, are:

A PN limit of 6.0×1011 km-1 over the NEDC / WLTC test cycle (Euro 6c)
RDE testing for PN emissions with a conformity factor of 1.5, i.e., PNRDE = 9.0×1011 km-1 (Euro 6d-TEMP)

China 6 regulations also include a WLTC PN emission standard of 6.0×1011 km-1, effective from July 2020, as well as RDE PN requirements from July 2023.
The Chinese PN standards are not limited to GDI, but apply to all gasoline vehicles.

The above standards could also be met—at least in certain types of vehicles—via in-cylinder controls such as fuel injection strategies, without particulate
filters. However, the GPF has several advantages compared to in-cylinder controls:

Effectiveness under all operating conditions—While in-cylinder strategies tend to be more effective under certain modes of operation, the GPF provides
PN emission control under all engine operating conditions—an advantage that is especially important in RDE testing.
Control of emissions from engine faults—Increased PN emissions can occur as a result of engine faults and malfunction, such as increased lube oil
consumption. These emissions can be effectively controlled by particulate filters [Czerwiński 2017].
Control of unregulated emissions—The GPF can control certain unregulated emissions, including polynuclear aromatic hydrocarbons (PAH). GDI
engines, even though equipped with three-way catalysts (TWC), may produce significant levels of toxic PAH emissions [Muñoz 2017].

Gasoline particulate filters are not expected to be widely adopted in North America, where particle emissions are regulated through mass-based PM limits only.
The US Tier 3 PM limit of 3 mg/mi, as well as the 2025 California LEV III limit of 1 mg/mi will likely be met through in-cylinder control technologies.

Commercial Status. Gasoline particle filters were first launched in a mass production application by Daimler, who introduced a GPF on their Mercedes-Benz
S500 luxury sedan in early 2014 [Lanzerath 2017]. The number of GPF applications has increased rapidly since 2017, as a result of the PN RDE testing
requirements that became effective at the Euro 6d-TEMP stage. Filters were introduced on additional models by Daimler, as well as by Volkswagen, BMW,
Peugeot and other manufacturers. By mid-2018, one GPF manufacturer—Corning—supplied one million GPFs for the European market, which indicates that
within about one year the technology reached a market penetration of at least 10% of gasoline vehicles. GPFs may be also adopted for some port fuel injected
(PFI) engines, even though PFI vehicles are not subject to European PN/PM emission standards.

Most early GPF applications included an uncoated GPF positioned downstream of a TWC catalyst. As the technology matured, GPFs have been also coated
with a three-way catalyst. This catalyst coated GPF configuration is sometimes referred to as the 4-way catalyst. One of the early applications of a coated GPF
was the 1.0 TSI engine used in the 2018 VW up! GTI city car.

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 2/23
12/19/2019 Gasoline Particulate Filters

While the GPF and DPF technologies are closely related, there are a number of differences in the filter configuration, operation and control strategy, which are
due to the differences in the operating conditions, and the particulate emission rates and composition between gasoline and diesel engines. These aspects of
GPF technology are discussed in the balance of this paper. An in-depth discussion of strategies and technologies to reduce PM emissions from gasoline engines
can be found in the literature [Boger 2018].

2. Materials and Configuration

2.1 GPF Substrates


GPFs utilize cellular wall-flow monoliths first developed for diesel particulate filters. Because of less frequent and lower peak temperature regeneration events
required in GDI engines with a GPF compared to diesels with a DPF, cordierite substrates can be used for GPFs. In contrast, filters for light-duty diesel engines
typically use substrates made of materials of higher thermal durability such as silicon carbide.

The filtration efficiency of the GPF depends on the cellular parameters and material properties of the substrate. GPF substrates are designed to provide PN
filtration efficiency on the order of 70-80%, which is generally sufficient for meeting the applicable PN emission standards, while minimizing the GPF pressure
drop. However, meeting the applicable PN limits can be challenging with fresh filters (no soot or ash deposits) in RDE testing with hard accelerations and
transient driving, and may require a zero-mileage GPF efficiency of above 80% [Yoshioka 2019], which can be achieved through substrate optimization. It was
also shown that fresh filters pre-loaded with a small amount of thermally stable submicron alumina particles (“artificial ash”) were able to achieve ~90% PN
filtration efficiency with only a small increase in pressure drop [Liu 2019].

Low pressure drop has been one of the key GPF design targets, due to its potential adverse effect on high load operation and the rated engine power. In general,
in contrast to the diesel engine, exhaust backpressure has little effect on fuel consumption in the throttled gasoline engine, unless the engine is operated at full
load—a condition rarely encountered in light-duty vehicles. However, GDI engines tend to be more sensitive to backpressure than port-injected engines. This is
further discussed below.

Among the cellular properties of the filter substrate, the wall thickness has a strong impact on filtration efficiency, Figure 1. The substrate cell density, on the
other hand, has a relatively small impact on filtration efficiency [Kattouah 2014].

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 3/23
12/19/2019 Gasoline Particulate Filters
100 100

PN Filtration Efficiency, %

PN Filtration Efficiency, %
Cell density 300 cpsi Wall thickness 12 mil

90 90

80 80

70 70

60 60
6 8 10 12 14 16 200 250 300 350 400
Wall Thickness, mil Cell Density, cpsi

Figure 1. Impact of wall thickness and cell density on PN filtration efficiency


NEDC cycle, 1.4 L GDI engine, uncatalyzed GPF, φ118.4×127 mm, mean pore size 15-25 µm, porosity 65%

Important material properties that affect filtration efficiency include material porosity and the mean pore size (MPS). Lower MPS increases the GPF filtration
efficiency, while higher porosity decreases the pressure drop and allows to accommodate higher washcoat loads in coated filters. Materials of 48% porosity
(Figure 2) have been used for uncatalyzed GPFs, while higher porosities such as 65% are intended for catalyzed filters. As apparent from Figure 2, materials of
higher porosity experience a smaller pressure drop increase from the addition of the catalyst washcoat [Kattouah 2014]. Cell densities of 200 cpsi (8 mil/200
cpsi) have been suggested for uncoated GPFs, while 300 cpsi substrates were more suitable for catalyst coating [Yoshioka 2019]. The higher cell density in
catalyzed filters is desired to enhance the TWC catalyst performance.

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 4/23
12/19/2019 Gasoline Particulate Filters
5.0

4.0 +25%

Pressure Drop, kPa


+16%
+7%
3.0

2.0

Catalyzed

Catalyzed

Catalyzed
Bare

Bare

Bare
1.0

0.0
48 59 65
Wall Porosity, %

Figure 2. GPF pressure drop as a function of porosity


Cold flow bench, 6 Nm3/min, 25°C; GPF 12 mil/300 cpsi, φ118.4×127 mm; Equal washcoat loading.

The increase of GPF pressure drop due to catalyst coating presents a challenge for the application of TWC-coated filters [Inoda 2017]. To achieve low tailpipe
NOx and HC emissions, high loadings of the catalyst washcoat are required, which can result in unacceptably high pressure drop over the coated GPF.

For higher catalyst washcoat loads—in excess of 100 g/L—the material’s MPS increasingly affects the GPF pressure drop. The pressure drop significantly
increases for MPS below 15 μm [Kattouah 2014].

2.2 Catalyst Coating


The catalyst coating in catalyzed GPFs is typically based on three-way catalyst technology using a Pd+Rh catalyst and ceria-based oxygen storage capability
(OSC) [JM 2017]. Catalyst coated GPFs may eventually evolve to provide improved TWC performance—matching that of conventional three-way catalysts—
while maintaining an acceptable level of pressure drop and the required PN filtration efficiency. This would enable a wider use of catalyzed GPFs and more
compact emission systems.

GPF coating can also enhance soot oxidation, due to the presence of a PGM catalyst and/or the OSC component of the washcoat. The latter can promote soot
oxidation under rich conditions using the stored oxygen. Powder reactor studies confirmed that the presence of PGM and ceria-containing OSC materials can
reduce soot combustion temperatures and that the OSC material can influence peak soot combustion temperatures by more than 50ºC [Morgan 2015]. Ceria-
containing OSC washcoat without a PGM catalyst can also be used to promote soot oxidation in the GPF, but appears unusual in production vehicles.

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 5/23
12/19/2019 Gasoline Particulate Filters

The GPF catalyst coating technology can be optimized to control pressure drop and enhance catalyst performance. For instance, by using zone coating, the
precious metal catalyst may be preferentially located close to the GPF inlet, aiming at early warm-up. By changing the rheology and the particle size of the
washcoat slurry, it is also possible to place more catalyst washcoat inside the wall pore network, or on the wall inside the filter channels, Figure 3 [Xia 2019].
Deep coating into the wall porosity was reported to help maintain larger pore diameters and enlarge the flow route of the gas; on the other hand, coating on-the-
wall improved the GPF filtration efficiency [Inoda 2017]. Another study found that GPF pore network structures with a higher number of narrow paths for the
gas flow produced lower pressure drop. A coating process was developed to make the washcoat penetrate into the wall and increase the path numbers [Tanaka
2018].

Figure 3. SEM photo of GPF coating on-the-wall (left) and in-the-pores (right)
(Source: SPM Catalyst)

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 6/23
12/19/2019 Gasoline Particulate Filters

Catalyst poisoning is another potential issue with catalyst coated filters. The catalyst would remain in contact with the ash accumulating in the GPF, which is
derived primarily from the engine lube oil additive package. Several of the lube oil ash components have been known TWC catalyst poisons.

2.3 GPF System Configuration


GPFs are installed on GDI engines that are normally equipped with three-way catalysts. Hence, there are two options to incorporate the GPF:

Uncatalyzed GPF—An uncoated GPF is added downstream of the TWC catalyst.


Catalyzed GPF—A GPF coated with a three-way catalyst can either replace the TWC or be used in series with a TWC converter.

Uncatalyzed GPF. Several configurations are possible, including a close-coupled TWC+GPF, or a GPF in an underfloor location. The underfloor location can
provide more flexibility with the installation space, but the GPF is exposed to lower exhaust gas temperatures which makes it more difficult to regenerate. In
the close-coupled configuration, the TWC and GPF can be packaged in one housing, as shown in Figure 4. Alternatively, the TWC and the GPF can be
packaged separately, with the GPF installed immediately downstream of the TWC [Van Nieuwstadt 2017].

Switching O2 Pressure
sensor sensor
Lambda Temperature
sensor sensor

TWC GPF

Figure 4. Catalytic converter containing a TWC and a GPF

An example of a GPF in the underfloor location is shown in Figure 5, which depicts the exhaust system of the 2014 Mercedes-Benz S500, the first mass
produced vehicle with a GPF. The S500, powered by the 8-cylinder M278 Mercedes GDI engine, utilizes two uncatalyzed particulate filters, one in each
exhaust bank [Lanzerath 2017]. A GPF substrate of 48% porosity and 6 mil wall thickness was selected to meet the engineering targets for the filter pressure
drop and a 70% PN filtration efficiency. The GPFs have a 4.33" (110 mm) diameter and a 1.5 L volume each. The total GPF volume of 2-3 L per vehicle was
required to retain a sufficient ash holding capacity with increasing mileage. The underfloor GPF location was chosen to allow installation in the existing vehicle
model.

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 7/23
12/19/2019 Gasoline Particulate Filters

Figure 5. Twin GPF in the exhaust system of 2014 Mercedes-Benz S500


(Source: Daimler)

Catalyst Coated GPF. The catalyst coated GPF (cGPF) has advantages in terms of installation space and cost—particularly if the catalyzed GPF can simply
replace the TWC converter. However, if a coated GPF is used, the washcoat load must be relatively low to ensure low GPF pressure drop. If the catalyst load on
the GPF is not sufficient for meeting the applicable emission limits for gaseous pollutants, another flow-through TWC must be added in the emission system. A
number of configurations are possible—three examples are compared in Table 1 [Tang 2019].

Table 1
Example system configurations with catalyzed GPF

Configuration Advantages Disadvantages


Lower thermal stress
Saves space in engine compartment
TWC cGPF Soot regeneration more challenging
Less catalyst required
Lower pressure drop
Requires space in engine compartment
Increased thermal stress
TWC cGPF Soot regeneration less challenging More catalyst required to accommodate higher thermal
stress
Higher pressure drop
Very high catalyst loading
cGPF Soot regeneration least challenging Highest pressure drop
Light-off activity of TWC poorer due to lower cell density

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 8/23
12/19/2019 Gasoline Particulate Filters

The last configuration in Table 1, a single TWC-coated GPF, remains uncommon due to the technical challenges shown in the table. Two-stage configurations
with a coated GPF and a flow-through TWC in series are typical for commercial applications of coated GPFs. In these two-stage systems, the TWC can be
positioned upstream (as shown in Table 1) or downstream of the cGPF. The upstream cGPF location has the disadvantage of a higher catalyst light-off
temperature due to the lower cell density—a GPF would typically have a cell density of 200-300 cpsi, while the TWC could have a cell density of 600 cpsi or
higher.

Therefore, configurations with TWC in the upstream position are preferred for applications with more stringent NOx emission limits, such as China 6 [Tang
2019]. On the other hand, if NOx limits can be met with the cGPF positioned upstream, this configuration can provide a more reliable cGPF regeneration. One
of the first coated GPF applications—the 2018 EU market VW up! GTI 1.0 TSI engine—utilized a close-coupled cGPF followed by an underfloor TWC that
provided a sufficient overall TWC catalyst loading to meet gaseous emission limits, even under heavy loads.

3. Emissions and Engine Performance

3.1 Pollutant Emissions


The effect of a GPF on particle emissions can be quantified by the measurement of particle number (PN) or particulate mass (PM) emissions. The European PN
standards are expressed in terms of numbers of solid particles above 23 nm, according to the PMP procedure. This is the obvious metric to determine GPF
performance in the context of Euro 6 and related emission regulations. The measurement of particle mass, on the other hand, could be more appropriate in the
context of US and California emission standards. Some authors neglect to specify their method of measurement, which makes their reports confusing.

Figure 6 shows the PN filtration efficiency of the Mercedes S500 with GPF [Lanzerath 2017]. The vehicle was tested over the distance of 400 km by driving
repeated NEDC cycles. The filtration efficiency of the particulate filters increased from 73% to 81% due to the deposition of soot and ash over the distance.

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 9/23
12/19/2019 Gasoline Particulate Filters
100

90

Filtration Efficiency, %
80

70

60

50
0 50 100 150 200 250 300 350 400 450
Distance, km
Figure 6. PN filtration efficiency as a function of distance driven
Repeated NEDC cycles

Since particle emissions in GDI engines occur mostly at the cold start and at very high load conditions, the PN emission levels as well as the GPF performance
depend on the drive cycle. Figure 7 compares PN emissions of a GDI vehicle without and with a GPF, as determined over three test cycles [Kato 2014]. In the
NEDC test, the vehicle emissions without GPF were already below the regulatory limit. However, PN emissions exceeded the Euro 6c limit when the NEDC
test was run at a cold temperature, as well as over the WLTC test that is characterized by a higher engine load. In all cases, the GPF reduced PN emissions to
well below the regulatory limit.

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 10/23
12/19/2019 Gasoline Particulate Filters
1.0e+13

no GPF
with GPF

Particle Number, 1/km


Euro 6c limit
1.0e+12

1.0e+11

1.0e+10
NEDC -7°C NEDC WLTC
Figure 7. Effect of GPF on particle number emissions
1.4 L and 1.8 L GDI engine, λ=1, uncatalyzed GPF, φ118.4×127 mm

The effect of a GPF on particle size distribution is shown in Figure 8 [Kato 2014]. The tests were conducted over the WLTC and the NEDC cycles, with the
WLTC producing higher engine-out particle numbers. With the GPF, PN emissions were reduced over the whole range of particle size. The post-GPF particle
size distributions over the WLTC and NEDC were similar.

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 11/23
12/19/2019 Gasoline Particulate Filters
1.6e+05

1.4e+05
WLTC w/o GPF
1.2e+05 NEDC w/o GPF

dN/d(logDp), 1/cm3
1.0e+05 WLTC w/ GPF
NEDC w/ GPF
8.0e+04

6.0e+04

4.0e+04

2.0e+04

0.0e+04
10 100 1000
Particle Diameter, nm

Figure 8. Effect of GPF on particle size distribution


1.8 L GDI engine, λ=1, uncatalyzed GPF, φ118.4×127 mm

The results in Figure 8, especially those for the NEDC cycle, suggest that the GPF was more efficient for smaller size particles. A similar behavior was seen in
another study that compared GPF efficiency using the PMP method with two different particle diameter cut-off points: 23 nm and 7 nm. The GPF PN reduction
efficiency ranged from about 60% to 80% for particles above 23 nm, while for particles above 7 nm, the GPF efficiency reached the range from 70% to more
than 90% [Andersson 2017].

Polynuclear Aromatic Hydrocarbons. There is some evidence that GPFs can reduce PAH emissions from GDI engines. In a study with a Euro 5 GDI vehicle,
four GPFs were tested, including two catalyzed and two uncatalyzed units [Muñoz 2017]. The GPFs reduced the PAH toxicity equivalent concentrations (TEC)
by 20-80%. It is unclear whether the TWC catalyst coated over two of the GPFs played a role in the PAH conversion. Even with the GPF, the PAH
concentrations from the GDI vehicle were still higher than those from a reference Euro 5 diesel car equipped with a DOC + DPF emission system.

In another study, two 2016 GDI vehicles (a 2.0 L LEV III SULEV30 and a 1.5 L LEV II) were tested in their original configuration and with a catalyzed GPF
replacing the underfloor TWC (the original close-coupled TWCs were retained) [Yang 2018]. The GPFs were catalyzed with precious metal loadings typical of
underfloor catalysts, matching the certification levels of the two vehicles. Both GPFs were φ4.66”×4.5”, with 8 mil/300 cpsi cells, catalyzed with a TWC at 1.0
g/L metal loading (Pd:Rh 4:1). Emissions of PAH and nitrated PAH species were quantified in both the vapor and particle phases of the PM. In emissions from
the stock vehicles, 2–3 ring compounds and heavier 5–6 ring compounds were observed in the PM, whereas the vapor phase was dominated mostly by 2–3 ring
aromatic compounds. In tests with the GPF-equipped vehicles, the total concentration of PAH species from the two vehicles was reduced by 97% and 99%,
https://www.dieselnet.com/tech/gasoline_particulate_filters.php 12/23
12/19/2019 Gasoline Particulate Filters

respectively. The concentration of vapor-phase PAH was reduced between 84% and 35%—depending on the PAH category and the vehicle—suggesting that
vapor-phase PAHs are not eliminated as effectively over the catalyzed GPF as those associated with the particle phase. Emissions of total nitro-PAHs from the
two vehicles were reduced by 91% and 77%, respectively.

3.2 CO2 & Fuel Economy


The GPF pressure drop contributes to the exhaust backpressure, which can have adverse effect on engine operation, fuel consumption and CO2 emissions—
especially at high load [Lee 2013]. GDI engines can be especially sensitive to increased backpressure due to the need to ensure scavenging of residual gases.

At part load operation, there may be little pressure drop difference between an optimized GPF and a TWC, and thus little impact on fuel consumption and CO2
emissions over regulatory drive cycles dominated by part-load operation [Chan 2016][Yang 2018]. However, high load operation similar to that represented by the
Common Artemis Driving Cycle (CADC) in Figure 9 [Kattouah 2013], where the pressure drop can become a significant consideration, is often encountered in
real world driving and RDE testing with downsized GDI engines.

Original system With GPF

200
120 km/h 130 km/h 160 km/h
180

160

140

120
CO2 Emissions, g/km

100

80

60

40

20

NEDC WL TC Artemis

Figure 9. The effect of GPF on CO2 emissions over different test cycles
Vehicle: 1.4 L stoichiometric GDI; GPF: 48% porosity, φ118.4×127 mm, 6 mil/220 cpsi, uncatalyzed; underfloor position; Maximum speed for each cycle as indicated

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 13/23
12/19/2019 Gasoline Particulate Filters

4. GPF Operation and Control

4.1 Overview
The GPF operation resembles that of the diesel particulate filter. The filter accumulates particulate matter, with the initial accumulation occurring within the
pore network in the substrate walls (depth filtration in Figure 10). Once the pore volume becomes filled, a filtration cake forms over the surface of the filter
channels. The pressure drop build-up due to soot accumulation in the GPF, shown in Figure 10, is very similar to that seen in diesel particulate filters [Lanzerath
2017]. With the operating time, lube oil ash becomes deposited in the pores of the GPF substrate, which reduces the role of depth filtration and favors surface
filtration.

Depth
Surface filtration
filtration
GPF Pressure Drop, kPa
4

0
0 2 4 6 8 10 12
Time, h
Figure 10. GPF pressure drop during soot loading
Soot loading of a clean GPF on the Mercedes S500 M278 engine
Soot loading: exhaust gas flow 200 kg/h, GPF inlet temperature 450°C
ΔP measurement: gas flow 650 kg/h, temperature 750°C

The soot accumulated in the GPF must be oxidized in a process known as filter regeneration. The regeneration can be either passive (occurring during regular
operation of the engine) or active (triggered by purposely increasing the exhaust temperature). The regeneration also requires a sufficient oxygen concentration
in the exhaust gas—a condition typical for the diesel engine, but not necessarily so in GDI engines.

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 14/23
12/19/2019 Gasoline Particulate Filters

A major difference between diesel and gasoline emissions is that gasoline exhaust includes much less carbonaceous particulates (soot) than diesel exhaust. This
has important implications for the GPF operation. The lower soot content, in combination with higher exhaust temperatures, make gasoline particulate filters
easier to regenerate.

Most GPFs regenerate passively, using excess oxygen that is available during fuel shut-off events. A regeneration strategy may be added to address soot
accumulation during challenging, low temperature drive cycles. The typical maximum soot load in the GPF is 1.5-2 g/L [Van Nieuwstadt 2017]. In contrast, light-
duty diesels rely predominantly on active DPF regeneration, with the typical maximum soot load of 5-10 g/L.

4.2 Ash Accumulation


Since GDI emissions include less soot than diesel emissions, ash accumulation can be relatively more important in the GPF operation. A number of GPF
studies have focused on ash accumulation, examining ash impacts on GPF performance and pressure drop, the oxidation of soot in the GPF [Choi 2016], as well
as ash effects on the filter-coated TWC catalyst.

Ash loading in the GPF was examined in a study by Ford [Lambert 2017]. A number of (catalyzed) filters were road, dyno, and/or burner-aged, using fuel
dopants including ZDDP, calcium sulfonate and organophosphates. In addition, artificial ash was used.

An interesting finding in the Ford study was that very low ash loadings of approximately 1-2 g significantly improved the filtration efficiency. An increase in
efficiency from about 60% to 80% was observed at a relatively small increase of pressure drop from about 15 kPa to 20 kPa (600 kg/h, 750°C). The increase in
efficiency was attributed to bridging the filter pores by ash. Figure 11 shows SEM images showing submicron dendritic ash structures on the wall surface of
samples from two GPFs (3k1 and 3k2). Both filters were road aged over 3,000 km.

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 15/23
12/19/2019 Gasoline Particulate Filters

Figure 11. Ash particles bridging GPF pores


(Courtesy of Ford Motor Company)

The study confirmed that ash penetrates into the GPF wall over the life of the filter. Considering the passive regeneration mechanism of the GPF, this is
consistent with the ash deposition patterns in the DPF, where ash in passively regenerated filters tends to migrate into the filter walls. In contrast, more ash is
collected at the plugged channel ends in actively regenerated diesel filters.

Ash accumulation in the GPF pores may have unexpected beneficial effects on the pressure drop. An examination of aged (3,000 km) and fresh coated filters
showed that filters with ash showed lower pressure drop than fresh filters, due to the cake filtration mechanism in ash-loaded filters, which produces less
pressure drop than depth filtration in the pore network [Liu 2018].

To study GPF ash morphology, cross sections of filter samples were prepared using the argon ion milling (CSIM) and gallium ion source (FIB) techniques
[Lambert 2017]. SEM images of the samples, Figure 12, show that ash particles penetrated up to about 60 µm into the inlet wall pores. The penetration was
evident already at 3,000 km (samples 3k1 and 3k2), resulting in improved filtration efficiency early in GPF life.
https://www.dieselnet.com/tech/gasoline_particulate_filters.php 16/23
12/19/2019 Gasoline Particulate Filters

Figure 12. Cross-section of filter samples with ash penetrating into the pores
Ash deposits are outlined in red in each image
(Courtesy of Ford Motor Company and MIT)

Some reports suggest that the ash distribution changes, with more ash being deposited at the plugged channel ends, as the overall ash load in the GPF increases.
A significant growth of ash plugs was observed when the GPF ash load increased from 10 g/L to 20 g/L [Seong 2017].

Hydrated Ash. Crystallites of hydrated ash material have been found firmly anchored to the GPF substrate, Figure 13 [Lambert 2017], which were identified as
CaSO4·nH2O, where 0.5 ≤ n ≤ 2 (compounds related to gypsum, CaSO4·2H2O). These hydrated CaSO4 species appeared to retain water up to temperatures as
high as 900°C. The fact that these crystals could stay in a hydrated form under most of the temperatures experienced in a GPF may explain their ability to grow
to very large sizes—some of the measured crystallites had dimensions in excess of 300 μm.

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 17/23
12/19/2019 Gasoline Particulate Filters

Figure 13. Hydrated ash structures at GPF inlet channel


(Courtesy of Ford Motor Company and MIT)

The implications of this hydrated ash material may include increased ash density, decreased ash transport and potential filter pore clogging. It was further
suggested that higher GPF ash densities (relative to diesel) can be attributed to higher temperatures in gasoline exhaust and the lack of significant soot cake that
may change the ash density.

It may be noted that in some work by Argonne that used accelerated aging, CaSO4 was not identified as a component of ash from GDI combustion [Seong
2016]. Later research that used field aging confirmed that CaSO4 was indeed present in the GDI ash [Seong 2017]. CaSO4 is also a main component of ash from
diesel engines.

Ash-Catalyst Interaction. Compounds of several elements originating from lube oil additives, the upstream catalysts (if present) or from corrosion—including
phosphorus (P), calcium (Ca), zinc (Zn) and iron (Fe)—may show negative effects on the TWC performance and durability. Ash including these elements may
be distributed in and on the filter wall, and the pattern may depend on the washcoat loading, washcoat distribution, temperature history and other factors
[Lambert 2017].

In a study with TWC catalyzed, laboratory-aged GPFs, a deterioration of light-off temperatures was seen with ash accumulation, particularly below 300°C. The
light-off temperature increase was apparent already at low ash loadings of ~5 g/L. NOx conversion was most impacted, even under rich conditions [Seong 2017].

4.3 GPF Regeneration and Control

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 18/23
12/19/2019 Gasoline Particulate Filters

4.3.1 Overview
Under typically encountered operating conditions, GPFs are regenerated passively when oxygen concentrations increase during fuel shut-off events. The term
passive regeneration usually refers in diesel engines to soot oxidation by NO2. However, due to the low NO2 concentration in gasoline exhaust, passive
regeneration of the GPF occurs via oxygen chemistry, at temperatures of 600°C and above [Van Nieuwstadt 2017].

GPF regeneration can be promoted by a three-way catalyst, Figure 14, through the catalytic effect of the noble metals and/or the oxygen storage capability of
ceria. In the uncoated filter only 50% of the soot was removed at a peak temperature of 600ºC, increasing to 90% at 650ºC. In contrast, soot oxidation in the
coated GPF increased rapidly at temperatures above 550ºC, with complete removal of the soot at ~570ºC [Morgan 2015].

100

cGPF
80
bare GPF

Soot Burnout, % 60

40

20

0
450 500 550 600 650 700

Filter Inlet Temperature, °C

Figure 14. Effect of a TWC GPF coating on soot combustion under fuel cut-off conditions
Soot burnt after 10 vehicle deceleration fuel cut-off events. NEDC cycle with 10 repetitions of the EUDC segment, where the final EUDC deceleration from 120 km/h triggers a fuel
cut-off. Peak filter inlet temperatures controlled by varying the dynamometer gradient.

Due to the lower soot load in the GPF, the exotherm during regeneration is lower compared to diesel systems. During the development of the Mercedes S500
GPF system, filter regenerations were induced by initiating a fuel cut phase of approximately 30 s. At the filter inlet temperature of 700°C, the maximum GPF
temperature reached 1150°C due to the soot burn-off [Lanzerath 2017].

4.3.2 Active Regeneration Assist

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 19/23
12/19/2019 Gasoline Particulate Filters

An active GPF regeneration mechanism is desired to ensure the filter does not become plugged during periods of challenging, low temperature driving. GPF
heating can be achieved through such strategies as spark retard and air-to-fuel ratio (AFR) modulation. GPF heating can be further supported by a number of
secondary actuators, including cam phasing, idle speed increase, turbine bypass, injection mode, and rail pressure [Van Nieuwstadt 2019].

Spark retard decreases combustion efficiency and increases the fraction of fuel energy converted into exhaust gas enthalpy [Van Nieuwstadt 2017]. At vehicle
speeds below ~50 km/h, a moderate spark retard can increase the exhaust gas temperature by 50°C without a noticeable impact to the driver. Aggressive spark
retard, beyond 20°CA, can achieve a temperature increase of up to 200°C at the expense of very poor fuel economy and objectionable NVH. If aggressive spark
retard is used, the driver may have to be notified about the condition via the vehicle dashboard.

AFR modulation is another actuator that can be used for GPF regeneration [Van Nieuwstadt 2019]. By introducing a pulsating λ bias, Figure 15, an exotherm on
the order of 80°C could be generated over the TWC to assist the regeneration of a downstream GPF. In addition, in V-configuration engines with one GPF
serving both exhaust banks, a bank-to-bank AFR modulation is possible, where an exotherm is generated over the TWC-coated GPF itself. Unlike spark retard,
AFR modulation affects the performance of the TWC—care must be taken to ensure that emission limits for gaseous pollutants are not exceeded.

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 20/23
12/19/2019 Gasoline Particulate Filters

Figure 15. Example AFR modulation strategy


3.5 L V6 GTDI engine with two TWCs per bank and one GPF filtering exhaust from both banks. Oxygen breaks through the TWC after 1-2 s; the λ bias is maintained for 9-10 more
seconds.
(Courtesy of Ford Motor Company)

A combination of spark retard and AFR modulation could achieve GPF regeneration temperatures in a Ford F150 with a 3.5 L V6 turbocharged, direct injected
(GTDI) engine over two challenging drive cycles: stop-and-go driving below 30 mph, and steady-state driving at 40 mph or above [Van Nieuwstadt 2019].

A soot estimation algorithm is needed to trigger an active regeneration event. Since most of the GDI soot is emitted during cold start, with the quantity
generally proportional to fuel, a simple soot estimation algorithm can be based on the fuel consumption and the engine coolant temperature [Van Nieuwstadt
2017]. Additionally, a GPF pressure drop based soot estimation can be implemented using a differential pressure sensor.

4.3.3 Protection Strategy


The GPF requires a protection strategy that includes at least two components. First, the engine needs to be protected against high back pressure levels that can
cause hanging exhaust valves and engine damage. This can be realized using a pressure sensor to limit torque and therefore the air flow and GPF pressure drop.

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 21/23
12/19/2019 Gasoline Particulate Filters

Second, the GPF must be protected against thermal events caused by the simultaneous presence of a high soot load, oxygen, and high temperature. If these
conditions occur, they may lead to cracking or melting of the GPF substrate. In the gasoline engine, the soot combustion rate and the exotherm can be easily
controlled by metering the oxygen flow into the GPF. More particulars on GPF regeneration control can be found in the literature [Van Nieuwstadt 2017][Van
Nieuwstadt 2019].

Acknowledgements

We are grateful to Christine Lambert of Ford Motor Company and Carl Justin Kamp of MIT who provided ash images (Figure 11, Figure 12, Figure 13) as well
as valuable comments on this paper. The section on hydrated ash is based in part on the communication with Dr. Kamp, and includes some observations from
his research on the morphology of DPF/GPF ash that was not yet published at the time of writing.

References
Andersson, J., Demuynck, J., Hamje, H., 2017. Proceedings from the 21st ETH Conference on Combustion Generated Nanoparticles, June 19-22, 2017, Zürich, Switzerland,
http://nanoparticles.ch/archive/2017_Andersson_PR.pdf
Boger, T., Cutler, W., 2018. “Reducing Particulate Emissions in Gasoline Engines”, SAE International, Warrendale, PA, USA, ISBN 978-0-7680-9417-6
Chan, T.W., et al., 2016. “Characterization of Real-Time Particle Emissions from a Gasoline Direct Injection Vehicle Equipped with a Catalyzed Gasoline Particulate Filter During Filter Regeneration”, Emission
Control Science and Technology, 2(2), 75-88, doi:10.1007/s40825-016-0033-3
Choi, S., Seong, H., 2016. “Lube oil-dependent ash chemistry on soot oxidation reactivity in a gasoline direct-injection engine”, Combustion and Flame, 174, 68-76, doi:10.1016/j.combustflame.2016.09.019,
https://www.researchgate.net/publication/308796840
Czerwiński, J., Comte, P., et al., 2017. “Nanoparticle Emissions of GDI Car with Increased Lube Oil Consumption, Potentials of GPF”, Proceedings from the 21st ETH Conference on Combustion Generated
Nanoparticles, June 19-22, 2017, Zürich, Switzerland, http://nanoparticles.ch/archive/2017_Czerwinski_PR.pdf
Davis, S.C., Williams. S.E., et al., 2017. “2016 Vehicle Technologies Market Report”, Oak Ridge National Laboratory, Report ORNL/TM-2017/238,
http://cta.ornl.gov/vtmarketreport/pdf/2016_vtmarketreport_full_doc.pdf
Inoda, S., Nomura, Y., Ori, H., and Yabuzaki, Y., 2017. “Development of New Coating Technology Optimized for Each Function of Coated GPF”, SAE Technical Paper 2017-01-0929, doi:10.4271/2017-01-0929
JM, 2017. “PGM Market Report”, Johnson Matthey, May 2017, http://www.platinum.matthey.com/documents/new-item/pgm%20market%20reports/pgm_market_report_may_2017.pdf
Kato, K., Thier, D., Kattouah, P., Ohara, E., Vogt, C.D., et al., 2014. “Advanced Catalyzed Gasoline Particulate Filter to Fulfill Future Emission Targets”, Proc. 18th ETH Conference on Combustion Generated
Nanoparticles, Zurich, June 22-25, 2014, http://www.nanoparticles.ch/archive/2014_Kato_PR.pdf
Kattouah, P., 2013. “Ceramic Wall Flow Filter for Particulate Emission Reduction of Petrol Engines”, Cambridge Particle Meeting, University of Cambridge, May 24, 2013,
http://www.cambridgeparticlemeeting.org/sites/default/files/Presentations/2013/PKattouah(NGK)_2013_Wall%20flow%20filter%20for%20particulate%20emission%20reduction%20of%20petrol%20engines.pdf
Kattouah, P., Kato, K., Thier, D., et al., 2014. “Advanced Gasoline Particulate Filter for Effective Gasoline Emission Control Beyond Euro 6”, 7th Emission Control Conference, Dresden, Germany, 22-23 May
2014, https://www.researchgate.net/publication/267335874
Lambert, C.K., Chanko, T., Jagner, M., Hangas, J. et al., 2017. “Analysis of Ash in Low Mileage, Rapid Aged, and High Mileage Gasoline Exhaust Particle Filters”, SAE Technical Paper 2017-01-0930,
doi:10.4271/2017-01-0930

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 22/23
12/19/2019 Gasoline Particulate Filters
Lanzerath, P., Wunsch, R., Schön, C., 2017. “The first series-production particulate filter for Mercedes-Benz gasoline engines”, 17. Internationales Stuttgarter Symposium; Proceedings; Springer, Wiesbaden,
Germany, 851-865, doi:10.1007/978-3-658-16988-6_68
Lee, K., et al., 2013. “Particulate Emissions Control by Advanced Filtration Systems for GDI Engines”, DOE Annual Merit Review & Peer Evaluation Meeting, May 15, 2013,
https://www.energy.gov/sites/prod/files/2014/03/f13/ace024_lee_2013_o.pdf
Liu, X., Chanko, T., Lambert, C., Maricq, M., 2018. “Gasoline Particulate Filter Efficiency and Backpressure at Very Low Mileage”, SAE Technical Paper 2018-01-1259, doi:10.4271/2018-01-1259
Liu, X., Szente, J., Pakko, J., Lambert, C. et al., 2019. “Using Artificial Ash to Improve GPF Performance at Zero Mileage”, SAE Technical Paper 2019-01-0974, doi:10.4271/2019-01-0974
Morgan, C., 2015. “Platinum Group Metal and Washcoat Chemistry Effects on Coated Gasoline Particulate Filter Design”, Johnson Matthey Technology Review, 59(3), 188-192, doi:10.1595/205651315X688109
Muñoz, M.M., 2017. “Are GDI vehicle exhaust genotoxic like non-treated diesel exhaust?”, Proceedings from the 21st ETH Conference on Combustion Generated Nanoparticles, June 19-22, 2017, Zürich,
Switzerland, http://nanoparticles.ch/archive/2017_Munoz_PR.pdf
Seong, H.J., Choi, S., 2016. “Ash-Durable Catalyzed Filters for Gasoline Direct Injection (GDI) Engines”, US DOE Annual Merit Review & Peer Evaluation Meeting, Washington, DC, June 9, 2016,
https://energy.gov/sites/prod/files/2016/06/f32/ace024_seong_2016_o_web.pdf
Seong, H.J., Choi, S., 2017. “Ash-Durable Catalyzed Filters for Gasoline Direct Injection (GDI) Engines”, 2017 Annual Merit Review, US DOE, Washington, DC, June 5-9, 2017,
https://energy.gov/sites/prod/files/2017/06/f34/acs024_seong_2017_o.pdf
Tanaka, A., Miyoshi, N., Sato, A., 2018. “Development of Low Pressure and High Performance GPF Catalyst”, SAE Technical Paper 2018-01-1261, doi:10.4271/2018-01-1261
Tang, W., et al., 2019. “On Developing Advanced Catalysts Systems to Meet China New Regulations”, SAE Technical Paper 2019-01-0978, doi:10.4271/2019-01-0978
Van Nieuwstadt, M., A. Shah, E. Serban, D. Martin, 2019. “Regeneration Strategies for Gasoline Particulate Filters”, SAE Technical Paper 2019-01-0969, doi:10.4271/2019-01-0969
Van Nieuwstadt, M., Ulrey, J., 2017. “Control Strategies for Gasoline Particulate Filters”, SAE Technical Paper 2017-01-0931, doi:10.4271/2017-01-0931
Xia, W., Yuan, X., Yang, D., Zheng, Y. et al., 2019. “Design of Catalyzed Gasoline Particulate Filter (cGPF) and Investigation of Its Durability Performance Using Accelerated Engine Aging”, SAE Technical
Paper 2019-01-0970, doi:10.4271/2019-01-0970
Yang, J., P. Roth, T.D. Durbin, K.C. Johnson, et al., 2018. “Gasoline Particulate Filters as an Effective Tool to Reduce Particulate and Polycyclic Aromatic Hydrocarbon Emissions from Gasoline Direct Injection
(GDI) Vehicles: A Case Study with Two GDI Vehicles”, Environ. Sci. Technol., 52(5), 3275-3284, doi:10.1021/acs.est.7b05641
Yoshioka, F., Kato, K., Aoki, T., Makino, M. et al., 2019. “Performance of Next Generation Gasoline Particulate Filter Materials under RDE Conditions”, SAE Technical Paper 2019-01-0980, doi:10.4271/2019-
01-0980

###

https://www.dieselnet.com/tech/gasoline_particulate_filters.php 23/23

Potrebbero piacerti anche