Atmospheric Environment 134 (2016) 109e120

Contents lists available at ScienceDirect

Atmospheric Environment
journal homepage: www.elsevier.com/locate/atmosenv

Characterisation of diesel particulate emission from engines using

commercial diesel and biofuels
T. Ajtai a, *, M. Pinte

r b, N. Utry b, G. Kiss-Albert b, G. Gulya
s c, P. Pusztai d, R. Puska

s d,

Bereczky e, Gy. Szabados f, G. Szabo
A.
a, Z. Ko

nya d, Z. Bozo

ki a
a
MTA-SZTE Research Group on Photoacoustic Spectroscopy, Szeged, H-6720, Hungary
b
Department of Optics and Quantum Electronics, University of Szeged, Szeged, H-6720, Hungary
c
Hilase Development, Production, Service and Trading Limited Company, Szeged, H-6727, Hungary
d
Department of Applied and Environmental Chemistry, University of Szeged, Szeged, Rerrich B

ela t

er 1, 6720, Hungary
e
Department of Energy Engineering, University of Technology and Economics, Budapest, H-1111, Hungary
f
KTI-Institute for Transport Sciences, Budapest, H-1119, Hungary

h i g h l i g h t s

We investigated the diluted exhaust emitted by a diesel engine.

The engine was operated with various mixtures of diesel and biodiesel fuels.

Concentration and size distribution strongly depend on engine condition and fuel type.

The thermal evolution of the emitted particulates was also investigated.

Changes in concentration and structure were both observed due to thermal treatment.

a r t i c l e i n f o a b s t r a c t

Article history: In this paper, the number concentration and the size distribution of diluted diesel exhaust particulate
Received 2 July 2015 matter were measured at three different engine operating points in the speed-load range of the engine as
Received in revised form follows: 1600 rpm; 50% load, 1900 rpm; 25% load, 1900 rpm; 75% load, adopted from the UN ECE Vehicle
21 March 2016
Regulation no. 49 (Revision 2) test protocol using pure diesel and biodiesel fuels, as well as their
Accepted 23 March 2016
Available online 26 March 2016
controlled blends. The emitted particulate assembly had lognormal size distribution in the accumulation
mode regardless of the engine operational condition and the type of fuel. The total number and volume
concentration emitted by the diesel engine decreased with increasing revolution per minute and rated
Keywords:
Diesel particulate
torque in case of all the fuel types. The mixing ratio of the fuels did not linearly affect the total emission
Biodiesel but had a minimum at 75% biodiesel content. We also studied the thermal evolution of the emitted
Size distribution particulates using a specially designed thermodenuder (TD) heated at specific temperatures (50  C,
Volatile particle 120  C, and 250  C). The first transition, when the temperature was increased from 50  C to 120  C
Thermodenuder resulted in lower number concentrations with small relative shifts of the peak position. However, in case
of the second transition, when the temperature reached 250  C the individual volatile particulates
adsorbed onto the surface of soot particles were completely or partly vaporised resulting in lower total
number concentrations with a substantial shift in peak position.
© 2016 Published by Elsevier Ltd.

1. Introduction scientific interest nowadays. The highly absorbing carbonaceous

Owning to their climate and health relevance diesel emitted
carbonaceous particulates are in the middle of gradually increasing
* Corresponding author.
E-mail address: ajtai@titan.physx.u-szeged.hu (T. Ajtai).
aerosol emitted by diesel engines have a positive and direct climatic
effect, while particles ejected by i.e. aircraft engines can influence
cirrus and contrail formation eventuating in indirect and negative
climate forcing respectively (Sekiguchi et al., 2003). According to
the latest scientific assessments, regarding its climate impact,
carbonaceous particulate matter (CPM), which is the by-product of

http://dx.doi.org/10.1016/j.atmosenv.2016.03.046
1352-2310/© 2016 Published by Elsevier Ltd.
110 T. Ajtai et al. / Atmospheric Environment 134 (2016) 109e120
incomplete combustion, is the second most important anthropo-
genic emission. Only the CO2 gas has larger climatic impact than
CPM (Bond et al., 2013). The optical response of aerosol strongly
depends on size and chemical composition. Residence time and
chemical reactivity also strongly depend on the characteristic
diameter of the aerosol especially in the fine size region (PM1).
Thus, size determines the climatic effect of ambient particulates
both directly and indirectly.
Diesel soot is a dominant source of CPM especially in the fine
particle size region (Braun et al., 2005). Due to their large surface
area per unit mass and high adsorption capability for toxic sub-
stances, diesel soot particles have considerable and adverse health
effects. Combustion related carbonaceous particulate matter is
considered to be a significant risk factor of cardiopulmonary and
cancer mortality (Lloyd and Cackette, 2001). Both the surface area
per unit mass and the pulmonary deposition efficiency increases
towards smaller particle sizes, therefore, the size of CPM is also a
considerable factor of air quality.
Diesel exhaust is composed of versatile mixtures of abundant
composites found in various mixing states. Diesel soot originates
from versatile sources having different physicochemical features.
Generally, diesel soot particles show high variety in chemical
composition, size as well as in morphology and microstructure. The
characteristic parameter set of diesel emitted aerosol strongly de-
pends on the type of fuel and on the operational condition of the
engine as well. Moreover, the chemical composition of the
exhausted particles is also strongly affected by the actual sampling
conditions. It is because the equilibrium state of vapour-particle
partitioning is mainly governed by thermal energy, resulting in
different vapour particle ratios at different temperatures. Owning
to the dynamic nature of vapour-particle partitioning in the tur-
bulent reactive plume of the engine tailpipe system, representative
sampling and accurate measurement of the exhausted particulate is
one of the major challenges in this field. Sampling directly from the
tailpipe provides useful data regarding the operational conditions
of the engine, however it only delivers limited climate and health
relevant information. In order to ensure measurement conditions,
which are more representative of real ambient processes, a
specially designed dilution chamber has to be applied. That way the
temperature of the exhaust gas and the concentration of the raw
exhaust are reduced close to ambient temperature and to a level
required by the applied measurement system, respectively. This
way, the initial dilution effect on particle formation can be accu-
rately investigated. To investigate the ambient effects on the che-
mophysical properties of the diesel particulates, specially designed
reaction or smoke chamber experiments are required. Modern
diesel engines are all equipped with after treatment devices like
EGR (Exhaust Gas Recirculation) and DPF (Diesel Particulate Filter)
units. However these after treatment devices do not only simply
reduce the amount of emission but also change the characteristics
of the emitted particles. Thus, the investigation of raw (engine-out)
emission is most essential for fuel and engine development, and
also important in order to better understand the masking phe-
nomena of these devices.
When engine exhaust encounters cool ambient conditions,
vapour species with high molecular weights are to condensate onto
non-volatile soot particles by heterogeneous condensation or form
new individual volatile particles through nucleated condensation.
Both of these phenomena take place until the thermally driven
equilibrium state is reached at a given temperature. At this dynamic
equilibrium the residual molecules remain in the gaseous phase.
These processes modify the chemical composition and the related
mixing state of the freshly emitted diesel particulates, masking
their number concentration and size distribution as well. The vol-
atile to non-volatile ratio, the rate of individual volatile particulates
with a homogeneous and spherical geometry to the particles
composed of the non-volatile soot fractal aggregates and the
condensed volatile material with core-shell mixing geometry are
essential both from the prospective of climate relevance and risks
to human health (Cheng, 2013). The mixing state and also the
mixing geometry of the volatile to non-volatile particulate struc-
tures can be classified by the different thermal stabilities of the
composites (Burtscher et al., 2001). According to this approach
volatile material can be removed from the exhaust by a thermo-
denuder heated at a given temperature. The evaporated substances
are subsequently absorbed by an activated carbon filter. Following
that it becomes possible to investigate the remaining particles with
modified number concentration and size distribution.
Due to gradually increasing scientific interest, the real-time
investigation of the number concentration and size distribution of
diesel emitted particulates has become an intensively studied issue
nowadays. Most of the studies are focussing on diesel particulate
size distribution measured directly at the exhaust pipe that is
crucial from the point of view of fuel and engine development,
however resulting in limited climate and health relevant informa-
tion. Many of the size distribution measurements carried out under
real ambient conditions that focus on traffic related aerosols are
also available in the literature. Although, the ambient factors and
the cross sensitivity of other sources can reduce the reliability and
the general validity of the source related information obtained from
these studies. The climate and health relevant source related
emission properties of the diesel particulates can be modelled and
investigated under controlled laboratory circumstances using a
standard dilution chamber at atmospheric pressure and tempera-
ture (Burtscher, 2005). There are a great number of source related
studies executed under well controlled laboratory conditions in a
specially designed dilution chamber available in literature.
Furthermore, size distribution measurements aiming to investigate
the dynamic changes in the volatile-non-volatile ratio have also
been intensively studied. Most of them are focussing on emission
characteristics using specific fuel types. Moreover, although the
biodiesel content in the diesel fuel is specifically regulated in all EU
countries, the effect of the diesel to biodiesel ratio on emission and
on volatility have also been investigated, but the number of
comprehensive studies using the same measurement condition,
engine type and experimental setup are still limited (Popovicheva
et al., 2014; Uhrner et al., 2011; Rothenbacher et al., 2008;
Bukowiecki et al., 2012; Liu et al., 2007; Abdul-Khalek et al.,
1998; Ahlvik et al., 1998; Trapel et al., 2005; Mathis et al., 2004;
Steiner et al., 2013).
The ratio of biodiesel to conventional diesel in the total fuel
blend can influence both the total number and volume concen-
tration and distribution of the particle emission. Therefore, inves-
tigating these parameters in the function of biodiesel content and
volatility is extremely important from the point of view of engine
and fuel development too. Therefore, the detailed investigation of
the size distribution and the volatility of diesel emitted particulates
as a function of biodiesel content of the applied fuel are timely and
important issues in the fields of climate and air quality research.
Hence the aim of this work is to evaluate and characterize the effect
of fuel composition on diesel particulate emission at different en-
gine loads and volatility test conditions in order to better under-
stand the characteristics of particulate emissions from vehicle
engines.
In this paper, we are studying the number concentration and
size distribution of diesel particulate matter in the function of the
operational conditions of the engine using diesel and biodiesel
fuels as well as their controlled mixtures. We also study the size
distribution of the emitted particulates in the function of thermal
stability under various measurement conditions.
 T. Ajtai et al. / Atmospheric Environment 134 (2016) 109e120
2. Experimental methods
The experimental set up of the sampling system is shown in


Fig. 1. A RABA D10 UTSLL 160 heavy duty diesel engine was used to
generate the exhaust particulates.
The diesel engine exhaust was directed into the full flow con-
stant volume sampling (CVS) system designed for the characteri-
sation of diesel emission according to the UN ECE Vehicle
Regulation no. 49 (Revision 2). The detailed description of the full
flow CVS technique can be found in the Regulation no. 49 (ECE,
2000). Briefly, the end of the engine exhaust pipe was connected
to the entry port of the full flow CVS system. Following that the
inner port having a diameter of 120 mm led the exhaust into the
centre of the dilution tunnel. The inner diameter of the tunnel was
600 mm. The sampling of the primary diluted raw exhaust was
carried out from the centre of the tunnel at a distance of 6000 mm
from outlet point of the inner port. The dilution of the raw exhaust
was executed by filtered air through the dilution inlet port. Exhaust
mass flow rates and dilution factors used during the measurement
points are presented in Table 4. In the next step the concentration of
the primarily diluted exhaust was further reduced by a factor of 10
by an ejector diluter (Palas GmbH VKL 10). There were two mea-
surement modes applied in the tests. During measurements aiming
to characterize diesel emission in the function of engine working
conditions and fuel types, the heating unit of the thermodenuder
was set to 50  C. In case of volatility measurements the thermo-
denuder unit was heated to 120  C and 250  C, respectively. The
detailed description of the basic phenomena and the instrumen-
tation of the low flow TD applied here are described in details
elsewhere (Burtscher et al., 2001). Most of the commercially
available TD units are operated at around 3 min1 l. However, to
fulfil real time and isokinetic sampling conditions, a lower flow rate
is required by aerosol size spectrometers such as the scanning
mobility particle sizer (SMPS) (0.3 min1 l flow rate) (Burtscher
et al., 2001). The low flow TD applied here was optimized for
these specific experimental conditions. This modular type TD
concept was developed especially for combustion exhaust mea-
surements where high volatile matter content is expected. Origi-
nally, the adsorption efficiency of the TD developed by Burtscher
et al. (2001) was increased by an active cooling system in which
Fig. 1. The experimental set up of the sampling system.
111
ambient air was directed through an activated charcoal during
operation. However, this methodological solution can cause sig-
nificant thermophoretic losses due to the fast temperature drop
inside the tube (Fierz et al., 2007). Therefore, to minimize these
losses we reduced length of the coupling tube between the
adsorber and desorber module as short as possible. Furthermore,
we thermally isolated the adsorber tube, and we neglected the
additional cooling of the activated charcoal. Thereafter, the particle
penetration efficiency and the temperature dependency of these
slight modifications were quantified using diffusion dried NaCl
particulates. Shortly, the size distribution and number concentra-
tion of the NaCl particles were measured before and after the TD at
different TD temperatures. For reference measurements a simple
connector tube was used having the same dimensions than that of
the one originally implemented into the TD (Stevanovic et al.,
2015). Due to our slight modifications, in case of our TD we
reached moderate temperature drop in the last section of the
desorber unit compared to the TD unit by Burtscher et al. (2001). As
a result of that the temperature dependency of particle losses in the
TD have been decreased. However, corresponding to the TD by
Burtscher et al. (2001) these modifications did not yield remarkable
improvement in particle diffusion losses. The number concentra-
tion and size distribution of the diesel emitted particles were
measured by a Scanning Mobility Particle Sizer (SMPS, GRIMM
Aerosol Technique, Germany). The SMPS consists of two parts. First,
the sampled particle assembly is led into a Classifier “Vienna” Type
Long Differential Mobility Analyser (LDMA, Model #5.500) in which
the particles are separated based on their electrical mobility. Then
the size classified particles are sent to the Condensation Particle
Counter (CPC Model #5.400) where the size segregated particulates
are counted. The sheath and aerosol flow rates were set to 3.0 and
0.3 L per minute, thus the SMPS measured the size distribution over
the range of 11.1e1083.3 nm with resolution of ±0.5 nm at 11.1 nm
to ±30 nm at 1083.3 nm. In order to minimize the shielding artefact,
coinciding correction of the measured data was also applied.
Finally, for representation purposes, the morphology and the
microstructure of the generated particulates were sampled deter-
mined by Transmission Electron Microscopy (TEM, FEI Tecnai G2 20
X-Twin). The TEM was operated at 200 kV accelerating voltage. The
TEM grid used for the investigation of the morphology was carbon
112 T. Ajtai et al. / Atmospheric Environment 134 (2016) 109e120
film coated 200 mesh copper grid (CF200-Cu, Electron Microscopy
Sciences, USA), while for the characterization of the microstructure
of the primary particles lacey carbon grids (LC200-Cu, Electron
Microscopy Sciences, USA) were used. Before imaging, the grid was
placed in a polystyrene Petri dish and equilibrated in a desiccation
chamber at ~25  C and ~20% RH. To measure the degree of crys-
tallinity of the interrogated particles, Electron Diffraction (ED) and
high resolution electron microscopy (HRTEM) mode was used with
a point to point resolution of 0.26 nm. In order to avoid the fractal
displacement during the exposition on the grid in the HRTEM
mode, the sample was preheated before imaging. The commercially
available pure diesel EN (European Norm) 590, pure specified
biodiesel fuel (FAME, EN 14214) as well as their specified mixtures
were used in this study. The characteristic parameters of the test
engine and the fuels are listed in Tables 1 and 2. The measurements
were executed at three different engine operational conditions
adopted from the 13-mode steady state cycle of UN ECE Vehicle
Regulation no. 49 (Revision 2). The number concentration and size
distribution of the diesel exhaust were measured at each adopted
mode using pure and defined mixing ratios of diesel and biodiesel
fuels at three different TD temperatures. All the measurement
conditions presented in this study can be seen on Fig. 1 and are
summarized in Tables 3 and 4. For comparison purposes the data
demonstrated throughout this work are dilution and particle loss
corrected.
3. Results and discussion
3.1. Dynamic changes of diesel emission
The size distributions in number representation measured at all
engine operating conditions and at all defined fuel mixtures
defined in Table 3 are shown in Fig. 2. As demonstrated by these
examples the agglomerated particulate assembly has lognormal
size distribution in the accumulation mode regardless of the engine
operational condition and the fuel type. While the geometric mean
diameter is relatively insensitive to the engine operational condi-
tion and the type of fuel, the geometric standard deviation shows
higher values in engine operational condition no. 3, than in all other
cases regardless of the used fuel type. The characteristic parameters
of all the measured sized distributions are summarized in Table 5.
The relative deviations in total number and volume concentra-
tions are plotted on Fig. 3 as a function of fuel mixing rate at the
specific engine operational conditions. The volume of the particu-
late was deduced simply from the diameter using a spherical ge-
ometry approximation. The real morphology of the diesel soot have
complex fractal structure built up by small monomers with
spherical geometry. However, the spherical approximation means
limitation regarding the data interpretations, it's still a generally
accepted approach especially when the objective is to discuss
relative deviations (Ban-Weiss et al., 2010; Rissler et al., 2012). The
Table 1
Characteristic parameters of the test engine.
Bore 120.5 mm
Stroke 150 mm
Number of cylinders 6 number representation measured in case of the different engine
Engine layout inline
Compression ratio 15.2
Displacement 10,350 cm3
Injection direct injection
Fuel supply system Bosch
Charging yes
Rated power 160 kW/1900 rpm
Rated torque 920 Nm/1600 rpm
total number and volume concentration emitted by the diesel en-
gine decreased with increasing revolution per minute and rated
torque. However, surprisingly, the mixing ratio of the fuels did not
linearly affect the total emission but had a minimum at 75% FAME
content (Fig. 3). This anomalous behaviour is not consistent with
the existing data in the literature. However, since, the operational
condition of the diesel engine is not optimized for high FAME
content and also since the regulation restricts the maximum FAME
content of diesel fuel below 10%, the available studies focus mainly
on the intercomparison of pure FAME and diesel fuel or diesel fuel
with limited FAME content. Detailed analysis of biodiesel blends
with high mixing ratio is limited (Puzun et al., 2011; XuSheng et al.,
2009; Turrio-Baldassarri et al., 2004; Tinsdale et al., 2010; Lin et al.,
2008). Based on those data lower total number-concentration is
predicted in case of pure FAME than in case of pure diesel fuel if
linear relation is assumed between fuel blend and total number-
concentration. However, similar behaviour have also been recog-
nised elsewhere but with different types of biofuel and PM relation
regardless of a detailed explanation of the possible theoretical
background (Lin et al., 2008). Moreover, many earlier studies have
experimentally confirmed that the emission characteristics of
different diesel engines can vary even if the same operational
condition and fuel are applied (Abdul-Khalek et al., 1999; Burtscher,
2005; Lloyd and Cackette, 2001; Cheng, 2013). The statistical
summary of the relative changes in total emission are summarized
in Table 6. Comparing the pure fuels, the total number concentra-
tion of the emitted particulates was higher, while the total volume
of those was lower using pure FAME fuel than the same values in
case of the commonly used pure diesel fuel at all engine operating
conditions. Similarly, the changes in total number-concentration
exceeded 25% when the engine operating condition turned from
no. 1 to no. 2 using pure FAME fuel, meanwhile the changes in total
volume concentration associated with the same transition
remained below 12% (Fig. 3). These phenomena can be explained by
the changes in particle population statistics. In the first case, the
different median diameter together with the near identical
geometrical standard deviation and total number concentration
resulted in substantially different (lower) volume concentration
values. Similarly, in the second case, although the median value was
higher in engine operating condition no. 2 than in engine operating
condition no. 1, the lower total number concentration together with
the higher geometric standard deviation rearranged the relative
weight of number and volume concentrations at each size bins
issuing negligible deviation in total volume. The statistical sum-
mary of total number and volume concentration measured at all
the presented engine operating conditions and fuel types refer to
pure diesel fuel can be found in Table 6.
3.2. The particle size distribution as a function of exhaust
temperature
The thermal behaviour of the agglomerated particulates was
investigated in case of all engine conditions and types of fuels at
different temperatures defined in Table 3. The first set temperature
point (50  C) indicated a reference “not-heated” condition corre-
sponding to the specified sampling condition stated in UN ECE
Vehicle Regulation no. 49 (Revision 2). The size distribution in
conditions and fuel types at three different temperatures are
plotted in Fig. 4, As the temperature increased from the base con-
dition (when the heater is set to 50  C in the thermodenuder unit)
to 120  C, the number concentration decreased significantly in case
of all engine operational conditions and fuel types. However, the
peak position of the curves shifted slightly towards both directions
without a clear trend in the function of measurement conditions.
 T. Ajtai et al. / Atmospheric Environment 134 (2016) 109e120 113

Table 2
Characteristic parameters of the test fuels (Szabados and Bereczky, 2015).

Fuel type Density [kg/dm3] 15  C Kin. viscosity [mm2/s] at 40  C Lower heating value [MJ/kg] Flash point (open cup method) [ C] Flash point (closed cup method) [ C]

Diesel 0.837 2.98 42.12 90 70

FAME 0.877 5.05 36.29 189 201

Table 3 resulted in particulate number decrement, while the later one is

Measurement conditions.
# Fuel type
I. 100% Diesel
II. 25% FAME
III. 75% FAME
IV. 100% FAME
# Engine operating condition
1. 50% load, 1600 rpm
2. 25% load, 1900 rpm
3. 75% load, 1900 rpm
# Thermodenuder temperature
a 50  C
b 120  C
c 250  C
The statistical summary of the characteristic parameters of the
measured distributions at 120  C are summarized in Table 7.
Further increasing the temperature to 250  C remarkable structural
changes could be observed in the measured size distribution under
all engine conditions and fuel types (Fig. 4). Two characteristic
modes with various intensities and peak positions have been
identified at this temperature. The statistical summary of the
characteristic parameters of the measured distributions at 250  C
are also summarized in Table 8. The results demonstrated in Fig. 4
and the data in Tables 7 and 8 have already been corrected by the
losses of the TD, so the changes presented here are real and exceed
the TD effect itself. Possible explanation of such thermal behaviour
is generally based on the following approximation. Due to the fast
temperature drop in the dilution system the volatile and con-
densable molecules formed individual volatile particulates having
various compositions likely in a spherical geometry, while the
condensable substances having higher molecular weight adsorbed
onto the surface of soot particulates resulting in core-shell micro-
structures (Ishigoru et al., 1997).
The phenomenon characteristic of the first transition, when the
temperature was increased from 50  C to 120  C, during which
there were changes in number concentrations and peak positions,
have already been observed in many earlier works and explained by
the differences in thermochemical reactiveness of the individual
volatile particles and that of core-shell microstructures (Burtscher
et al., 2001; Burtscher, 2005; Cheng, 2013; Guo et al., 2014).
Namely, that the individual volatile particulates having low ther-
mochemical refractiveness completely vaporized, while the core-
shell microstructure only partly vaporized. The former could have
thought to have manifested in the shift of peaks. However, this
approximation is neglecting the generally accepted conception that
in the accumulation mode most of the particulates are solid which
cannot be completely vaporized at this temperature (Kittleson,
1998.). The fact that the changes in population statistic resulted
in a slight shift in peak position really support the theory of the
evaporation of molecules previously adsorbed at lower tempera-
ture in the CVS system onto the soot core. The mostly size inde-
pendent particle losses rather indicate other dominant phenomena
like limited pyrolysis of particles having limited or no graphitic
microstructure. However, the limited pyrolysis is more expected in
case of FAME than in case of diesel fuel, we have recognised
remarkable changes in all cases regardless of the type of fuel, which
is consistent with many already previous studies (Lapuerta et al.,
2008a, 2008b; Turrio-Baldassarri et al., 2004; Tinsdale et al.,
2010; Lin et al., 2008).
At the second transition, when the temperature reached 250  C,
structural changes were observed resulting in bimodal distribu-
tions under some measurement conditions. The peak positions of
the modes shifted towards the smaller size region compared to
baseline conditions (T ¼ 50  C) in all cases. Anyway, both of these
modes can be individually fitted with lognormal distribution in
almost all cases. Moreover, the envelope of the sum of these indi-
vidual modes can be well fitted to the original curve indicating a
distinct dominancy of particles that fall into these modes in the
whole particle assembly (Fig. 4.)
Although the monomodal to bimodal transition of the measured
size distribution were definitely recognized and quantified here
(Table 7), the adequate theoretical explanation of such thermal
behaviour of temperature treated diesel exhaust has not been
found earlier in the literature. The shift of peak position towards the
smaller size region at higher temperature e with remaining mon-
omodal distribution e have been documented in many studies until
recently (Cheng, 2013; Burtscher, 2005). It could be explained by
the rearrangement in population statistics due to the evaporation
of condensable compounds which were previously adsorbed onto
the surface of soot particles (Cheng, 2013). Although it is possible to
find small solid particles after temperature treatment even below
the 40 nm size range such as primary soot particles or soot fractal
aggregates with limited number of monomers, or different kinds of
metallic ashes (Giechaskiel et al., 2014; Giechaskiel and Martini,
2014), the generally accepted most possible scenario of the origin
of such small particulates is new particle formation by homoge-
neous nucleation. According to the authors, the cause of the
structural changes observed here is the re-condensation of

Table 4
Mass flow rates of the exhaust and primary dilution factors.

Type Engine operating condition 1 Engine operating condition 2 Engine operating condition 3
of fuel
Mass flow of Mass flow of diluted Dilution Mass flow of Mass flow of diluted Dilution Mass flow of Mass flow of diluted Dilution
exhaust gas [kg/h] exhaust gas [kg/h] factor exhaust gas [kg/h] exhaust gas [kg/h] factor exhaust gas [kg/h] exhaust gas [kg/h] factor

I. 470.3 5210.9 10.081 668.2 6165.5 8227 946.4 7690.9 7127

II. 467.8 5208.7 10.134 674.3 6125.9 8085 949 7686.9 7.1
III. 469.3 5231.5 10.147 655.8 6129.7 8347 936.9 7599.3 7111
IV. 473.4 5216.3 10.019 665.4 6140.2 8228 934.9 7682.3 7218
114 T. Ajtai et al. / Atmospheric Environment 134 (2016) 109e120

Fig. 2. The size distribution in number representation provided at all engine operating condition using various fuel mixing.

Table 5
Characteristic parameters of the size distributions measured at all engine operating condition and at defined fuel mixtures. CMD stands for count median diameter.

Fuel type CMD [nm] Total number concentration (*108) [cm3] Geometric standard deviation [nm]

Engine operating condition no.1.

I. 98.7 ± 0.55 4.51 73.7 ± 0.48
II. 101.8 ± 0.71 3.09 69.6 ± 0.62
III. 106.3 ± 0.68 2.16 75.3 ± 0.56
IV. 88.1 ± 0.59 4.67 61.6 ± 0.6
Engine operating condition no.2.
I. 107.6 ± 0.57 2.80 84.8 ± 0.45
II. 107.3 ± 0.88 1.83 76.0 ± 0.73
III. 105.3 ± 0.57 1.21 72.5 ± 0.55
IV. 91.1 ± 0.56 3.63 62.9 ± 0.55
Engine operating condition no.3.
I. 125.1 ± 3.13 1.03 132.9 ± 1.7
II. 108.2 ± 1.13 0.53 108.0 ± 0.78
III. 106.9 ± 0.87 0.34 91.0 ± 0.66
IV. 83.5 ± 0.92 1.71 74.0 ± 0.89

previously desorbed molecules originating from the surface of solid
soot particles which are not adsorbed on the activated coal in the
second stage of the denuder. The charcoal of the denuder was
activated and altered when the fuel or the operational condition of
the engine were changed during the thermal treatment measure-
ment program, in which the measurements at different tempera-
tures were carried out subsequently in increasing order. Therefore,
it is possible that the adsorption efficiency of the activated chair
coal at least partly decreased at the highest temperature making re-
condensation process possible inside or after the denuder.
For quantitative comparison purposes, the losses of total num-
ber and volume concentration compared to baseline condition are
plotted in Fig. 5. As shown by these examples, at the first temper-
ature transition (from 50  C to 120  C), the TD loss corrected
particle losses in total number concentration increased substan-
tially up to around 80% using fuel type no. I. no. II. and no. IV.
regardless of the engine operational conditions. However, fuel type
no. III. indicates smaller loss in all cases (Fig. 5a). These findings are
consistent with the earlier published data using commercial diesel
or other biodiesel fuels (Burtscher et al., 2001; Burtscher, 2005;
Cheng, 2013; Guo et al., 2014). At the subsequent temperature
transition (from 120  C to 250  C) the pure fuels regardless of the
engine operational condition and the mixing rate at measurement
condition no. III. provided negligible further losses in number-
concentration (Fig. 5). However, unexpected results have been
observed at engine operating conditions no. 2 and no. 3 when fuel
types no. II. and no. III. were used. In these cases, the total number
concentrations increased compared to total number concentrations
 T. Ajtai et al. / Atmospheric Environment 134 (2016) 109e120 115

Fig. 3. The relative deviation in total number (a) and volume concentration (b) as a function of fuel mixing rate and engine operational condition.

Table 6
Statistical summary of total number (a) and volume concentration (b) measured at all the applied engine operating conditions and fuel types.

Fuel type Engine operating condition no. 1 Engine operating condition no. 2 Engine operating condition no. 3

a) Relative changes in particle number concentration [%]

I. e e e
II. 30.40 33.44 47.68
III. 50.95 55.83 66.20
IV. 4.76 30.61 66.22
b) Relative changes in total volume concentration [%]
I. e e e
II. 30.57 37.99 55.79
III. 43.01 60.47 74.72
IV. 31.61 24.83 42.51

measured at 120  C. These observations can also be explained by
new particle formation caused by the denuder mentioned above.
However, re-calculating the total number concentration neglecting
the contribution of the particles' residue in the smaller mode also
resulted in similar trends at these particular cases (Fig. 5). There-
fore, other physical phenomena, than the denuder caused re-
condensation could be behind this particular behaviour. One
possible effect that can explain this behaviour is the fragmentation
of loosely bended sub-fractal fraction (Harris and Maric, 2002;
Rothenbacher et al., 2008). However, these physical phenomena
are more likely to be expected in case of FAME than in case of diesel
fuel, some earlier studies suggest that primary emitted particles can
be fragmented due to the low thermal stability adsorbates being
removed from their surface even in case of commercial diesel fuel
(Harris and Maric, 2002; Rothenbacher et al., 2008). This could
further attest the relevance of the fragmentation process, although
such particles could be identified during our measurements we
have not found evidence of superaggregates present in pure diesel
exhaust in literature.
To further support the essential of this interpretation total vol-
ume concentration and its relative changes measured at different
temperatures were also calculated (Fig. 5b.). In consonance with
number-concentration, the total volume concentration increased
significantly at the 50  Ce120  C transition and increased further or
remained almost constant at the 120  Ce250  C temperature
transition in all cases (Fig. 5b.). The relative changes in total
volume-concentration observed here are also consistent with the
earlier published data (Burtscher et al., 2001; Sakurai et al., 2003;
Ro€ nkko€ et al., 2007). Since the total volume concentration also
remained almost constant in the temperature transition from
120  C to 250  C, it is possible that the soot fractal aggregates or
super aggregates which are formed when two or more primary
aggregates collide with each other (Chakrabaty et al., 2014) suffer
fragmentations resulting in higher particle number in the whole
assembly but with unchanged total number concentration. At this
point, we also have to note, that this fragmentation phenomena can
observed only in case of 25% and 75% FAME content, while this type
of fragmentation is most expected in case of higher or pure FAME
content.
Besides, the on-line measurements, the morphology and the
microstructure of engine exhaust particulates were investigated by
Transmission Electron Microscopy (TEM). Due to the limited
number of samples collected at the particular measurement con-
ditions as well as the inherent nature of TEM methodology such as
the limited number of trapped particulates or evaporation of vol-
atile particulate from the grid by the applied beam energy, the TEM
pictures here serve only representation purposes and are analysed
only in the context of the discussion of the SMPS data. Neither the
types of fuel nor engine operational conditions have not shown any
characteristic features for any of the particular measurement con-
ditions. In all cases, we found dominancy of soot fractal aggregates
consisting of 7e10 nm primary particulates having spherical ge-
ometry (Figs. 6 and 7a). We have also found some evidence of soot
fractal superaggregate-like structures having more complex and
extended morphology even in case of commercial diesel fuel
(Figs. 6 and 7b). The primary particulates investigated by HRTEM
show amorphous turbostratic structure with randomly oriented
graphitic domains regardless of the fuel type (Figs. 6 and 7c). Apart
from carbon structures unidentified particulates with different
morphology have also been found among the samples (Figs. 6 and
7d-e). The electron diffraction pattern of the amorphous turbos-
tratic particulates consists of diffuse rings implying the lack of
crystallinity. The diffraction of the randomly oriented graphitic
domains cannot be discerned in the images since it is largely
masked by the diffuse diffraction of the amorphous particulates
(and the carbon film of the sample holder grid). In contrast to the
116 T. Ajtai et al. / Atmospheric Environment 134 (2016) 109e120

Fig. 4. The size distribution in number representation measured at different engine operating condition and fuel types at three different temperature (from top to bottom: 100%
diesel (I), 25% FAME (II), 75% FAME (III), 100% FAME (IV)).

diffuse diffraction of carbon structures distinct spot patterns were
obtained for the unidentified particulates demonstrating their
crystalline nature. Therefore the presented ED images confirm that
the fractal and superaggregates have amorphous turbostratic
structure with randomly oriented graphitic domains, while the
other type particles (Figs. 6 and 7e-d) have specific crystalline
structures.
So, these additional investigations have confirmed that in the
engine exhaust the diesel and biodiesel soot fractal aggregates
having graphitic or turbostratic molecule structure dominantly fall
into the submicron size domain. Besides, fractal super aggregates
with extended dimensions and complexity can also be identified
regardless of the applied fuel. Moreover, other types of particulates
were also recognised having crystallite molecule structure and
unknown thermal stability.
4. Summary and outlook
The number concentration and the size distribution of the
diluted diesel exhaust particulate matter were measured at three
different engine operating points from the speed-load range of the
engine as follows: 1600 rpm; 50% load, 1900 rpm; 25% load,
 T. Ajtai et al. / Atmospheric Environment 134 (2016) 109e120 117

Table 7
The statistical summary of the characteristic parameters of the measured distributions at 120  C. CMD stands for count median diameter.

Fuel type CMD [nm] Total number concentration (*108) [cm3] Geometric standard deviation [nm]

Engine operating condition no. 1.

I. 95.66532 1.22 65.84
II. 99.2583 0.97 64.17
III. 104.22142 1.11 67.49
IV. 85.08072 1.24 56.86
Engine operating condition no. 2.
I. 101.34134 0.64 69.61
II. 101.04047 0.47 63.91
III. 98.10056 0.52 60.05
IV. 87.4088 0.85 54.61
Engine operating condition no. 3.
I. 80.69511 0.23 60.64
II. 85.73022 0.095 57.51
III. 94.68666 0.12 58.15
IV. 72.06986 0.32 46.34

Table 8
The statistical summary of the characteristic parameters of the measured distributions at 250  C. CMD stands for count median diameter.

Fuel type CMD [nm] Total number Geometric standard deviation [nm]
concentration (*108)
Mode 1 Mode 2 Mode 1 Mode 2
[cm3]

Engine operating condition no.1.

I. 37.9 76.1 1.20 21.35 25.09
II. 37.2 73.3 1.07 19.7 23.2
III. 37.1 73.3 1.04 20.2 25.1
IV. 34.9 73.0 1.30 23.4 26.68
Engine operating condition no.2.
I. 36.6 75.4 0.76 24.8 26.8
II. 34.6 72.1 0.89 20.4 22.5
III. e 94.68 0.90 e 55.36
IV. 32.0 67.0 1.06 21.5 23.6
Engine operating condition 3.
I. 61.9 119.7 0.15 10.52 100,3
II. 55.3 88,6 0.31 11.42 96.41
III. 27 57.4 0.31 37.54 14.62
IV. 27.5 53.1 0.29 20.9 14.5

1900 rpm; 75% load, adopted from the UN ECE Vehicle Regulation
no. 49 (Revision 2) test protocol using pure diesel and biodiesel
fuels, as well as their controlled mixtures. The thermal evolution of
the diluted diesel particulates was also investigated at all the pre-
sented engine operational conditions and fuel types. The data
presented here further confirmed experimentally that diesel soot
provides lognormal distribution under ambient relevant conditions
with around 100nm count median diameter independently from
the fuel types and engine operating conditions (Fig. 2). However,
the total number and volume concentrations strongly, while the
peak positions weakly depend on fuel types and engine conditions
(Table 5). It has also been recorded that the total number and
volume emission is not monotonically affected by the biodiesel fuel
content, but has a minimum at 75% biodiesel to diesel ratio (Fig. 3).
The thermal treatment of the diluted particulate assembly (tem-
perature increase from 50  C to 120  C) resulted in significant
decrement in number concentration, while the peak position has
slightly changed (Fig. 5 and Table 7). This behaviour can be partly
explained by thermal desorption of volatile matter previously
adsorbed onto the active surface of solid soot and partly by the
limited pyrolysis of other material containing no soot (Fig. 6) At
higher temperature (250  C) structural changes (monomodal to
bimodal transition) have been observed and quantified in size
distribution in almost all cases (Fig. 6 and Table 8). The observed
particulates of the smaller size range were most likely formed as a
result of the denuder effect, while the particulates accumulating in
higher mode represent soot with graphitic structure without
adsorbed volatile matter. However, in some cases, higher value of
total number concentrations were experienced at 250  C compared
to 120  C, the total volume concentrations corresponding to this
transition remained constant (Fig. 5). The authors suggest that one
possible explanation of this behaviour could be the fragmentation
mechanism of the aggregates which is governed by the thermal
energy (Fig. 5). However, it is really curious that it only occurs in
case of fuel blends, while in case of the pure fuels it is definitely not
observed.
The extent of the biodiesel content in the total fuel amount can
influence the total volume and number of the emitted particulates
as well. According to the presented results, development of new
fuels can also be a powerful tool in order to further reduce the
diesel emissions and also to fulfil the continuously tightening
emission standards. The demonstrated data has also experimen-
tally shown that the changes in fuel composition can induce
changes in size distribution and mixing of soot particulates.
Therefore, we further confirmed that the biodiesel content of the
applied fuel can modify the characteristics of the emission. The
mixing state and the mixing geometry of the volatile to non-
volatile compounds measured in engine out emission serve as
input parameters for the development of EGR or DPF system.
Therefore they strongly influence the climate and health impact of
tailpipe emission. Investigating the thermally classified diesel
particulate matter under controlled ambient conditions by the
recently available high efficiency and low particle loss character-
istic thermodenuders (Huffmana et al., 2008; Fierz et al., 2007), is
118 T. Ajtai et al. / Atmospheric Environment 134 (2016) 109e120

Fig. 5. Relative losses of total number (a) and volume (b) concentration compared to baseline condition. Dotted lines represent values, where the smaller mode originating from the
TD effect was neglected in calculation.

Fig. 6. TEM, HRTEM and linked ED (inset) images illustrating the different kinds of particles identified in the exhaust in case of pure diesel. Scale values are 100, 500, 10, 100, 50,
50 nm, respectively.

now a possibility and worth to further investigate in the future. as fuel composition and sampling condition dependence. In order
Emission characteristics have engine, operational condition as well to acquire generally accepted, comprehensive data, it is essential to
 T. Ajtai et al. / Atmospheric Environment 134 (2016) 109e120
Fig. 7. TEM, HRTEM and linked ED (inset) images illustrating the different kinds of particles identified in the exhaust in case of biodiesel blends. Scale values are 100, 100, 10 nm, and
1, 1, 1 mm, respectively.
further improve sampling units and also to more rigidly stan-
dardize sampling procedures.
Acknowledgements
Financial support is greatly acknowledged from projects GOP-


1.1.1-11-2012 and TAMOP-4.2.2.B-15/1/KONV-2015-0006.
References
Abdul-Khalek, I., Kittelson, D., Graskow, B., Wei, Q., 1998. Diesel Exhaust Particle
Size: Measurement Issues and Trends. http://dx.doi.org/10.4271/980525. SAE
Technical Paper 980525.
Abdul-Khalek, I., Kittelson, D., Brear, F., 1999. The Influence of Dilution Conditions
on Diesel Exhaust Particle Size Distribution Measurements. http://dx.doi.org/
10.4271/1999-01-1142. SAE Technical Paper 1999-01-1142.
Ahlvik, P., Ntziachristos, L., Keskinen, J., Virtanen, A., 1998. Real Time Measurements
of Diesel Particle Size Distribution with an Electrical Low Pressure Impactor.
http://dx.doi.org/10.4271/980410. SAE Technical Paper 980410.
Ban-Weiss, G.A., Lunden, M.M., Kirchstetter, T.W., Harley, R.A., 2010. Size-resolved
particle number and volume emission factors for on-road gasoline and diesel
motor vehicles. Aerosol Sci. 41, 5e12.
Bond, T.C., Doherty, S.J., Fahey, D.W., Forster, P.M., Berntsen, T., DeAngelo, B.J.,
Flanner, M.G., Ghan, S., K€ archer, B., Koch, D., Kinne, S., Kondo, Y., Quinn, P.K.,
Sarofim, M.C., Schultz, M.G., Schulz, M., Venkataraman, C., Zhang, H., Zhang, S.,
Bellouin, N., Guttikunda, S.K., Hopke, P.K., Jacobson, M.Z., Kaiser, J.W.,
Klimont, Z., Lohmann, U., Schwarz, J.P., Shindell, D., Storelvmo, T., Warren, S.G.,
Zender, C.S., 2013. Bounding the role of black carbon in the climate system: a
scientific assessment. J. Geophys. Res. Atmos. 118, 1e173.
Bukowiecki, N., Kittelson, D.B., Watts, W.F., Burtscher, H., Weingartner, E.,
Baltensperger, U., 2012. Real-time characterization of ultrafine and accumula-
tion mode particles in ambient combustion aerosols. Aerosol Sci. 33, 1139e1154.
Braun, A., Shah, N., Huggins, F.E., Kelly, K.E., Sarofim, A., Jacobsen, C., Wirick, S.,
Francis, H., Ilavsky, J., Thomas, G.E., Huffman, G.P., 2005. X-ray scattering and
spectroscopy studies of diesel soot from oxygenated fuel under various engine
load conditions. Carbon 43, 2588e2599.
Burtscher, H., 2005. Physical characterization of particulate emissions from diesel
engines: a review. Aerosol Sci. 36, 896e932.
Burtscher, H., Baltensperger, U., Bukowiecki, N., Cohn, P., HuKglin, C., Mohr, M.,
Matter, U., Nyeki, S., Schmatloch, V., Streit, N., Weingartner, E., 2001. Separation
of volatile and non-volatile aerosol fractions by thermodesorption:
119
instrumental development and applications. Aerosol Sci. 32, 427e442.
Chakrabaty, R.K., Beres, N.D., Moosmüller, H., China, S., Mazzoleni, C., Dubey, M.K.,
Liu, L., Miscchenko, M.I., 2014. Soot superaggregates from flaming wildfires and
their direct radiative forcing. Nat. Sci. Rep. 4, 5508.
Cheng, M.-D., 2013. Classification of volatile engine particles. Aerosol Air Qual. Res.
13, 1411e1422.
ECE, 2000. Uniform Provisions Concerning the Approval of Compression-ignition
(CI) and Natural Gas (NG) Engines as Well as Positive-ignition (PI) Engines
Fuelled with Liquefied Petroleum Gas (LPG) and Vehicles Equipped with CI and
NG Engines and PI Engines Fuelled with LPG, with Regard to the Emissions of
Pollutants by the Engine. Regulation No. 49, Revision 3. E/ECE/324. Economic
Commission for Europe.
Fierz, M., Vernooijb, M.G.C., Burtschera, H., 2007. An improved low-flow thermo-
denuder. J. Aerosol Sci. 38 (11), 1163e1168.
Giechaskiel, B., Manfredi, U., Martini, G., 2014. Engine exhaust solid sub-23 nm
particles: I. Literature survey. SAE Int. J. Fuels Lubr. 7 (3), 950e964. http://
dx.doi.org/10.4271/2014-01-2834.
Giechaskiel, B., Martini, G., 2014. Engine exhaust solid sub-23 nm particles: II.
Feasibility study for particle number measurement systems. SAE Int. J. Fuels
Lubr. 7 (3), 935e949. http://dx.doi.org/10.4271/2014-01-2832.
Guo, X., Nakayama, T., Yamada, H., Inomata, S., Tonokura, K., Matsumi, Y., 2014.
Measurement of the light absorbing properties of diesel exhaust particles using
a three-wavelength photoacoustic spectrometer. Atmos. Environ. 94, 428e437.
Harris, S.J., Maric, M.M., 2002. The role of fragmentation in defining the signature
size distribution of diesel soot. J. Aerosol Sci. 33 (6), 935e942.
Huffmana, J.A., Ziemannb, P.J., Jaynec, J.T., Worsnopc, D.R., Jimeneza, J.L., 2008.
Development and characterization of a fast-stepping/scanning thermodenuder
for chemically-resolved aerosol volatility measurements. Aerosol Sci. Technol.
42 (5).
Ishigoru, T., Takatori, Y., Akihama, K., 1997. Microstructure of diesel soot particles
probed by electron microscopy: first observation of inner core and outer shell.
Combust. Flame 108, 231e234.
Kittleson, D.B., 1998. Engines and nanoparticles: a review. J. Aerosol Sci. 29 (5/6),
575e588.
Lapuerta, M., Rodríguez-Ferna
ndez, J., Agudelo, J.R., 2008a. Diesel particulate
emissions from used cooking oil biodiesel. Bioresour. Technol. 99 (4), 731e740.
Lapuerta, M., Armas, O., Rodríguez-Ferna
ndez, J., 2008b. Effect of biodiesel fuels on
diesel engine emissions. Prog. Energy Combust. Sci. 34 (2), 198e223.
Lloyd, A.C., Cackette, T.A., 2001. Diesel engines: environmental impact and control.
J. Air Waste Manage Assoc. 51 (6), 809e847.
Lin, Y.-C., Lee, C.-F., Fang, T., 2008. Characterization of particle size distribution from
diesel engines fueled with palm-biodiesel blends and paraffinic fuel blends.
Atmos. Environ. 42, 1133e1143.
Liu, G.Z., Vasys, V.N., Kittleson, D.B., 2007. Nuclei-mode particulate emissions and
120 T. Ajtai et al. / Atmospheric Environment 134 (2016) 109e120
their response to fuel sulfur content and primary dilution during transient
operations of old and modern diesel engines. Environ. Sci. Technol. 41 (18),
6479e6483. http://dx.doi.org/10.1021/es0629007.
Mathis, U., Ristima €ki, J., Mohr, M., Keskinen, J., Ntziachristos, L., Samaras, Z.,
Mikkanen, P., 2004. Sampling conditions for the measurement of nucleation
mode particles in the exhaust of a diesel vehicle. Aerosol Sci. Technol. 38 (12),
1149e1160. http://dx.doi.org/10.1080/027868290891497.
Popovicheva, O.B., Kireeva, E.D., Steiner, S., Rothen-Rutishauser, B.,
Persiantseva, N.M., Timofeev, M.A., Shonija, N.K., Comte, P., Czerwinski, J., 2014.
Microstructure and chemical composition of diesel and biodiesel particle
exhaust. Aerosol Air Qual. Res. 14, 1392e1401.
Puzun, A., Wanchen, S., Guoliang, L., Manzhi, T., Chunjie, L., Shibao, C., 2011. Char-
acteristics of particle size distributions about Emissions in A Common-rail
diesel engine with biodiesel blends. Proc. Environ. Sci. 11, 1371e1378.
Rissler, J., Swietlicki, E., Bengtsson, A., Boman, C., Pagels, J., Sandstrom, T.,
Blomberg, A., Londahl, J., 2012. Experimental determination of deposition of
diesel exhaust particles in the human respiratory tract. J. Aerosol Sci. 48, 18e33.
Rothenbacher, S., Messerer, A., Kasper, G., 2008. Fragmentation and bond strength
of airborne diesel soot agglomerates. Part. Fibre Toxicol. 5, 9.
€ nkko
Ro € , T., Virtanen, A., Kannosto, J., Keskinen, J., Lappi, M., Pirjola, L., 2007.
Nucleation mode particles with a nonvolatile core in the exhaust of a heavy
duty diesel vehicle. Environ. Sci. Technol. 41 (18), 6384e6389.
Sakurai, H., Tobias, H.J., Park, K., Zarling, D., Docherty, K.S., Kittelson, D.B.,
McMurry, P.H., Ziemann, P.J., 2003. On-line measurements of diesel nano-
particle composition and volatility. Atmos. Environ. 37 (9e10), 1199e1210.
Sekiguchi, M., Nakajima, T., Suzuki, K., Kawamoto, K., Higurashi, A., Rosenfeld, D.,
Sano, I., Mukai, S., 2003. A study of the direct and indirect effects of aerosols
using global satellite data sets of aerosol and cloud parameters. J. Geophys. Res.
108 (D22).
Steiner, S., Czerwinski, J., Comte, P., Popovicheva, O., Kireeva, E., Müller, L., Heeb, N.,
Mayer, A., Fink, A., Rothen-Rutishauser, B., 2013. . Comparison of the toxicity of
diesel exhaust produced by bio- and fossil diesel combustion in human lung
cells in vitro. Atmos. Environ. 81, 380e388.
Stevanovic, S., Miljevic, B., Madl, P., Clifford, S., Ristovski, Z., 2015. Characterisation
of a commercially available thermodenuder and diffusion drier for ultrafine
particles losses. Aerosol Air Qual. Res. 15, 357e363.
Szabados, Gy, Bereczky, A.,
2015. Comparison tests of diesel, biodiesel and TBK-
biodiesel. Period. Polytech. Mech. Eng. 59 (3), 120e125.
Tinsdale, M., Price, P., Chen, R., 2010. The impact of biodiesel on particle number,
size and mass emissions from a Euro4 diesel vehicle. SAE Int. J. Eng. 3 (1),
597e608.
Trapel, E., Mayer, C., Schulz, C., Roth, P., 2005. Effects of Bio Diesel Injection in a DI
Diesel Engine on Gaseous and Particulate Emission. http://dx.doi.org/10.4271/
2005-01-2204. SAE Technical Paper 2005-01-2204.
Turrio-Baldassarri, L., Battistelli, C.L., Conti, L., Crebelli, R., De Berardis, B.,
Iamiceli, A.L., Gambino, M., Iannaccone, S., 2004. Emission comparison of urban
bus engine fueled with diesel oil and ‘biodiesel’ blend. Sci. Total Environ. 327,
147e162.
Uhrner, U., Zallinger, M., von Lo €wis, S., Vehkama €ki, H., Wehner, B., Stratmann, F.,
Wiedensohler, A., 2011. Volatile nanoparticle formation and growth within a
diluting diesel car exhaust. J. Air Waste Manag. Assoc. 61 (4), 399e408.
XuSheng, Z., Hui, Z., ZongJie, H., ZhiJun, W., LiGuang, L., 2009. Effect of biodiesel on
the particle size distribution in the exhaust of common-rail diesel engine and
the mechanism of nanoparticle formation. Sci. China Ser. E-Tech Sci. 52 (9),
2773e2778. http://dx.doi.org/10.1007/s11431-009-0100-x.