Fuel 141 (2015) 18

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Effects of oxygenated fuel blends on carbonaceous particulate

composition and particle size distributions from a stationary diesel
engine
Zhi-Hui Zhang, Rajasekhar Balasubramanian
Department of Civil and Environmental Engineering, Faculty of Engineering, National University of Singapore, 1 Engineering Drive 2, E1A 02-19, Singapore 117576, Singapore

h i g h l i g h t s
 Oxygenated fuel blends inuence particulate composition and size distributions.
 They reduced EC and DPM emissions with butanol being the most effective.
 They showed different effects on counts of nanoparticles and lager particles.
 Both the oxygen content and the type of oxygenates affect DPM characteristics.
 The change in thermo-physical properties affects the level of DPM emissions.

a r t i c l e

i n f o

Article history:
Received 23 July 2014
Received in revised form 22 September 2014
Accepted 13 October 2014
Available online 22 October 2014
Keywords:
Diesel engine
Oxygenates
OC/EC
Particle size distributions

a b s t r a c t
A systematic study was conducted to evaluate and compare the effects of blending ve different oxygenated compounds, diglyme (DGM), palm oil methyl ester (PME), dimethyl carbonate (DMC), diethyl adipate (DEA) and butanol (Bu) with ultralow sulfur diesel (ULSD), on engine performance, particulate
mass concentrations, organic (OC) and elemental (EC) carbon fractions of the particles and particle size
distributions from a single cylinder, direct injection stationary diesel engine with the engine working
at a constant engine speed and at three engine loads. A small increase in the brake specic fuel consumption (BSFC) and brake thermal efciency (BTE) was observed with the use of oxygenates blended with
ULSD. All ve oxygenates were found to be effective at reducing particulate mass emissions at medium
and high engine loads, with butanol being the most effective and DGM being the least effective. Analysis
of the relative contribution of changes in the OC and EC emissions to the reduction of particulate matter
indicated that under the same oxygen content, EC made a dominant contribution to the reduction of particulate mass. The results also indicated that reduction in both particle mass and number emissions was
affected not only by the oxygen content, but also by the chemical structure and thermophysical properties of oxygenates as well as engine operating conditions.
2014 Elsevier Ltd. All rights reserved.

1. Introduction
Research on sustainable and cleaner fuels continues to receive
considerable attention because of the motivation to reduce dependence on fossil fuels and to mitigate environmental and health
impacts associated with particulate emissions from both on-road
and non-road diesel engines [1]. Numerous oxygenated compounds in the form of alcohols, ethers, and esters have been investigated as additives to conventional diesel fuel in the context of
reducing soot and diesel particulate matter (DPM) emissions
Corresponding author. Tel.: +65 65165135; fax: +65 67744202.
E-mail address: ceerbala@nus.edu.sg (R. Balasubramanian).
http://dx.doi.org/10.1016/j.fuel.2014.10.023
0016-2361/ 2014 Elsevier Ltd. All rights reserved.

[19]. There is a general consensus that the fuel oxygen content

has a signicant inuence on soot formation and DPM emissions.
However, there are conicting reports on the relative impacts of
different oxygenated functional groups [29]. For example,
Miyamoto et al. [2], Choi and Reitz [3], and Cheng et al. [4]
concluded that reduction in soot or particulate mass was mainly
dependent on the percentage of oxygen in the fuel blend while
other researchers reported that apart from the oxygen content,
the chemical structure of oxygenates also inuenced the amount
and toxicity of DPM emissions [510]. Moreover, different kinds
of oxygenates have different thermophysical properties, which
tend to affect not only the fuel injection and combustion processes
but also the DPM emissions [11,12]. In addition, even with the

Z.-H. Zhang, R. Balasubramanian / Fuel 141 (2015) 18

same oxygen content, the extent of soot and DPM emissions reduction with oxygenated fuel blends may also be dependent on the
engine types and operating conditions [3,12]. However, no systematic studies have been conducted to make a comparative evaluation of several oxygenates with a range of thermophysical
properties on DPM emissions from stationary diesel engines, such
as diesel generator sets. These engines are widely used, and emit a
substantial fraction of DPM on a global level because they have
limited emission control measures. For example, the US Environmental Protection Agency (EPA) estimates that non-road diesel
engines contribute to about 44% of the DPM emissions nationwide
[13].
DPM consists mostly of carbonaceous materials (total carbon,
TC), which are often classied as elemental carbon (EC) and organic
carbon (OC) [14]. EC stems from fuel droplet pyrolysis in the fuelrich zone under high pressure and temperature, while OC originates from unburned fuels, lubricating oil, and combustion
byproducts [14]. The previous work reported in the literature
mainly focused on studying the soot or DPM emissions [28], with
little investigation on the variation of EC and OC in response to the
types and amounts of oxygenates used [15]. Furthermore, from the
environmental perspective, EC suspended in the atmosphere can
effectively absorb solar radiation, thus affecting the energy redistribution and global climate [16], while some semi-volatile hydrocarbons in the OC fraction are suspected human carcinogens [17].
To improve our understanding of the environmental and health
effects of oxygenates, there is a strong need to conduct a systematic investigation of changes in carbonaceous matter composition
of particulate emissions when the petroleum diesel blended with
oxygenates is used in diesel engines.
DPM typically consists of ne particles (PM2.5) inclusive of a
large number of ultrane particles (UFP) and nanoparticles (NP)
with their aerodynamic diameter 6 100 nm and 50 nm, respectively. UFP and NP have recently drawn considerable research
attention due to their higher rate of pulmonary deposition, ability
to travel from lung to systemic sites, higher inammation potential, and more biological activities than those of particles with larger sizes [17]. UFP and NP dominate the number concentration
proles of DPM, although they do not make a signicant contribution to the total mass concentration. In previous studies, the use of
oxygenates as blending agents showed reduction in smoke opacity
or DPM emissions, caused by a reduction in larger particles, while
there might be an increase in sub-micron sized particles [12,18].
Thus, the increased emission of the sub-micron particles and the
reduced emission of larger size particles would change particle size
distributions in ambient air as well as the characteristics of the
inhalation exposure [17]. Because of these reasons, the mass-based
emission reduction from oxygenated fuels may not be effective in
mitigating their inuence on the environment and human health.
However, systematic studies and a comparative evaluation of the
potential impacts of different oxygenated fuels on UFP and NP
emission characteristics are currently lacking.
The aim of this study was to characterize and compare the
effects of ve oxygenates (diglyme (DGM), palm oil methyl ester
(PME), dimethyl carbonate (DMC), diethyl adipate (DEA) and butanol (Bu)) with a range of thermophysical properties on carbonaceous particulate emissions from a stationary diesel engine. To
this end, the oxygenates were blended with ultralow sulfur diesel
(ULSD) fuel at 2 wt.% and 4 wt.% oxygen. The inuence of oxygen
content in the blends, the specic types of oxygenates, and the
engine operation parameters on carbonaceous particulate composition, particle number and size distributions were investigated.
The outcome of the study may offer insights into the effect of oxygenated fuel blends on diesel particulate emissions, and prove to
be useful for controlling and regulating DPM emissions from
non-road diesel engines.

2. Experimental section
2.1. Test engine and fuels
A schematic of the experimental system employed in this study
is shown in Supplementary Fig. S1. Experiments were carried out
on a single cylinder, naturally aspirated, four-stroke, direct
injection diesel engine (L70AE, Yanmar Corporation) connected to
a 4.5 kW generator. The diesel engine has a displacement of
296 mL with bore and stroke of 78 mm and 62 mm, a xed speed
of 3000 rpm (revolutions per min). The main specications of the
engine are shown in Supplementary Table S1. In this study, ultralow sulfur diesel fuel with less than 10-ppm (parts per million)
by weight of sulfur was used as the baseline fuel. Five oxygenates,
namely, PME as biodiesel, DMC, DGM, DEA and Bu which represent
different chemical groups, namely ester, ethers and alcohols were
used as oxygenated fuels. The PME used in this study was obtained
from a palm oil-based biodiesel plant in Malaysia operated by
Vance Bioenergy, and the fuel properties were provided by the biodiesel supplier. The other four oxygenate compounds were purchased from Sigma-Aldrich. The major properties of the tested
fuel are listed in Supplementary Table S2. The properties of each
fuel blend can be estimated on the basis of properties of the
baseline fuel and the oxygenated compounds using the method
recommend by Wang et al. [19], with the results shown in Supplementary Table S3. Each fuel blend having the same oxygen concentrations of 2% and 4% by weight was used for this study. The same
amounts of oxygen in fuel blends have been used in many other
studies, and also proved to be effective in reducing DPM emissions
signicantly without the need for engine modication as documented in many previous studies [2023].

2.2. Particulate sampling and testing

A two-stage Dekati mini-diluter (DI-2000, Dekati Ltd.) was used
for diluting the engine exhaust gas for DPM sampling. The diluter
provides primary dilution in the range of 8:16:1, depending on
the engine operating conditions, while the secondary dilution system provides a further dilution of 8:1. The actual dilution ratio was
evaluated based on measured CO2 concentrations in the raw
exhaust, in the background air and in the diluted exhaust. The
CO2 concentration was measured with a non-dispersive infrared
analyzer (MRU Vario Plus, Germany, 0.5% accuracy). This measurement was done for every test, and all data presented in this
article have been dilution-corrected to represent engine-out
condition.
DPM emissions from the rst-stage diluter were collected onto
47 mm Teon lters (Pall Life Sciences, Ann Arbor, MI) and precombusted (650 C for 12 h) 47 mm quartz ber lters (Whatman,
USA), by using two Mini-Vol particulate samplers (Air metrics Ltd.;
5 L min 1 ow rate), respectively. Particles collected on the Teon
lters were used for gravimetric analysis, while those collected on
quartz ber lters were processed for subsequent OC/EC analysis.
Before and after sampling, the lters were allowed to equilibrate
in a humidity-controlled cabinet at 25 C and 30% relative humidity (RH), and weighed using a microbalance (Sartorius MC5, accuracy of 1 lg) for quantifying total particulate mass emissions.
After being weighed, the lters were kept in glass petri dishes
and stored under refrigeration at 20 C for the subsequent analysis. A thermal/optical carbon aerosol analyzer (Sunset Labs, Forest
Grove, Oregon, USA) was used to quantify EC and OC emissions
according to the NIOSH 5040 reference method [24]. The number
concentration and size distribution of particulate emissions in
the secondary dilution stage were measured by a Fast Mobility Particle Sizer (FMPS, Model 3091, TSI Incorporated, USA) in the size

-1
Bu4

Bu2

DEA4

DEA2

DMC4

DMC2

PME4

Fuel type
5

(b) 50% load

BSFC
BTE

3
2
1

0
-1
Bu4

Bu2

DEA4

DEA2

DMC4

DMC2

PME4

PME2

DGM4

DGM2

Fuel type
5

(C) 75% load

BSFC
BTE

4
3
2
1
0
-1
-2

Bu4

Bu2

DEA4

DEA2

DMC4

DMC2

PME4

PME2

DGM4

DGM2

For each testing mode, the volumetric ow rates of the fuels

were measured and then converted into the mass consumption
based on the density of each fuel. The brake specic fuel consumption (BSFC) and the brake thermal efciency (BTE) were then calculated and the percentage changes in them compared to that of the
baseline fuel are shown in Fig. 1. The BSFC generally increased with
an increase in oxygen content in the blends, which is consistent
with the reduction of the lower heating value (LHV) of the fuel
with addition of the oxygenates. Meanwhile, for all test conditions,
the addition of oxygenates showed an increase in BTE due to the
improvement of the combustion process caused by oxygen enrichment. As shown in Supplementary Table S3, compared to ULSD, the
LHV from PME-ULSD blends was only slightly reduced. Therefore, a
minor reduction in the LHV combined with the improved BTE from
PME-ULSD blends resulted in a moderate increase in the BSFC only
with 4% oxygen addition; for 2% oxygen addition there was either
no change or a minor decrease. However, for DGM, DMC, DEA and
Bu, the improved BTE combined with somewhat higher reduction
in the LHV with their addition to ULSD led to higher BSFC with
an increase in the oxygen content in the fuel blends. Overall, variations in BSFC among the fuel blends were less than 5% because the
addition of the oxygenates did not change the LHV of ULSD significantly, as shown in S3. Ren et al. [20] investigated the combustion
and emissions of a direct injection diesel engine fuelled with oxygenated fuel blends, including DGM, DMC and DEA investigated in
this study. Their results showed a slight increase in both BSFC and
BTE with an increase in oxygen content in blends. A similar variation of BSFC and BTE with oxygenated fuel blends was also
reported by Lin et al. [21], Di et al. [22], Dogan [23], Cheung
et al. [25], and Zhu et al. [26] , while using diesel blended with different fractions of PME, DGM, DMC and DEA, and Bu, respectively.
As proposed by Ren et al. [20] and subsequently conrmed by Cheung et al. [25] and Zhu et al. [26], the effective release of heat with
the oxygenated fuel blends contributes to the increase in BTE.
Moreover, the long ignition delay caused by DMC, DEA and Bu
addition results in improved combustion of fuels in the premixed
mode and thus in the enhancement of the BTE [23,25,26].

BSFC
BTE

PME2

3.1. Engine performance

(a) 25% load

DGM4

3. Results and discussion

DGM2

Experiments were performed at a constant engine speed of

3000 rpm and at three engine loads, corresponding to 25%, 50%
and 75% related power, respectively. At each mode of operation,
the engine was allowed to run for at least half an hour until the
exhaust gas temperature and the power output attained steadystate values; these data were measured subsequently. For particulate sampling and analysis, three tests were carried out for each
operating condition and the results were found to agree with each
other at 95% condence level. For determining particle number
concentrations and size distributions, ve measurements were
taken at each mode of operation and the average values are presented. The relative standard deviations ranged from 1.8% to
2.5%, 2.0% to 3.5%, 3.0 % to 3.7%, 4.6% to 5.8%, and 2.5% to 3.1%
for fuel consumption, particulate mass, EC, OC, and particle number, respectively. Results obtained from different fuels were compared using the two-sided Students t-test to examine if they are
signicantly different from each other at 95% condence level.

Percentage change in BSFC (%) and BTE (%)

2.3. Test procedure

Percentage change in BSFC (%) and BTE (%)

range of 5.6560 nm. In this setup, two diluters were used in series, with the rst stage being heated to 190 C in order to minimize
thermophoretic deposition.

Percentage change in BSFC (%) and BTE (%)

Z.-H. Zhang, R. Balasubramanian / Fuel 141 (2015) 18

Fuel type
Fig. 1. Effect of oxygenated fuel blends on the percentage change in brake specic
fuel consumption (BSFC) and brake thermal efciency (BTE).

However, the higher latent heat of evaporation of DMC, DEA and

Bu could lower the combustion temperature with reduction of
the BTE, especially at lower engine load with lower in-cylinder
combustion temperature. The different characteristics of fuels
combined with their unique combustion properties led to variations of the BTE of the fuel blends with no systematic trends in this

Z.-H. Zhang, R. Balasubramanian / Fuel 141 (2015) 18

study. Overall, the results on engine performance indicated that

the use of these oxygenated fuel blends does not require signicant
modication of existing diesel engines. However, minor variations
in the fuel delivery system may be needed due to the unique fuel
properties of oxygenates, including that of changing the materials
of polymers parts and enhancing wear resistance of moving and
rotating parts.
3.2. Emissions of carbonaceous matter
DPM has been classied into three fractions in this study; EC,
OC and other non-carbonaceous materials (such as sulfates,
nitrates, metals and ash). The variation of their concentrations
50

0.15
Others
EC
OC

40

0.00
20
-0.05
10

-0.10
-0.15
Bu4

Bu2

DEA4

DEA2

DMC4

DMC2

PME4

PME2

DGM4

DGM2

Baseline

Fuel type
0.4

75

(b) 50% load

Others
EC
OC

60

0.2

45
0.0
30

EC/OC

Concentration (mg/m3)

EC/OC

0.05
30

-0.2

15

-0.4
Bu4

Bu2

DEA4

DEA2

DMC4

DMC2

PME4

PME2

DGM4

DGM2

Baseline

Fuel type

120
Others
EC
OC

(c) 75% load

100

Concentration (mg/m3)

0.10

0.9

0.6
80
60

0.3

EC/OC

Concentration (mg/m3)

(a) 25% load

40
0.0
20
-0.3
Bu4

DEA2

DEA4

DMC4

DMC2

PME2

PME4

DGM2

DGM4

Baseline

Bu2

Fuel type

Fig. 2. Effect of oxygenated fuel blends on DPM (DPM = EC + OC + Others) and

carbonaceous species (EC (elemental carbon) and OC (Organic carbon)) emissions.

and EC/OC ratios are shown in Fig. 2. In general, both OC and EC,
and thus the TC (TC = OC + EC) and DPM concentrations increased
with an increase in engine load for each fuel, which is consistent
with the higher fuel usage, lower air/fuel ratio, and higher pressure
and temperature at higher loads [14]. Moreover, the increase in EC
concentrations with the increased engine load was higher than
that of OC concentrations, resulting in the increased EC/OC ratios
with the engine load. The addition of oxygenates showed no significant change in these emissions at the low engine load. However, at
medium and high engine loads, the reduction of these emissions
became signicant (P < 0.05) with the higher oxygen content in
the blends resulting in the lower emissions of OC, EC and DPM.
In terms of an average reduction of DPM across all test conditions,
Bu was most effective, with an average reduction of 10.9% and
19.4%, corresponding to 2% and 4% oxygen addition, followed by
DEA at 9.4% and 18.2%, DMC at 7.8% and 15.4%, and PME at 6.6%
and 13.5%. DGM was least effective across all three test conditions,
with an average reduction of 3.9% and 8.6% with 2% and 4% oxygen
contents, respectively. A similar effect of oxygenated fuel blends on
the EC emissions was also observed in this study. Specically, the
corresponding reduction in EC emissions was 37.3% and 56.3% for
Bu, 29.5% and 44.9% for DEA, 30.3% and 44.8% for DMC, 19.1%
and 29.7% for PME, and 2.6% and 6.1% for DGM. However, this
study showed a modest change in OC emissions between different
fuel blends. The OC emission reduction was in the range of 3.99%
and 8.713.1% with 2% and 4% oxygen in the blends, respectively.
Therefore, the difference in DPM emission reduction can be caused
predominantly by the difference in EC emission reduction. Moreover, the different changes in EC emissions coupled with the modest reduction in OC emissions may contribute to the variation of
EC/OC ratios with different fuel blends. For each engine load,
DGM showed no signicant changes in EC/OC ratios, whereas other
kinds of oxygenated fuel blends showed lower EC/OC ratios, indicating the higher proportion of adsorbed semi-volatile organics
in the particulate matter.
Many researchers have reported that the fuel oxygen content is
one of the main factors affecting EC (soot) and DPM emissions. For
example, a numerical study by Cheng et al. [4] provided a broad
explanation on the relationship between the oxygenates and the
soot (EC) formation as follows. First, compared with diesel, the
pyrolysis and decomposition of oxygenates could reduce the soot
precursors. Second, the increase in free radical concentrations
(e.g. O, OH, etc. produced by oxygenates) promotes the carbon oxidation to CO and CO2 within the premixed ame zone, thus limiting the carbon available for the formation of soot precursor species.
Last, the high radical concentrations serve to limit aromatic ring
growth and soot particle inception. Moreover, Cheung et al. [27]
indicated that oxygenated additives could inhibit the in-cylinder
EC production by both disrupting the carbon chain development
and promoting oxidation. The results obtained in this study support these ndings reported by Cheng et al. [4] and Cheung et al.
[27]. Specically, the higher oxygen content in the blends not only
leads to the lower concentration of soot precursors, but also
reduces the potential functions of these soot precursors to form
soot. Thus, the higher oxygen content in the blends results in the
lower EC and DPM emissions. In addition, the diluted aromatics
and sulfur concentrations resulting from the addition of oxygenates can also contribute to lower soot and DPM emissions
[12,19,28]. In this study, both the aromatic and sulfur contents in
the baseline fuel were quite low as shown in supplementary
Table S2, so only a marginal decrease in them was observed with
the use of oxygenated fuel blends. Therefore, changes in aromatics
and sulfur contents are not likely to be a major factor in the reduction of DPM emissions for the present study.
The results from this study also suggest that with the same oxygen content in blends, the alcohol group of Bu appears to be more

Z.-H. Zhang, R. Balasubramanian / Fuel 141 (2015) 18

effective than the ester group of PME, DMC and DEA, which is in
turn better than the ether group of DGM in reducing EC and DPM
emissions. This trend with the alcohol functional group being more
effective in inhibiting soot formation than that of ether and ester
groups is consistent with the previous observations reported by
Pepiot-Desjardins et al. [8], Wang et al. [19], Wang et al. [9], and
Zhang et al. [10]. However, Mueller et al. [5] and Westbrook
et al. [7] concluded that the ether structure was more effective in
reducing the formation of soot precursors than the ester structure.
As explained by Buchholz et al. [6] and Westbrook et al. [7], the
two oxygen atoms in esters are linked to the same carbon atom
and are more likely to form CO2, which means that not every O
atom in these compounds is able to eliminate a carbon atom from
the soot precursor pool, resulting in a higher fraction of the
remaining carbon atoms available to produce soot. This ester
decomposition mechanism was also proposed by Szybist et al.
[29] while discussing the outcome of their combustion experiments with diesel engines. However, the experimental data of this
study does not support this theory based on the fact that DGM
appeared to be less effective at reducing soot and DPM than all
tested ester group oxygenates. Moreover, both DMC and DEA were
considerably more effective than PME, although they belong to the
same functional group of ester. These ndings reveal that apart
from the oxygen content of the blends, the functional group of oxygenates, the extent of EC and DPM emissions reductions is also
affected by the other fuel properties and engine types.
An important nding of this study is that under the same oxygen content in blends, there is a high correlation between cetane
number of fuel blends and the concentrations of EC emissions,
and between cetane number and EC/OC ratios, as shown in Table 1.
This nding implies that under the same oxygen content in blends,
the oxygenates with the lower cetane number would lead to lower
EC emissions, but higher OC fractions in particles. For diesel
engines, it is known that the fuel burned in the diffusion mode is
much related to EC (soot) emission [28,30]. Variations of fuel
cetane numbers would have a direct impact on ignition delay
and the ratio between premixed and diffusion combustion, and
therefore leading to a change in EC emissions. Specically, oxygenated fuel blends with a low cetane number lead to an increase in
ignition delay together with an increase in the amount of fuel
burned in the premixed combustion phase, which also reduces
the amount of fuel blends burned in the diffusion mode and hence
further reduces the amount of soot formed. Boot et al. [31] and
Wang et al. [19] reported that the use of a cetane booster had a
negative inuence on the soot emissions because of a reduced
non-sooting premixed burn duration brought about by the shorter
ignition delay. Zhang et al. [30] investigated the relationship
between particulate mass emission and premixed combustion
ratios under different engine operating conditions. Their results
revealed that there is a decrease of particulate mass concentration
with increase of the premixed to diffusion combustion ratio for
each engine load. Our previous results and the results from the current study indicate that the lower cetane number has a positive

effect on EC emission reduction. Moreover, as explained by Gill

et al. [28], although the ether should in theory be more effective
than the ester, the effect of a higher cetane number may result in
an overall diminished benet. Therefore, DGM appeared to be less
effective at reducing EC and DPM than all tested ester group oxygenates used in this study. Similarly, PME has a higher cetane number than that of other kinds of ester group oxygenates, such as
DMC and DEA, and is thus less effective in EC and PM emission
reduction. Ren et al. [20], Cheung et al. [25] and Zhu et al. [26] also
found the increased ignition delay combined with the increase in
premixed combustion promoted the PM emission reduction from
both diesel-DMC and diesel-DEA blends, respectively.
There are several factors that can be used to explain the variation of OC emissions and the relationship between cetane number
and EC/OC ratios. On one hand, the oxygen content in blends might
lead to lower OC emissions due to the enhanced combustion.
Furthermore, the lower EC emissions for oxygenated fuel blends
provided a small surface area for adsorption of the volatile materials that comprised the OC, resulting in lower OC emissions. One the
other hand, the lower cetane number of fuel blends would mean
that they have a lower auto-ignition ability and leads to higher
unburned hydrocarbon (HC) emissions and thus the higher OC
emissions [19]. In this study, it is possible that the decreased ignitability and ammability, caused by the lower cetane number of
oxygenates, was mainly responsible for lower EC/OC ratios that
resulted from DMC, DEA and Bu addition. It should be emphasized
that apart from the oxygen content in fuel blends and their cetane
numbers, the other physical properties of each fuel could also inuence the OC emissions. For example, Bagley et al. [32] and Knothe
et al. [33] concluded that fuels with the higher boiling point, higher
viscosity and higher density could lead to a poor air/fuel mixture
and incomplete combustion, and thus higher OC emissions. The
oxygenates investigated in this study enhanced one or more of
these adverse factors, leading to the lower EC/OC ratios.
The results from this study also revealed that the addition of
oxygenates was more benecial at high engine load conditions.
At low load condition, DPM emissions were not signicantly
reduced with oxygenates addition. Two factors likely contributed
to this observation. The rst factor is that less fuel is injected at
low engine load and therefore less fuel is burned during the mixing-controlled phase of combustion. Since soot formation occurs
primarily during this mixing-controlled combustion phase, the
effect of these oxygenates on DPM emission would be less pronounced at this engine load. The second factor is that the overall
(absolute) level of soot at low engine load was small. The oxygenated blends produced higher hydrocarbon (HC) emissions, and the
contribution to the DPM mass from adsorbed or condensed HCs
may have masked a small reduction in DPM mass.
3.3. Particle number concentrations and size distributions
In this study, the particle number size distribution in the size
range of 5.6560 nm was studied (Fig. 3). In order to analyze the

Table 1
Statistical correlation between cetane number (CN) and EC or EC/OC ratios.
25% load
2% oxygen
CN vs EC

CN vs EC/OC

R2
P-value
Regression slope
y-intercept
R2
P-value
Regression slope
y-intercept

0.798
0.041
0.145
5.559
0.789
0.044
0.007
0.253

50% load
4% oxygen
0.880
0.018
0.112
4.138
0.880
0.018
0.005
0.192

75% load

2% oxygen

4% oxygen

2% oxygen

4% oxygen

0.891
0.016
0.516
18.416
0.876
0.019
0.016
0.558

0.902
0.013
0.373
12.510
0.886
0.017
0.012
0.395

0.806
0.039
0.825
13.503
0.703
0.087
0.013
0.014

0.855
0.024
0.686
10.628
0.732
0.065
0.014
0.098

5.0 x 1013

(b) 50% load

44
Dp>100 nm
Dp=50-100 nm
Dp=20-50 nm
Dp<20 nm

4.0 x 1013
3.0 x 1013

33
22

2.0 x 1013
11

1.0 x 1013

0.0
Bu4

Bu2

DEA4

DEA2

DMC4

PME4

DMC2

PME2

DGM4

Baseline

DGM2

100

GMD (nm)

55

GMD (nm)

dN/dlog(Dp) (#/m3 )

Fuel type

6.0 x 1013

0.0
10

Bu4

9.0 x 10

0
DEA4

12

0.0
Bu2

1.8 x 10

DEA2

13

1.0 x 10 13

DMC4

2.7 x 10

16

DMC2

13

24

PME4

3.6 x 1013

32

2.0 x 10 13

PME2

Baseline
DGM2
DGM4
PME2
PME4
DMC2
DMC4
DEA2
DEA4
Bu2
Bu4

Dp>100 nm
Dp=50-100 nm
Dp=20-50 nm
Dp<20 nm

3.0 x 10 13

DGM4

(a) 25% load

40

(a) 25% load

DGM2

4.5 x 1013

4.0 x 10 13

Baseline

variation of UFP and NP in detail, the particles were further classied into four groups: particles with diameter < 20, 2050, 50100,
and >100 nm. The effect of the fuel blends on particle number concentrations in each group is shown in Fig. 4. As can be seen from
Fig. 3(a)(c), the size of particle emissions was mostly below
100 nm. At low engine load, the particle size distributions were
bimodal, showing clearly a distinctive nucleation mode with the
peak diameters of 810 nm, and an accumulation mode with the
peak diameters of 4552 nm, respectively. With the increased
engine load, the relative intensity of the nucleation mode
decreased while that of the accumulation mode increased, resulting in a shift of the particle size distribution from bimodal to single
mode. Furthermore, with an increase in engine load, regardless of
the fuel used, the particle size distributions shifted upwards and
towards larger size, which may be responsible for an increase in

Particle number concentration (#/m3 )

Z.-H. Zhang, R. Balasubramanian / Fuel 141 (2015) 18

Particle number concentration (#/m3)

Fuel type

Baseline
DGM2
DGM4
PME2
PME4
DMC2
DMC4
DEA2
DEA4
Bu2
Bu4

dN/dlog(Dp) (#/m3)

(b) 50% load

7.2 x 1013
5.4 x 1013
3.6 x 1013
1.8 x 1013

dN/dlog(Dp) (#/m3)

10

100

Diameter (nm)
Fig. 3. Effect of oxygenated fuel blends on particle size distributions.

0
Bu4

0.0

0.0

Bu2

4.0 x 1013

10

DEA4

8.0 x 1013

2.0 x 10

DMC4

1.2 x 1014

20

13

DEA2

1.6 x 1014

40
30

4.0 x 1013

DMC2

Baseline
DGM2
DGM4
PME2
PME4
DMC2
DMC4
DEA2
DEA4
Bu2
Bu4

(c) 75% load

50

6.0 x 1013

PME2

2.0 x 10

Dp>100 nm
Dp=50-100 nm
Dp=20-50 nm
Dp<20 nm

8.0 x 1013

PME4

14

60

(c) 75% load

DGM4

100

Diameter (nm)

1.0 x 10

DGM2

10

70

14

Baseline

0.0

1.2 x 1014

GMD (nm)

9.0 x 1013

Particle number concentration (#/m3)

Diameter (nm)

Fuel type

Fig. 4. Effect of oxygenated fuel blends on size-segregated particle number

concentrations.

total particle number concentrations and geometric mean diameter (GMD), as shown in Fig. 4. Characteristics of the particle size
distributions and their dependence on operating conditions of
the engine observed in this study are similar to those previously
reported [34,35]. For each engine load condition, the size distribution curves shifted upwards after the addition of DGM and PME in
the baseline fuel, indicating an increase in the total particle concentration as shown in Fig. 4. In contrast, the particle size distribution proles in general became atter, with the peak shifting
slightly towards smaller size with an increase in DMC, DEA and
Bu in the blends, indicating a decrease of the total particle number
concentrations and GMD, as shown in Fig. 4. However, addition of
DMC, DEA and Bu to the baseline fuel caused an obvious increase in
particles with the diameter less than 20 nm at low engine load.

Z.-H. Zhang, R. Balasubramanian / Fuel 141 (2015) 18

As can be seen in Fig. 4, the particles of less than 20 nm

decreased while the particles of larger than 20 nm increased,
which led to an increase in the total particles counts with the
engine load for all the fuels used in this study. These variations
also caused an increase in GMD with engine load, as shown in
Fig. 4. For each engine load, the total particle number concentration with the addition of DGM and PME was larger than that of
the diesel fuel, which is in line with the ndings from previous
studies reported by Di et al. [22] and Zhu et al. [18], respectively.
However, with DMC, DEA and Bu addition, the total number concentration and the GMD of the particles were reduced considerably. Compared to the baseline fuel, the average reduction of the
total particle number concentrations on different engine loads
was 7.9% and 3.4%, 13.6% and 23.9%, 25.1% and 36.1% for DMC,
DEA and Bu with 2% and 4% oxygen in blends, respectively. The
corresponding percentage increase of total particle number emissions was 6.4% and 18.3%, and 19.0% and 34.8% for addition of
DGM and PME, respectively. Moreover, the reduction of total particle number emissions with the addition of DMC, DEA and Bu is a
result of a signicant reduction in the number of particles larger
than 20 nm, especially at medium and high engine loads. The
reductions in large diameter particles are responsible for the
decrease in particulate mass, as shown in Fig. 2, as they are more
inuential on mass. As suggested by Cheung et al. [25], Zhu et al.
[26] and Sukjit et al. [1], the reduced carbon content coupled with
an increase in oxygen content, the increased premixed combustion
and reduced diffusion combustion with DMC, DEA and Bu addition,
may contribute to the reduction of soot number emissions and
hence the total particle number concentrations. The reduction in
particle number concentration may also reduce the occurrence of
coagulation and agglomeration of the particles, leading to a reduction in the GMD. In contrast, the higher cetane number of DGM
and PME may result in the lower premixed combustion and thus
more particle formation and emissions. In addition, the higher viscosity and the higher boiling point of biodiesel would lead to the
variations of the fuel injection pressure of the mechanical fuel
injection system used in this study [18], and atomization characteristics, which would also affect the particle size emission
characteristics.
The emissions of small nanoparticles (<20 nm) show very different trends compared to those of larger particles. At low engine
load, DGM and PME addition showed a slight decrease in particles
of less than 20 nm, while DMC, DEA and Bu addition showed an
increase in these particles. At medium and high engine loads, these
oxygenated fuel blends showed only a slight or no signicant
change in particles of less than 20 nm. The results indicated that
the reduction of total particle number concentration as well as
the GMD associated with DMC, DEA and Bu addition was a result
of a signicant reduction in the number of larger particles. These
ndings are in line with those of Schneider et al. [36], Kittelson
et al. [37], and Sukjit et al. [1] who suggested that the soluble
organic fraction (SOF) in the particles may be transferred from
the gas phase to the particle phase by two paths: nucleation to
form new particles, or adsorption onto or absorption into existing
particles. If the number of accumulation mode particles is extremely reduced, there would not be sufcient surfaces available for
volatiles adsorption, and the high concentration of gaseous volatiles would lead to the formation of a nuclei mode. Thus, the
increase in the nanoparticle range for the DMC, DEA and Bu addition can be caused by a decrease in large particle emissions produced by the cleaner burning properties of the fuels. At medium
and high engine loads, the amount of accumulation mode particles
for each fuel was higher coupled with higher SOF than that of diesel, with oxygenates addition, which leads to the decrease in the
nanoparticles concentration but no signicant changes in them
between different fuels.

From the observations of this study, we can conclude that under

the same oxygen content in blends, DGM and PME addition
showed less effectiveness in DPM emission reduction because both
of them led to a lower DPM mass reduction efciency while showing a higher particle number concentration, compared to that of
DMC, DEA and Bu addition. This drawback is not only applicable
to DGM and PME blends but also to other oxygenates at low engine
load. The results from this study also reveal that when designing
the oxygenated fuel blends, in addition to the fuel oxygen content
and the molecular structure of the oxygenates, the cetane number
and other fuel properties should also be considered. The change in
thermophysical properties caused by the oxygenates addition
tends to affect the level of the EC, OC and particle number emissions. More research is warranted to identify and quantify specic
chemical groups of OC such as alkanes, polycyclic aromatic hydrocarbons (PAHs) and polar organic compounds and reduce the emission of toxic organic compounds such as PAHs by taking into
consideration combustion chemistry, types and amounts of oxygenates used and thermophysical properties of blended fuels.

4. Conclusions
The following major conclusions are drawn from this study.
(1) The brake specic fuel consumption was marginally
increased with the addition of oxygenates to the baseline
fuel, while a slight increase in brake thermal efciency was
observed.
(2) The use of oxygenated fuel blends caused a signicant
reduction in DPM emissions at medium and high engine
loads. This reduction can mainly be attributed to the EC
emission reduction. However, with the exception of DGM,
other blended fuels increased the OC fraction in particles.
(3) Under the same oxygen content in fuel blends, butanol addition gave the highest DPM and EC emission reduction as well
as the lowest particle number concentration, followed by
DEA and DMC addition. However, both DGM and PME addition showed a lower reduction potential of DPM mass emissions, but also a higher particle number concentration.
(4) The overall reduction of particle number emissions with the
use of oxygenated fuel blends appears to be mainly due to
the reduction of larger particles emissions, while the emissions of smaller particles are more associated with fuel properties than with the amounts of oxygenates.

Acknowledgements
Zhi-Hui Zhang would like to thank the SingaporePeking
Oxford Research Enterprise (COY-15-EWI-RCFSA/N197-1) for
providing scholarship in support of his doctoral study.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.fuel.2014.10.023.
References
[1] Sukjit E, Herreros JM, Piaszyk J, Dearn KD, Tsolakis A. Finding synergies in fuels
properties for the design of renewable fuels-hydroxylated biodiesel effects on
butanol-diesel blends. Environ Sci Technol 2013;47(7):353542.
[2] Miyamoto N, Ogawa H, Nurun NM, Obata K, Arima T. Smokeless, low NOx, high
thermal efciency, and low noise diesel combustion with oxygenated agents as
main fuel. SAE Paper, 980506.

Z.-H. Zhang, R. Balasubramanian / Fuel 141 (2015) 18

[3] Choi CY, Reitz RD. An experimental study of the effects of oxygenated fuel
blends and multiple injection strategies on DI diesel engine emissions. Fuel
1999;78(11):130317.
[4] Cheng AS, Dibble RW, Buchholz BA. The effect of oxygenates on diesel engine
particulate matter. SAE Paper, 200201-1705.
[5] Mueller CJ, Pickett LM, Siebers DL, Pitz WJ, Westbrook CK, Martin GC. Effects of
oxygenates on soot processes in DI diesel engines: experiments and numerical
simulations. SAE Paper, 2003-01-1791.
[6] Buchholz BA, Mueller CJ, Upatnieks A, Martin GC, Pitz WJ, Westbrook CK. Using
Carbon-14 isotope tracing to investigate molecular structure effects of the
oxygenate dibutyl maleate on soot emissions from a DI diesel engine. SAE
Paper, 2004-01-1849.
[7] Westbrook CK, Pitz WJ, Curran HJ. Chemical kinetic modeling study of the
effects of oxygenated hydrocarbons on soot emissions from diesel engines. J
Phys Chem A 2006;110:691222.
[8] Pepiot-Desjardins P, Pitsch H, Malhotra R, Kirby SR, Boehman AL. Structural
group analysis for soot reduction tendency of oxygenated fuels. Combust
Flame 2008;154:191205.
[9] Wang X, Cheung CS, Di Y, Huang Z. Diesel engine gaseous and particle
emissions fueled with dieseloxygenate blends. Fuel 2012;94:31723.
[10] Zhang ZH, Balasubramanian R. Physicochemical and toxicological
characteristics of particulate matter emitted from a non-road diesel engine:
comparative evaluation of biodieseldiesel and butanoldiesel blends. J
Hazard Mater 2014;264:395402.
[11] Stoner M, Litzinger T. Effects of structure and boiling point of oxygenated
blending compounds in reducing diesel emissions. SAE Paper, 1999-01-1475.
[12] Salvi AA, Assanis D, Filipi Z. Impact of physical and chemical properties of
alternative fuels on combustion, gaseous emissions, and particulate matter
during
steady
and
transient
engine
operation.
Energy
Fuels
2012;26(7):423141.
[13] US EPA. Final regulatory impact analysis: control of emissions from nonroad
diesel engines, May 2004, EPA420-R-04-007.
[14] Liu Z, Lu M, Brich MW, Keener TC, Khang S, Liang F. Variations of the
particulate carbon distribution from a nonroad diesel generator. Environ Sci
Technol 2005;39:78404.
[15] Xu Z, Li X, Guan C, Huang Z. Characteristics of exhaust diesel particles from
different oxygenated fuels. Energy Fuels 2013;27(12):757986.
[16] Bond TC, Doherty SJ, Fahey DW, Forster PM, Berntsen T, DeAngelo BJ, et al.
Bounding the role of black carbon in the climate system: a scientic
assessment. J Geophys Res 2013;118:5380552.
[17] Wichmann HE. Diesel exhaust particles. Inhalation Toxicol 2007;19:2414.
[18] Zhu L, Zhang W, Liu W, Huang Z. Experimental study on particulate and NOx
emissions of a diesel engine fueled with ultralow sulfur diesel, RME-diesel
blends and PME-diesel blends. Sci Total Environ 2010;408(5):10508.
[19] Wang J, Wu F, Xiao J, Shuai S. Oxygenated blend design and its effects on
reducing diesel particulate emissions. Fuel 2009;88(10):203745.
[20] Ren Y, Huang Z, Miao H, Di Y, Jiang D, Zeng K, et al. Combustion and emissions
of a DI diesel engine fuelled with diesel-oxygenate blends. Fuel
2008;87(12):26917.

[21] Lin YC, Lee WJ, Hou HC. PAH emissions and energy efciency of palm-biodiesel
blends fueled on diesel generator. Atmos Environ 2006;40(21):393040.
[22] Di Y, Cheung CS, Huang Z. Experimental investigation of particulate emissions
from a diesel engine fueled with ultralow-sulfur diesel fuel blended with
diglyme. Atmos Environ 2010;44(1):5563.
[23] Dogan O. The inuence of n-butanol/diesel fuel blends utilization on a small
diesel engine performance and emissions. Fuel 2011;90(7):246772.
[24] NIOSH. Analytical Method No. 5040. Cincinnati (OH): National Institute for
Occupational Safety and Health; 1995.
[25] Cheung CS, Zhu R, Huang Z. Investigation on the gaseous and particulate
emissions of a compression ignition engine fueled with diesel-dimethyl
carbonate blends. Sci Total Environ 2011;409(3):5239.
[26] Zhu R, Cheung CS, Huang Z, Wang X. Experimental investigation on particulate
emissions of a direct injection diesel engine fueled with diesel-diethyl adipate
blends. J Aerosol Sci 2011;42(4):26476.
[27] Cheung KL, Polidori A, Ntziachristos L, Tzamkiozis T, Samaras Z, Cassee FR,
et al. Chemical characteristics and oxidative potential of particulate matter
emissions from gasoline, diesel, and biodiesel cars. Environ Sci Technol
2009;43(16):633440.
[28] Gill SS, Tsolakis A, Herreros JM, York APE. Diesel emissions improvements
through the use of biodiesel or oxygenated blending components. Fuel
2012;95:57886.
[29] Szybist JP, Boehman AL, Haworth DC, Koga H. Premixed ignition behavior of
alternative diesel fuel-relevant compounds in a motored engine experiment.
Combust Flame 2007;149:11228.
[30] Zhang ZH, Cheung CS, Yao CD. Inuence of fumigation methanol on the
combustion and particulate emissions of a diesel engine. Fuel 2013;111:4428.
[31] Boot MD, Frijters PJM, Klein-Douwel RJH, Baert RSG. Oxygenated fuel
composition impact on heavy-duty diesel engine emissions. SAE Paper,
2007-01-2008.
[32] Bagley ST, Gratz LD, Johnson JH, McDonald JF. Effects of an oxidation catalytic
converter and a biodiesel fuel on the chemical, mutagenic, and particle size
characteristics of emissions from a diesel engine. Environ Sci Technol
1998;32(9):118391.
[33] Knothe G, Sharp CA, Ryan TW. Exhaust emissions of biodiesel, petrodiesel, neat
methyl esters and alkanes in a new technology engine. Energy Fuel
2005;20(1):4038.
[34] Sakurai H, Tobias HJ, Park K, Zarling D, Docherty KS, Kittelson DB, et al. On-line
measurements of diesel nanoparticle composition and volatility. Atmos
Environ 2003;37:1199210.
[35] Vaaraslahti K, Virtanen A, Ristimki J, Keskinen J. Nucleation mode formation
in heavy-duty diesel exhaust with and without a particulate lter. Environ Sci
Technol 2004;38(18):488490.
[36] Schneider J, Hock N, Weimer S, Borrmann S. Nucleation particles in diesel
exhaust: composition inferred from in situ mass spectrometric analysis.
Environ Sci Technol 2005;39(16):615361.
[37] Kittelson DB, Watts WF, Johnson JP. On-road and laboratory evaluation of
combustion aerosols part1: summary of diesel engine results. J Aerosol Sci
2006;37(8):91330.