Fuel 173 (2016) 189–208

Contents lists available at ScienceDirect

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

Review article

Challenges and opportunities of enhancing cold flow properties

of biodiesel via heterogeneous catalysis
Adeel Anwar ⇑, Arthur Garforth
Chemical Engineering and Analytical Science Department, Faculty of Engineering and Physical Science, The University of Manchester, Manchester M60 1QD, UK

h i g h l i g h t s

Lipid derived biofuel is primarily being investigated in two forms: renewable diesel and biodiesel.

The long straight chain fatty acid molecular composition found in naturally occurring oils produces fuel of a lesser quality.

Cold flow properties and oxidative stability are the two major concerns in terms of quality.

Hydroprocessing using heterogeneous catalysis is considered to be the most innovative solution.

Dewaxing (hydrocracking and hydroisomerisation) offers the most enticing prospects.

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

Article history: Concerns over climate change are necessitating the need for alternative greener fuel. Biodiesel is widely
Received 31 August 2015 considered to be a key contributor to alleviating this concern. At present the biodiesel sector faces a num-
Received in revised form 13 December 2015 ber of challenges including technical challenges such as improving the quality of biodiesel to match that
Accepted 13 January 2016
of crude oil derived diesel. One of the key challenges in the quality arena is improving the cold flow prop-
Available online 23 January 2016
erties to enable wider geographic and seasonal operability. This review discusses in depth the various
options which can be deployed to address this issue in an efficient and effective manner, with heteroge-
Keywords:
neous catalysis offering fascinating prospects.
Biofuel
Biodiesel
Ó 2016 Elsevier Ltd. All rights reserved.
Heterogeneous catalysis
Zeolites
Cold flow properties

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
1.1. Biodiesel versus renewable diesel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
1.2. Biodiesel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
1.3. Biodiesel challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
1.3.1. Availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
1.3.2. Sustainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
1.3.3. Economic feasibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
1.3.4. Quality. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
1.4. Biodiesel quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
1.4.1. Oxidative stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
1.4.2. Cold flow properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
1.4.3. Existing solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
2. Challenges and opportunities using heterogeneous catalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
2.1. Dehydrogenation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
2.2. Dewaxing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
2.2.1. Hydrocracking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
2.2.2. Hydroisomerisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201

⇑ Corresponding author.

http://dx.doi.org/10.1016/j.fuel.2016.01.050
0016-2361/Ó 2016 Elsevier Ltd. All rights reserved.
190 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208

3. Supplementary opportunities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205

3.1. Effects of stearic hindrance (HC, FA and FAME) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
3.2. Catalyst type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
3.3. Reactor type. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
3.4. Precise isomer identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
3.5. MP and OS for multiple branched molecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

1. Introduction low cost petroleum fuel production caused a decline in biofuel

The mass utilisation of biofuels from vegetable oil in internal
combustion engines dates back to the early 20th century (1920–
1930) in particular during WWII, with reported use from around
the globe including Germany, Argentina, Japan, Belgium, Italy,
France, the United Kingdom, Portugal, and China whom all tested
and used various types of biofuels [5]. However the emergence of
Fig. 1. Major drivers for the resurgence of biofuels. Adapted and modified from
Azapagic et al. [8].
research and made investigation into the biofuel production infras-
tructures redundant. In recent times concern over the depletion of
the finite fossil fuel reserves and with the transport sector now
accounting for approximately 25% of the overall energy consump-
tion and CO2 equivalent emissions produced globally a resurgence
is being witnessed in biofuel research and utilisation. Political,
environmental and social factors, highlighted in Fig. 1, are the pri-
mary instigators for the increasing demand for biofuels. When it
comes to the types of biofuel, numerous exist which can be pro-
duced from various feedstock using a variety of processes. The type
of processing required is largely determined by the desired product
as well as the composition of the feedstock, as shown in Fig. 2
below. Of the various types of biofuel, according to a recent study,
ethanol from lignocellulosic waste, biodiesel from waste oil and
palm oil are touted as the most economically competitive alterna-
tives [6]. The overall price of biofuels usually take into considera-
tion the costs associated with the feedstock, conversion process,
fuel distribution and value of co-products; with Table 1 entailing
some typical retail prices of international biofuels, courtesy of
IEA [7]. The units of price are displayed in US dollars per litre of
gasoline equivalent (USD/lge) due to the disparity in the energy
content of biofuels in comparison to that of fossil fuels.
The focus of the proposed research is on biofuels produced from
lipids which can be extracted from various forms of waste and
crops [9,10]. Lipids are typically triglycerides composed of a series

Fig. 2. Biofuel options based on different feedstock and processes. Adapted from Yue et al. [9].
 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208 191

Table 1 of fatty acids; a commonly found fatty acid is stearic acid (C18:0)
Biofuel retail prices (untaxed) and projected prices (USD/lge) till 2020. Adapted from which has the dimensions of approximately 24.5 by 2.5 Å as illus-
IEA-ESTAP and IRENA [7].
trated in Fig. 3 [11]. The molecular dimensions of fatty acids are of
2010 2020 extreme relevance when it comes to heterogeneous catalytic appli-
Sugarcane ethanol 0.62–0.64 0.6–0.7 cations involving porous materials such as zeolites; this is due to
Corn ethanol 0.71–0.76 0.7–0.8 their respective pore openings- for instance ZSM-5 and Zeolite-
Cellulose ethanol 1.0–1.1 0.9–1.05 Beta have pore sizes of 5.6 Å and 7.4 Å respectively [12,13]. Every
Rapeseed biodiesel 0.98–1.03 0.95–1.1
Biomass to liquid biodiesel 1.0–1.2 0.9–1.05
oil source consists of a number of fatty acid components thereby
Biosynthetic gas 0.9 0.85–0.95 exhibiting its own fatty acid profile; examples of which have been
Fossil gasoline 0.53–0.54 0.7 included in Table 2. Generally any fuel derived from an oil reflects
the profile of its original parent oil [14].

Fig. 3. Cylindrical molecular dimensions of a stearic acid (C18:0) molecule [11].

Table 2
Fatty acid profile of various sources of biodiesel (Normalised). Adapted and modified from Hoekman et al. [4].

Fay Acid Methyl Ester Composion

Yellow
Common Name Abbrev. Camelina Canola Coconut Corn Jatropha Palm Rapeseed Safflower Soy Sunflower Tallow
Grease
Caprioc 6:0 0.59 0.10
Caprylic 8:0 6.74 0.77
Capric 10:0 0.09 5.35 0.48 0.56 0.10
Lauric 12:0 0.36 47.26 0.09 0.29 0.09 0.09 0.10 0.19 0.19
Tridecylic 13:0
Myrisc 14:0 2.56 18.33 0.28 1.05 0.10 0.09 0.10 2.55 0.78
Myristoleic 14:1 0.29
Pentadanoic 15:0 0.58 0.09
Pentadecenoic 15:1 0.10
Palmic 16:0 5.79 4.05 9.02 11.43 14.36 41.59 4.04 8.14 11.07 6.31 23.85 16.03
Palmitoleic 16:1 0.28 0.10 0.20 0.93 0.19 0.09 0.10 0.18 0.10 2.55 0.85
Hexadecadienoic 16:2
Hexadecatrienoic 16:3
Heptadecanoic 17:0 0.09 0.09 0.10 0.09 0.09 0.10 1.35 0.09
Heptadecenoic 17:1 0.09 0.10 0.10 0.58 0.09
Stearic 18:0 2.66 1.93 2.65 1.88 5.88 4.11 1.54 2.48 3.72 3.55 17.86 6.90
Oleic 18:1 15.96 58.30 6.74 26.45 38.94 40.41 57.26 14.15 22.61 21.39 41.42 43.34
Linoleic 18:2 16.15 20.46 2.06 58.36 34.89 9.30 20.69 74.04 51.33 65.35 4.32 24.39
Linolenic 18:3 33.81 9.27 0.10 0.59 0.28 0.29 8.08 0.10 5.63 1.46 0.88 1.07
Stearidonic 18:4 0.39 0.47
Arachidic 20:0 1.33 0.65 0.10 0.30 0.19 0.29 0.37 0.10 0.27 0.29 0.20 0.28
Gondoic 20:1 13.68 1.45 0.10 0.09 0.10 2.02 0.27 0.19 0.59 0.47
Eicosadiensic 20:2 1.42 0.09 0.09
Eicosatrienoic 20:3 0.72
Eicosatetraenoic 20:4
Eicosapentaenoic 20:5
Behenic 22:0 0.81 0.28 0.10 0.19 0.10 0.28 0.27 0.58 0.10 0.38
Erucic 22:1 2.94 0.47 0.10 0.09 0.46 0.09 0.10 0.10 0.09
Docosatetraenoic 22:4
Docosapentaenoic 22:5
Docosahexaenoic 22:6
Ugnoceric 24:0 0.66 0.19 0.10 2.51 0.10 0.09 0.09 0.19 0.19
Nervonic 24:1 0.18 0.19 0.98 0.09 0.09 0.27 4.28
Other/Unknown ? 0.95 2.12 0.30 1.11 0.86 4.14 0.79 3.91 0.10 1.92
Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00
Total Saturates SFA 14.18 7.29 90.03 13.81 23.58 48.85 7.06 10.82 15.70 11.22 46.87 24.94
Total Unsaturates UFA 84.87 90.59 9.97 85.90 75.31 50.28 88.80 88.39 80.39 88.69 51.21 75.06
Dominant species in FAME Composion
Other major species (>10%) in FAME Composion

Green- edible crops forming primary biofuels. Red- non-edible crops or waste oils forming secondary biofuels.
192 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208
1.1. Biodiesel versus renewable diesel
At present there are primarily two main routes adopted in pro-
ducing renewable fuel from lipids. These are either trans-
esterification to produce biodiesel (fatty acid methyl esters) or
hydroprocessing in order to produce renewable diesel (essentially
hydrocarbons); as summarised in Fig. 4 [15–25].
Alongside trans-esterification, hydroprocessing is another
widely studied area for the treatment of lipids. This entails pro-
cess steps such as hydrogenation, hydrocracking, deoxygenation,
decarbonylation and decarboxylation of triglycerides. These steps
essentially produce what is commonly referred to as renewable
diesel though inherently often these result in the formation of
undesired products like water, CO, CO2 and light ends namely
propane. It must be noted however, since Knothe et al. coined
the term renewable diesel, in recent times various manufacturers
have used different terms to describe what is essentially the same
thing. These discrepancies may stem from branding and trade
names of the product or simply the continual use of terms first
adopted at the conceptualisation stage. Several global players
are in on the act of producing renewable diesel with Neste Oil
Corporation (Finland) being one of the market leaders by produc-
ing ‘NExBTL’ (an acronym for ‘‘next generation bio-to-liquid’’) in
2005, which was latter classified as ‘hydro-treated vegetable oil
(HVO)’ in 2007 before finally being called renewable diesel
2008 onwards. Conocophillips (United States, Ireland) and Syn-
troleum (United States) commonly refer to it as HVO whilst
Universal Oil Products (UOP)-Eni (UK, Italy) produce renewable
diesel using a process called Ecofining with the product trade
name being ‘Green Diesel’. Nippon Oil (Japan) and SK energy
(Korea) on the other hand appear to have an affinity for the term
‘hydro-treated biodiesel’ (HBD). This panmictic use of terminol-
ogy is slowly being eradicated with ‘renewable diesel’ consensu-
ally being used [26].
The focus of this article nevertheless is on biodiesel which is
regarded as a much more pragmatic and sustainable solution as
a renewable fuel source due to its simpler production process of
transesterification which requires minimal process steps and sig-
nificantly milder operating conditions.
Fig. 4. Main process routes for lipid derived biofuels. Adapted and modified from Knothe [21].
1.2. Biodiesel
Historically, the United States in 1978 and South Africa in
1981 are reported to have started exploration into vegetable oils
as fuel for possible commercial applications, with methyl esters
derived from rapeseed being produced in Germany and Austria
in 1982 along with a small pilot biodiesel production plant being
established in Austria in 1985. A brief history of biodiesel produc-
tion and research has been presented in Fig. 5. Commercial pro-
duction of biodiesel yet only came into existence as late as
1990 in Europe whereby targets were set of 2.5% of total trans-
port fuel, in other words 5 million tonnes of biodiesel for 2005,
5.75% (11 million tonnes) for 2010 and 10% (20 million tonnes)
for 2020. The largest global producer of biodiesel is the United
States accounting for 19% of the overall production whilst Ger-
many (16%) is the leading producer in Europe as seen in Fig. 6.
The EU mandate of 10% biofuel by 2020 is rather interesting as
it aims to limit production from food crops to just 5% and the
remaining 95% is expected to come from waste, residues and
completely new types of raw material. Although these may seem
slightly ambitious targets good progress is being made with just
over 2.7 million tonnes being produced in 2003 and 8.6 tonnes
reported to have been manufactured in 2011. As for the US, their
legislation requires 20% content by 2020 and they plan to reach a
production capacity of 3.3 million tonnes by 2016. On a global
perspective the biodiesel demand reached 21.8 million tonnes in
2012 and this is projected to rise to a staggering 35 million ton-
nes per annum by 2020 [5,27]. As a consequence of the imposed
targets the biodiesel sector has witnessed a rapid increase in pro-
duction over the last decade (Fig. 7), which is expected to
increase even further over the coming years. Although the global
statistics shown only date to 2012, using the latest accessible
data from the EIA U.S. reports coupled with the average U.S. glo-
bal contribution, the annual global production values of 593 and
558 thousand barrels a day can be estimated for 2013 and 2014
respectively [28,29]. Despite a clear increase in 2013 for U.S. pro-
duction a drop in 2014 is apparent – which can with a great deal
of certainty be attributed to the widely witnessed fall in crude oil
prices. Such substantial drops in fossil fuel prices naturally result
 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208 193

Fig. 5. History of Biodiesel. Adapted and modified from Lin et al. [31]; Image [32].

Fig. 6. Top biodiesel producing countries in 2011 [33].

194 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208
Fig. 7. Biodiesel Production from 2000 to 2012 in thousand barrels per day. Adapted and modified from Eia [34].
in less investment into the renewables sector due to the less
perceived pressure. That said in the first 9 months, for the U.S.,
figures show that 948 million gallons of biodiesel were produced
in 2015 which is more than that in the first 9 months in 2014
(916 million gallons) and just shy of values reported in 2013
(958 million gallons) [28,29]. This clearly indicates positive signs
of recovery for the biodiesel production sector and now owing
to the UN Climate Change Summit Paris 2015 agreement rapid
expansion in the market, in particular in the developing countries,
is anticipated [30].
1.3. Biodiesel challenges
Despite the overwhelming interest and growth in biodiesel pro-
duction the widespread uptake is still hindered by a number of
political, environmental, economic, social and technical challenges.
For the purposes of this review only the technical aspects will be
discussed. Fundamentally from a technical perspective problems
associated with economic feasibility, availability, quality and sus-
tainability need to be addressed in order facilitate greater utilisa-
tion of biodiesel, as elaborated in Fig. 8.
1.3.1. Availability
Currently the global biodiesel production is dominated by the
use of edible oils, with approximately 84% being produced from
just rapeseed oil. Other major contributions come from sunflower
oil (13%), palm oil (1%), soybean oil and others (2%), therefore over
95% of biodiesel is derived from edible oils. This presents a sub-
Fig. 8. Key areas of technical biodiesel challenges.
stantial problem with food resources becoming scarce in certain
parts of the world and large-scale global production of automotive
fuel based on edible oils almost certainly likely to contribute to an
imbalance in food supply and demand hence sparking the fuel ver-
sus food debate [35]. Additionally, assuming business as usual, not
only will increased demand for vegetable oil lead to unnecessary
clearing of forests but the deforestation process for plantation will
inevitably also disturb animal and plant ecosystems [36]. Based on
the information above, it is therefore imperative to avoid utilisa-
tion of first generation feedstock namely refined vegetable oils
derived from food crops, and explore second and third generation
biodiesel sources namely waste, non-edible crops and marine
algae. Another key aspect at a global level in displacing the petro-
leum economy is whether enough oil can in actual fact be obtained
to transform into biodiesel; principally can the oil supply meet the
fuel demand. From Fig. 9 it is apparent that micro algae presents
the greatest opportunity to address this concern.
1.3.2. Sustainability
There are obvious significant reductions in emissions by utilis-
ing biodiesel as opposed to conventional petroleum and diesel fuel
as displayed in Fig. 10. However despite this based on the existing
infrastructure from a life cycle analysis point of view, the sustain-
ability of most alternative fuels particularly biodiesel can be
deemed questionable. This is due to many of the raw materials
and equipment being synthesised from crude oil diesel. This is
epitomised by the methanol used to produce biodiesel as it is lar-
gely a product of the oil and gas industry as opposed to being pro-
 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208 195

Fig. 9. Annual oil yield of various feedstock. Adapted and modified from Atabani et al. [37].

Fig. 10. CO2 emissions produced (g per MJ of energy) of biodiesel and current crude oil derived fuels. Adapted and modified from DfT [38].

duced from bio-origins. Being able to produce substantial amounts
of methanol economically using syngas derived from gasification
of residues and crops is a key challenge in answering the sustain-
ability issue.
1.3.3. Economic feasibility
Current biodiesel production would most probably cease to
exist were it not for various tax exemptions and government sub-
sidies; as it gives them a chance of competing with the relatively
much less costly fossil fuel derived diesel. The overall production
cost of biodiesel consists of raw material (production and
processing), catalyst, biodiesel processing (energy, consumables
and labour), transportation (raw materials and final products) as
well as local and national taxes. With refined vegetable oil being
the main feedstock for most biodiesel facilities at present, the cost
of feedstock is subsequently reported to have an astonishingly con-
tribution of almost 80% on the overall biodiesel production cost, as
shown in Fig. 11. Therefore if biodiesel is to ever penetrate and
dominate the global market the choice of feedstock will be the
determining factor as it is the most crucial variable affecting the
final biodiesel price [39].
Another game changer in terms of improving economic feasibil-
ity is to expand the value chain by utilising by-products such as
glycerol to establish a series of commodity bio-chemical and bio-
polymer industries [33,40,41]. Current research into catalytic
transformation of glycerol in to value adding products is offering
196 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208

poor oxidative stability and cold flow properties exhibited by bio-

diesels. Addressing these issues is the heart of this specific research
area and review article.
It must nevertheless be noted that the aforementioned factors
are interdependent and only through solving the quality, availabil-
ity, sustainability and economic feasibility issues collectively will
there be a momentous shift towards a bio-economy.

1.4. Biodiesel quality

The quality of all commercially produced pure biodiesel (B100)

Fig. 11. General cost breakdown for biodiesel production. Adapted and modified
from Atabani et al. [37].
enticing opportunities in particular through the formation of
polyglycerols and their derivatives which have numerous applica-
tions in the food, pharmaceutical and cosmetic industries [33]. A
host of reactions such as oxidation, dehydration, acetylation, ether-
ification, esterification, acetalisation and ammoxidation are
reported to yield typical commodity chemicals like citric acid, lac-
tic acid, 1,3-dihydroxyacetone, 1,3-propanediol, dichloro-2-
propanol, acrolein, hydrogen and ethanol [41].
1.3.4. Quality
Lastly the key challenge at present in producing direct ‘drop-in’
fuels is resolving the issue of quality which is mainly linked to the
is dictated by quality standards which ensure normal engine oper-
ation and aim to minimise damage and pollution. These are defined
by ASTM D 6751 in the United States of America and EN 14214 in
Europe. Table 3 shows the full list of properties and specifications
for biodiesel for both of these standards [42].
With regards to the properties of biodiesel the initial complex-
ity arises due to the varying nature in composition of the parent oil
feedstock, as shown in Fig. 12.
As aforementioned the focus of this article is specifically on the
oxidative stability and cold flow properties (CP, CFPP ad PP); with
Fig. 12 presenting a paradox. Although increased degree of unsat-
uration results in improved cold flow properties (illustrated by a
reduction in CP, CFPP and PP) a decline in oxidative stability is also
experienced. These terms will be discussed and concepts clarified
in the following passages.
1.4.1. Oxidative stability
The oxidative stability (OS) of a fuel is a measure of the duration
of time or shelf-life before the oils are oxidised by atmospheric
oxygen to form undesirable products which can damage the com-
bustion engine. Oxidative stability is crucial when it comes to the

Table 3
Properties and specification of the ASTM D6751 and EN 14214 standards. Adapted from Leung et al. [3].

Property Unit Limits Test method

ASTM D6751 EN 14214 ASTM D6751 EN 14214
FAME content % (m/m) 96.5 min EN 14103
Flash point °C 130.0 min 101.0 min D93 ISO CD3679e
Kinematic viscosity at 40 °C mm2/s 1.9–6.0 3.5–5.0 D445 EN ISO 3104
Cetane number – 47 min 51 min D613 EN ISO 5165
Sulphated ash content % (m/m) 0.020 max D874 ISO 3987
Copper strip corrosion – No. 3 max Class 1 D130 EN ISO 2160
Acid value mg KOH/g 0.80 max 0.5 max D664 pr EN 14104
Free glycerol % (m/m) 0.020 max D6584 pr EN 14105m pr EN 14106
Total glycerol % (m/m) 0.240 max 0.25 max D6584 pr EN 14105m
Phosphorous content % (m/m) 0.001 max 0.01 max D4951 pr EN 141101
Carbon residue
ASTM D6751 (100% sample) % (m/m) 0.050 max – D4530 –
EN 14214 (10% bottoms) – 0.3 max – EN ISO 10370
Cloud point °C Report customer – D2500 –
Density at 15 °C kg/m3 – 860–900 – EN SIO 3675 EN SIO 12185
Distillation T90 AET °C 360 max – D1160 –
Sulfur (S 15 Grade) ppm 0.0015 max – D5453 –
Sulfur (S 500 Grade) ppm 0.05 max – D5453 –
Sulfur content mg/kg – 10 max – –
Water and sediment %vol. 0.050 max – D2709 –
Water content mg/kg – 500 max – EN ISO 12937
Total contamination mg/kg – 24 max – EN 12662
Oxidation stability at 110 °C h – 6 min – pr EN 14112
Iodine value – – 120 max – pr EN 14111
Linolenic acid methyl ester % (m/m) – 12 max – pr EN 14103d
Polyunsaturated (P4 double bonds) methyl esters % (m/m) – 1 max – pr EN 14103
Ester content % (m/m) – 96.5 min – pr EN 14103d
Methanol content % (m/m) – 0.2 max – pr EN 141101
Monoglyceride content % (m/m) – 0.8 max – pr EN 14105 m
Diglyceride content % (m/m) – 0.2 max – pr EN 14105 m
Triglyceride content % (m/m) – 0.2 max – pr EN 14105 m
Alkaline metals (Na + K) mg/kg – 5 max – pr EN 14108 pr EN 14109
 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208
Fig. 12. Degree of unsaturation of various biodiesels and their effect on properties. Adapted and modified from Hoekman et al. [4].
utilisation of biodiesel, since unstable fuel tends to increase in vis-
cosity and form gums, sediments and other deposits. The composi-
tions of the biodiesel, age of the mixture and storage conditions
also have a significant effect on the oxidative stability of the fuel
[43,44]. Biodiesels consisting of a higher degree of unsaturated
components lead to poorer stability, where the rate of oxidation
of unsaturated compounds depends upon the number and position
of the double bond(s) [10]. Typically the oxidation process begins
by extracting a hydrogen atom from the carbon adjacent to a dou-
ble bond therefore biodiesel compounds with greater number of
double bonds possess greater oxidative instability. The extraction
of hydrogen is followed by a rapid reaction with oxygen to form
an allylic hydroperoxide, which then further reacts to form prod-
ucts such as aldehydes, alcohols, and carboxylic acids. The form
in which the unsaturated geometric isomer exists also affect the
oxidative stability, with the trans configuration being more stable
than the cis configuration. In terms of oxidative stability naturally
obtained fats and oils tend to be largely composed of cis isomers
hence are relatively oxidatively instable [4,14]. Almost all biodiesel
composed of unsaturated esters require an antioxidant to meet the
American or European standards because even the most oxidative
stable mono-unsaturated compound, methyl oleate, in its pure
form is only stable for 2.5 h as opposed to the minimum 6 h
requirement stated in the EN 14214 standard or 3 h induction time
defined in ASTM 6751 [14]. Hence this necessitates the need for
some form of saturated components in the biodiesel mixture. The
oxidative stability is typically measured using either the Rancimat
method or Oxidative Stability Index (OSI) [45,46].
1.4.2. Cold flow properties
The cold flow properties (CFP) on the other hand are those
exhibited by a fuel at temperatures below ambient and are charac-
terised by three main indices; cloud point (CP), cold filter plugging
point (CFPP) and pour point (PP) [47].
The first index is the CP which is defined as ‘‘the temperature at
which the smallest observable cluster of wax crystals first appear”
197
with the diameter exceeding 0.5 lm and result in a hazy or cloudy
suspension [48,49].
The CFPP is the most widely accepted cold flow index and is
used in standards as it most accurately and efficiently predicts
how diesel fuels will perform in tanks and fuel systems in cold cli-
mates. It is defined as ‘‘the lowest temperature where 20 mL of fuel
passes safely through a 45 lm wire mesh filter under 200 mm H2O
(0.019 atm) vacuum within 60 s” [28,29]. CFPP is essentially the
point at which the fuel jams the filters due to the formation of
agglomerates of crystals, hindering the free flow through fuel
systems.
Lastly the PP is the temperature at which the fuel solidifies after
prolonged exposure at low temperatures causing crystals to grow
and gel together, eventually resulting in fuel ceasing to flow [48].
This also typically corresponds to the melting point of a mixture
[46].
Although naturally occurring oils contain unsaturated compo-
nents and in the cis configuration both of which favour good cold
flow properties, the presence of any saturated components results
in higher melting points and poorer cold flow properties, see
Table 4.
These two key quality parameters present a ‘Catch-22’ [46]; for
good fuel stability you need saturated and trans components
whilst at the same time for good cold flow properties you need
unsaturated and cis components.
The aim of this article is therefore to propose theoretical and
practical solutions in solving this apparent paradox associated with
the quality of biodiesel; with the focus being on the cold flow
properties.
1.4.3. Existing solutions
Currently this paradox is addressed using various methods
some of which have been outlined in Table 5. The oxidative stabil-
ity and cold flow properties are most commonly resolved by either
blending with crude diesel which is unsustainable as it still
requires fossil based fuel or using additives like iso-butene as
198 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208
Table 4
Molecular requirement for good OSI and CFP [1,2].
Good OS (Poor CFP)
Table 5
Existing solutions to improving CFP. Adapted and modified from Reaume et al. [53].
Approach Description
Blending with crude diesel Mixing biodiesel with crude oil derived diesel in various ratios
Modify oil content of seed Genetic engineering of seed to produce higher yields of unsaturated components
Additives Addition of pour point and cloud point depressants
Other alcohols Use longer or branched alcohols instead of methanol
Processing Winterisation or dry fractionation to remove fractions with high melting points
freezing point depressants and tert-butylhydroquinone as an
antioxidant which are both expensive [14,44,50,51]. Furthermore
in the case of cold flow properties processing to remove fractions
with high melting points is also utilised however it leads to a sub-
stantial losses in yield [52–56].
2. Challenges and opportunities using heterogeneous catalysis
The challenge therefore lies in finding a way to improve the cold
flow properties (CFP) without blending, using additives or process-
ing. An exciting opportunity which offers a practical solution is
through the modification of the chemistry of the biodiesel (fatty
acid methyl esters) molecules using heterogeneous catalysis. On
a molecular level there are numerous factors which contribute to
the cold flow properties of biodiesel. These include the carbon
chain length, odd or even number of carbons, the degree of satura-
tion, cis or trans geometric isomers, the position of double bonds
and the degree of branching. Fig. 13 shows the relationship
between the melting points of various FA and FAME and the afore-
mentioned factors.
When it comes to enhancing the cold flow properties of biodie-
sel via heterogeneous catalysis there are mainly three processes
dehydrogenation, hydrocracking and hydroisomerisation. Fig. 14
shows the effects of these processes on the molecular constituents
of the biodiesel mixture.
Whilst the application of zeolites to improve the cold flow prop-
erties using processes such as dehydrogenation and dewaxing in
the petro-chemical, oil and gas industry is an extremely well-
studied area; its application in the field of bio-chemistry has been
relatively less explored. The following section of the report outlines
details of these processes and includes a brief literature review of
associated studies to date.
Good CFP (Poor OS)
Limitations
Not a sustainable solution still reliant on fossil fuels
Expensive
Ethical issues
Does not resolve oxidative stability issue
Expensive hence increase cost of fuel
Not as widely available hence much more expensive
Loss in yield
Reduced oxidative stability
Reduced ignition quality
2.1. Dehydrogenation
Dehydrogenation involves removing hydrogen to induce a dou-
ble bond in the carbon chain length thereby resulting in a substan-
tial reduction in melting point. This is the most direct and obvious
method in improving the cold flow properties however the addition
of each double bond also leads to a significant reduction in oxida-
tive stability and cetane number. The example below, in Fig. 15,
of C 18 illustrates this whereby although going from C18:0 to
C18:3 leads to a reduction of 91 °C in melting point it leads to a dra-
matic reduction of oxidative stability from >40 h to 0.2 h.
The studies in attempting to dehydrogenate FAME using hetero-
geneous catalysis are limited with Macias et al. being one of the
few groups to have tried it in 2011. Upon testing numerous cata-
lysts, including aluminosilicates and silicoaluminophosphates,
only Ga–Nb oxides displayed signs of dehydrogenation. However
undesired reactions of hydrolysis from FAME to FA and dimeriza-
tion were also observed. Dehydrogenation also typically requires
higher reaction conditions of 220 °C in comparison to isomerisa-
tion which was reported to have been successfully carried out at
125 °C [61].
Nevertheless as stated previously and reiterated above,
although dehydrogenation helps improve the cold flow properties
at the expense of the oxidative stability it cannot be regarded as a
holistic solution in the context of generating a ‘drop-in’ renewable
alternative fuel. Therefore dewaxing is considered as the more
complete solution in addressing the issue of poor quality associ-
ated to biodiesel.
2.2. Dewaxing
Dewaxing of long chain molecules by hydrocracking and hydroi-
somerisation using zeolites to improve the cold flow properties has
 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208 199

Fig. 13. Melting points of various FA and FAME species [2,11,57–59].

Fig. 14. Molecular alteration options with respective processes to improve cold flow properties [60].

Fig. 15. Dehydrogenation products with melting point and oxidative stability properties [1,2].
200 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208
been a thoroughly studied area in the petro-chemical industry. For
the purposes of this review the results of one classical study by
Weitkemp et al. have been shown below in Figs. 16 and 17. The
reactions in this study were carried out in a small scale fixed bed
reactor with a bed volume of 2 cm3 using catalyst 0.5 wt% Pt/Ca
loaded Y zeolite at 39 bar pressure, a hydrogen: n-alkane molar
ratio of 17:1 and feed flowrate of 12  103 mole h1 [62].
The findings of this study albeit carried out on hydrocarbons
can be considered useful to bio-chemical applications as well. For
instance it can be deduced that cracking requires a higher temper-
ature than isomerisation, that longer chain molecules react at
lower temperatures as compared to shorter molecules and that
the overall maximum conversion of hydroisomerisation can be
expected to be less than hydrocracking for a given molecule. A
study such as this is pivotal to obtain bearings when investigating
the bio-chemical application and also in gaining a better analytical
insight to studies already conducted on fatty acids and fatty acid
methyl esters. The sub sections below will discuss the hydrocrack-
Fig. 16. Effect on hydroisomerisation conversion of hydrocarbons with varying
chain lengths at increasing temperatures. Adapted from Weitkamp [62].
Fig. 17. Effect on hydrocracking conversion of hydrocarbons with varying chain
lengths at increasing temperatures. Adapted from Weitkamp [62].
Fig. 18. Cracking products with melting point and oxidative stability properties [1,2,63].
ing and hydroisomerisation theory in a bio-chemical context along
with summarising key relevant literature articles to have emerged
relatively recently.
2.2.1. Hydrocracking
The ability to breakdown long chain FAME to smaller ones,
using hydrocracking shown in Fig. 18, provides an ideal solution
in terms of improving properties as this can result in substantial
reduction in melting point whilst maintaining the oxidative stabil-
ity. This can also be viewed as organic in-situ blending thereby
eliminating the need to post blend biodiesel with crude oil derived
diesel.
A possible limitation of cracking is the formation of light ends
such as propane which are useful in the petro-chemical industry
however the existing infrastructure of bio-refineries do not offer
much prospects. Therefore research into the utilisation of light
ends to add value to bio-refineries is an opportunity worth explor-
ing. Based on the current situation excessive light-end formation
can largely be deemed as undesirable as it will simply result in
losses in yields.
Nevertheless if a catalyst can be found which can selectively
crack long chain molecules in to two medium size ones rather than
one large and small (e.g. C 18 into C 10 and C 8, as shown in Fig. 18,
rather than C 2 and C 16), this would solve the issue of formation of
light ends. A potential problem it may however create, in the case
of biodiesel, is that a part of the molecule will be a fatty acid
methyl ester, a polar compound, whilst the other part a hydrocar-
bon, a non-polar, resulting is potential solubility issues. These
could nonetheless be addressed through blending techniques.
Finally due to the presence of the carboxylic acid group, any reac-
tion at high temperatures is likely to occur here first as opposed to
the hydrocarbon chain. It is therefore hypothesised that the decar-
boxylation reaction will be thermodynamically favoured as
opposed to cracking.
Cracking of FAME using heterogeneous catalysis is also an
understudied area with Macias et al. and Yori et al. having reported
observing small degrees of cracking using sulphated zirconia and
niobium oxides at temperatures greater than or equal to 200 °C
[50,61]. It is however more extensively studied in oil and gas and
renewable diesel applications particularly in China, Korea and
Japan [20,64–66].
The reaction mechanism for hydrocracking uses a bi-functional
catalyst and typically involves five distinct steps [64,67]:
1. Saturated molecule is dehydrogenated creating an alkene.
2. Alkene is protonated to create carbocation at double bond site.
3. C–C bond dissociation (b-scission) of carbenium ion to form an
alkene and a primary free radical.
4. Proton detaches from primary free radical resulting in another
alkene.
5. Pair of alkenes are hydrogenated to alkanes.
These have been visually illustrated in Fig. 19.
 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208
Fig. 19. Hydrocracking reaction mechanism.
2.2.2. Hydroisomerisation
The skeletal isomerisation of linear FAME to branched chain
FAME offers the most interesting possibility of reducing the freez-
ing temperature as it only involves altering the chemistry of the
molecule so there is no loss in yield which is possible in cracking
through the formation of light end products or massive reduction
in oxidative stability which is observed in dehydrogenation.
Hydroisomerisation simply refers to isomerisation of hydro-
genated reactants for example paraffin or saturated FA/FAME, an
example of which can been seen in Fig. 20.
In terms of isomerisation and hydroisomerisation of FAME this
is a well-studied area in recent times in comparison to dehydro-
genation and cracking. The findings of the numerous studies have
been summarised and discussed below.
In the last decade a number of studies have been conducted
investigating the isomerising of FAs and FAMEs over solid acid cat-
alysts to form skeletal branches in an attempt to improve the cold
flow properties.
Zhang et al. working for Akzo Nobel appear to have filed some of
the earliest patents and published articles on this subject dating as
far back as 2001. Their first patent was filed back in 2001 titled
‘‘Skeletal Isomerization of Fatty Acids” and published in January
of 2003 [68]. A year later another patent called ‘‘Fatty Acid Isomeri-
sation with Mesoporous Zeolites” was published along with a jour-
nal article looking at branched-chain fatty acids [69]. The published
article paper in 2004 highlighted an improved process for skeletal
isomerisation of unsaturated fatty acids using large-pore zeolites
(>6 Å) such as zeolite-beta which results in greater ethyl branching
in comparison to methyl branching. The formation of multiple
branched acids in large pore zeolites is said to occur due to the
double bond migrating away from the existing branching position
thereby enabling consecutive isomerisation and increased
Fig. 20. Hydroisomerisation product with melting point and oxidative stability properties [1,2].
201
branching. Furthermore, it was also reported that the utilisation
of mesostructured zeolites (>15 Å) or more specifically zeolites
with post synthetic modification to produce mesopores in microp-
orous structures, not only maintain a high isomerisation activity
but also display a much slower deactivation rate to the extent that
catalyst deactivation isn’t considered a significant issue [70].
Following on, Yori et al. in 2006 used mordenite and sulphated
zirconia catalyst in an attempt to isomerise linear oleic acid to
branched chain isomers in a batch reactor system. Temperature
drops of between 4 °C and 6.5 °C were observed, with sulphated
zirconia displaying the best result at 125 °C. In addition it was
reported that isomerisation had very little effect on the saturated
FAMEs of methyl palmitate and stearate, whilst unsaturated frac-
tions had the greatest reactivity, with C18 FAMEs (methyl- oleate,
linoleate and linolenate) suffering the greatest amount of skeletal
branching. However, catalyst deactivation due to the poisoning
by the sulphonic groups and severe cracking of the unsaturated
components were the main issues [71].
In 2007, Zhang et al. published yet another paper titled ‘‘skeletal
isomerisation of unsaturated fatty acids: the role of mesopores in
HBeta zeolites” which explored a number of commercial and labo-
ratory generated zeolite-beta samples in the isomerisation of oleic
acid. Amongst the commercial samples the sample with a Si/Al of
13.5 displayed the greatest activity and was found to contain the
largest amount of acid sites. As the isomerisation of unsaturated
fatty acids is frequently postulated to be a Brønsted acid catalysed
reaction, Zhang et al. state that for fatty acids the most accessible
sites are expected to be located inside the mesoporous channels
of the zeolites. Based on their findings it was concluded that fatty
acid skeletal isomerisation can be effectively achieved by zeolite-
beta that have both high acid site density and also accessible meso-
porosity. Therefore neither a microporous zeolite with a high acid
202 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208
site density nor a large pore zeolite with a low acid site density can
have a high isomerisation activity [72].
Furthermore, Ngo et al. in 2007 investigated the isomerisation
of oleic acid to branched chain isomers using modified Ferrierite
zeolite catalyst. The high surface area and high density of Brønsted
acid sites on the Ferrierite made it an active catalyst for skeletal
isomerisation. High conversions of 98% and a high selectivity of
85% were obtained as well as a reusability cycle of three cycles.
The oleic acid was converted to saturated branched chain isomers
(bc-SFA) namely isostearic acid isomers using modified Ferrierite
zeolites with a Si/Al ratio of 17.5–20 at a reaction temperature of
250 °C and time of 6 h in a batch reactor system. The isomerised
products then underwent Pd-catalysed hydrogenation and esterifi-
cation to form methyl esters resulting in bc-sFAME isomers namely
iso-stearate isomers. The best result was obtained using 2.5 wt%
Ferrierite catalyst with a small amount (1 mL) of deionised water
as a co-catalyst, which is thought to increase the number of active
Brønsted acid sites since it displayed the lowest Si/Al ratio [73].
Since then Ngo et al. have synthesised and characterised iso-oleic
and iso-stearic acids using Ferrierite and have found that branch-
ing results in improved cold flow properties, oxidative stability
and lubricity [74–76]. They have recently further gone on to inves-
tigate the process development and associated costs of saturated
branched chain isomer (iso-stearic acid) production. The overall
process is composed of 3 major steps; catalyst treatment, isomeri-
sation and product recovery as shown in Fig. 21 [77,78]. Although
studied in a batch reactor on a laboratory scale they have modelled
the process using CSTRs in the simulator.
The minimum production cost was determined to be
$2.375 kg1 with 51% of reduction in cost being obtained due to
the ability of regenerating Ferrierite at least 15 times. However
in order to be competitive in the market further cost reductions
were deemed to be necessary and they believe that through iden-
tification of more efficient and selective zeolite catalysts which
enable lower operating temperatures (<250 °C), the elimination
of a recrystallization and distillation steps; commercial viability
is certainly possible [77,78].
In essence, branched SFA have better CFP than linear SFA and
have better OSI than linear UFA. As Ngo et al. have stated that a sin-
gle monobranched C18 isomer is still a solid at room temperature
however a mixture of branched isomers i.e. mixture of methyl
branched positional isomers or multiple or longer branches pos-
sess a melting point that is significantly reduced. This enables
the mixture to behave as a fluid at ambient temperature but retain
Fig. 21. Process flow diagram for the production of branched-chain isomers from unsaturated fatty acids. Adapted from Ngo et al. [77].
its characteristics of being oxidatively stable. This phenomenon is
associated to the branched species being unable to orientate them-
selves routinely in an orderly fashion to form a crystal lattice [74].
These unique and special characteristics exhibited by a mixture of
branched chain fatty acids make them impeccable value adding
chemicals especially for the biodiesel and biosurfactant industry.
At present iso-stearic acid is in fact commercially being produced
by CRODA Co. (United Kingdom) under the trade name PrisorineTM
(P3505 ISAC) and by Arizona Chemicals (United States) as Cen-
turyTM 1105 and 1107 [78].
An Inorganic Chemistry and Catalysis research group (Wiede-
mann et al.) based in Utrecht University (Netherlands) in collabo-
ration with Croda appear to have adopted Ngo’s approach and
expanded on the findings of skeletal isomerisation of oleic acid
using ferrierite by studying the deactivation and influence of active
sites. The studies once again were performed in a batch reactor
system using 1.5 wt% catalyst at 260 °C which was initially pres-
surised to 1 bar and upon gradual heating at the rate of 6 °C/min
for roughly 40 min the pressure is reported to have increased to
11 bars [79,80].
The deactivation study published in 2014 suggests that deacti-
vation occurs predominantly in two ways. Firstly through coking
which is experienced very early on in the reaction, as little as min-
utes, leading to blocked pores and only the pore mouth being avail-
able for activity. The major culprits for coking up the catalyst were
identified to be dodecylbenzene and other long chain alkylben-
zenes reported formed from oleic acid via a cyclic conjugated
ketone. Secondly, the acid sites in the pore mouth are believed to
be poisoned by polyenylic carbocations which are formed during
the hydrogen transfer reactions between reactant species. The ini-
tial level of coking and poisoning of the active sites is sufficient to
considerably slow down the skeletal isomerisation. The effects of
coking are said to increase with higher concentrations of polyun-
saturated acids in the feed [79].
The most recent publication by Wiedemann et al. focusing on
the influence of active sites of the H-Fer catalyst showed that initial
activity associated with pore is due to the Brønsted acidity in the
10-MR channels. A greater ratio of Brønsted to Lewis sites are
reported to explain the high levels of BcUFAs observed whilst the
high selectivity to form mono-branched chain UFA have been
attributed to the presence of a few but strong Brønsted active sites.
Moreover, having a low density of external acid sites is said to
favour the use of H-Fer as it limits the undesirable oligomerisation
reaction and in turn enhancing the isomerisation reaction. Using
 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208
H-Fer is reported to generate a profile of at least 30 alkyl isomers
consisting predominantly of mono-branched UFA with the methyl
group having a tendency of being located on the 8–14 positions
along the carbon chain length. Only minute amounts of multiple
branched were detected which mainly exist on positions 11, 14
and 14, 15. It is also claimed that the optimal cold flow properties
are achieved by synthesising a single methyl-branched molecule
located towards the centre of the carbon chain as opposed to the
formation of multi-branched isomers [80]. This notion is based
on a series of studies by Cason et al., particularly one on melting
points of mono-branched acids which has been reproduced in
Fig. 22 [81–83].
From a biodiesel application perspective although adopting
such an approach of isomerisation followed by hydrogenation for
a biodiesel derived from a predominantly unsaturated source
would be extremely effective, economical and an enticing solution;
such a process would not be applicable to saturated biodiesels such
as coconut nor heterogeneous ones such as palm and beef tallow.
To address the latter two, hydrogenation followed by hydroiso-
merisation is the desired route to creating bc-sFAMEs. Further-
more, in the greater scheme of things with a trend towards multi
feedstock biorefineries having a standard process of hydrogenation
followed by hydroisomerisation offers a better prospect, despite
isomerisation followed by hydrogenation requiring significantly
milder operating conditions, due to its wider applicability [84].
In summary to the studies discussed thus far it appears that the
solution lies in hydrogenating the oils in order to attain the oxida-
tive stability and then seeking to form multi-branched species (see
Fig. 23), ideally located towards the centre of the molecule, using
the acid heterogeneous catalyst to improve the CFP. Additionally
having mesostructured zeolites would facilitate the reaction and
increase conversion.
Macias et al. in 2011 studied the catalytic transformations of a
commercial sample of biodiesel and methyl oleate model com-
pound over solid acid catalysts (H-ZSM-5 [Si/Al ratio of 15 and
40], H-Beta, H-Y, SAPO-11 and H-MCM-22) and bifunctional cata-
lysts (gallium–niobium oxides) which possess both acidic and
Fig. 22. MP of positional isomers of methyl-octadecanoic acids [82].
Fig. 23. Types of multiple branching [1,2].
203
redox properties. It was observed that the products of the catalytic
reaction of biodiesel (methyl- oleate, linoleate and linolenate) con-
sist of both double bond and skeletal isomers, dehydrogenation
products of both the saturated and C 18:1 FAME in the feed and
dimers of the methyl-oleate. Significant amounts of cracked prod-
ucts (C 14 and C 16) were also observed especially over the alumi-
nosilicate zeolite catalysts. Furthermore, whilst the large-pore
zeolites (H-Beta, H-Y and H-MCM-22) were found to be more reac-
tive, a greater selectivity for skeletal isomerisation was observed
over the medium-pore, weaker acidic, SAPO-11 for both the reac-
tions of the biodiesel mixture and model compound methyl oleate.
As for the mesoporous Ga–Nb oxides, acid catalysed reactions, like
double bond migration, skeletal isomerisation and dimerisation as
well as redox reactions like the dehydrogenation of the methyl-
oleates to linoleates, methyl-palmitates to palmitoates were
observed. Low conversions of biodiesel and methyl oleate were
seen with the less acidic catalysts and at low reaction tempera-
tures whilst cracking was observed when the reaction temperature
exceeded 220 °C. Among the catalysts tested GaNb4 and SAPO-11
displayed the greatest selectivities for skeletal isomerisation. It
was also noted that although skeletal isomerisation and dehydro-
genation lowered the melting point of the biodiesel, the formation
of dimers and free fatty acids on the other hand degraded the cold
flow properties [85].
Very recently the subject of enhancing CFP of biodiesel via
heterogeneous catalysis has been explored in depth by Reaume
et al. who have four publications under their name since 2011.
In their first publication Reaume et al. (2011) studied the iso-
merisation of oleic acid and methyl oleate to their branched chain
products using a zeolite-B catalyst with a Si/Al ratio of 180. The
process for forming branched chain fatty acids methyl esters pre-
esterification consisted of three steps involving hydrolysis, skeletal
isomerisation and esterification whilst for post-esterification only
two steps were required namely trans-esterification and skeletal
isomerisation. A variety of operating conditions were tested with
temperatures of 200–300 °C, pressures of 0.1–3.0 MPa N2 and co-
catalysts (water and methanol) between 0 and 2 wt%. Although
204 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208
the cloud point was unsuccessfully reduced an interesting observa-
tion was made in the formation of unbranched trans isomers which
led to an increase in the overall cloud point since the trans isomers
possess a melting point 29 °C greater than their cis counterparts. It
was also observed that performing isomerisation on esters was
preferred over fatty acids based on less trans isomer formation
and thus better cloud point results. Reaume et al. acknowledged
the fact that although there were benefits in creating branched
chain isomers, identifying ways to limit the production of
unbranched trans isomers was imperative [86].
In the second article Reaume et al. (2012) investigated the
skeletal isomerisation of oleic and palmitic acid mixtures. More
specifically, the effects of isomerisation and hydroisomerisation
reactions on the most common saturated fatty acid (palmitic)
and unsaturated fatty acid (oleic) was studied. Reaume et al. claim
to have tested the fatty acids instead of the fatty acid methyl esters
as they have greater reaction rates in the isomerisation and hydroi-
somerisation reactions; this is overrules their preference to react
esters due to their greater thermal stability as eluded to in their
previous study. The reaction conditions for isomerisation were
250 °C and 1 MPa H2 using zeolite-B (12.5) with a reaction time
of 6 h whilst for the hydroisomerisation reaction conditions of
300 °C and 4 MPa H2 using 0.5 wt% Pt – zeolite-B (12.5) with a
reaction time of 16 h. Geometric isomerisation of unsaturated fatty
acid from cis to trans isomer resulted in an increase in cloud point
whilst branching caused a reduction. The other negative side reac-
tion was direct hydrogenation of oleic acid to stearic acid without
branching. Overall the thermodynamic stability for the trans iso-
mer formation and hydrogenation reaction was noted to be much
greater than the tendency to form branched isomers through the
isomerisation and hydroisomerisation of pure oleic acid only and
hence displaying a negative effect on cloud point. As for the palmi-
tic acid, the two undesired reactions did not exist due to the
absence of a double bond and whilst isomerisation had a negligible
effect, hydroisomerisation had a positive impact resulting in a drop
in cloud point. An increase in the percentage of branching resulted
in a linear trend in reduction of cloud point whilst surprisingly
when increasing the percentage of unsaturated compounds in a
mixture no significant reduction in cloud point was observed up
until a mixture concentration of 60% unsaturated compounds. A
similar trend was also reported in hydrocarbons where the addi-
tion of double-bond alkenes did not significantly affect the melting
point until concentrations of almost 75% was reached. In this arti-
cle the cloud point was successfully reduced with the best result
being a cloud point reduction of 7.5 °C with an esters ratio of 55
(oleic acid): 45 (palmitic acid), however the issue of trans isomer
formation still exists [87].
The latest two articles are both published in 2013. The first of
the two looked at hydroisomerisation of saturated fatty acids and
methyl esters. More specifically the hydroisomerisation of palmitic
acid, methyl palmitate and hexanoic acid were studied using
0.5 wt% Pt – zeolite-B (12.5) catalyst at reaction conditions 200–
285 °C and 1–4 MPa H2 pressure. Although there is extensive liter-
ature published on hydroisomerisation of long chain hydrocarbons
this article is amongst a very few that have studied hydrosomeri-
sation of fatty acid and fatty acid methyl esters. The carbonyl
groups of COOH and COOCH3 for the acids and esters respectively
are said to hinder the catalytic hydroisomerisation reaction result-
ing in low conversions to the iso-species. The findings indicated
that performing hydroisomerisation on fatty acids was preferable
to that of esters as significantly greater conversions (approxi-
mately 30%) into branched products were experienced at high tem-
peratures of around 280 °C. A substantial increase in conversion
was noted from the reaction at 250 °C compared to the one at
285 °C which suggested that the activation energy for the hydroi-
somerisation reaction was likely to be greater than 250 °C. Another
noteworthy observation included that an increase in pressure from
1.0 to 2.5 MPa led to a 15% increase in conversion whilst when it
was further increased to 4.0 MPa the conversion only increased
by another 6%. The best result achieved was a cloud point reduc-
tion of 10 °C on palmitic acid at reaction conditions of 285 °C and
4 MPa H2 pressure giving a yield of 42% after 16 h. Although a rel-
ative good yield of branched products was obtained and the cloud
point reduced, a reaction time of 16 h is extremely long coupled
with the fact that a high pressure was also required meaning that
if developed into an industrial process it would be energy intensive
and costly [88].
In their most recent publication Reaume et al. explored the use
of isomerisation and hydroisomerisation reactions on biodiesel
derived from various oils. Isomerisation was carried out at 260 °C
and 1.5 MPa H2 pressure using zeolite-B (12.5) for 6 h and hydroi-
somerisation at 300 °C and 4 MPa H2 using 0.5 wt% Pt – zeolite-B
(12.5) for 16 h. Each oil underwent a four step process beginning
with hydrolysis; to release the fatty acids as Reaume et al. claim
that the triglycerides are too large in their original state to undergo
isomerisation effectively, this was followed by isomerisation,
hydroisomerisation and esterification. The best result gave a cloud
point reduction of 16.5 °C with coconut oil, a predominately satu-
rated oil, as the reactant. Palm oil was the only other oil to experi-
ence a cloud reduction and that was of 4.7 °C. It was concluded that
whilst branching of low unsaturated fatty acids containing oils can
lead to a cloud point reduction, biodiesels derived from oils with a
high unsaturated fatty acids content undergo hydrogenation and
trans isomer formation side reactions which significantly increased
the cloud point mitigating any positive effects contributed by
branching. Once again other than the high operating conditions
and lengthy reaction times, the several processing steps require
optimisation [89].
Reaume’s research highlights key drawbacks, besides the seem-
ingly undesirable reactions of trans isomer formation and hydro-
genation needing to be controlled, the extremely high operating
conditions and lengthy reaction times of 260 °C, 15 bar for 6 h
and 300 °C, 40 bar for 16 h for isomerisation and hydroisomerisa-
tion respectively are not ideal. Furthermore, the several extra pro-
cessing steps introduced must be minimised in order to make the
overall process of improving the properties of biodiesel economi-
cally worthwhile on an industrial scale.
Based on the discussion thus far it appears that other than lim-
iting the undesired reactions of dimerisation, hydrogenation,
hydrolysis and trans isomer formation; the key challenges in this
approach remain finding a continuous process, improving the qual-
ity of a heterogeneous mixture and analysing the specific branched
chain isomers formed.
Nevertheless it is critical to note that although when merely
trying to improve the cold flow properties hydrogenation is cor-
rectly considered an undesirable reaction but when attempting
to improve the properties of the biodiesel as a whole (including
oxidative stability) hydrogenation becomes an integral reaction.
Following on, to improve the cold flow properties of saturated
FAME hydroisomerisation is required as opposed to solely isomeri-
sation. The reaction mechanism for hydroisomerisation typically
involves five steps [90]:
1. Saturated molecule is dehydrogenated creating an alkene.
2. Alkene is protonated to create carbocation at double bond site.
3. Carbocation rearranges to form an iso-species by branching.
4. Proton detaches resulting in an iso-alkene.
5. Iso-alkene is hydrogenated to an iso-alkane.
These have been visually illustrated in Fig. 24.
Using specific process and reaction conditions, a metal loaded
acid heterogeneous catalyst for instance a metal loaded zeolite
 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208 205

Fig. 24. Hydroisomerisation reaction mechanism.

could in fact be considered as a tri functional catalyst whereby the
metal hydrogenates unsaturated components, dehydrogenating/r
e-hydrogenating of the saturated components pre and post iso-
merisation and the acid enables the branching to occur.
All in all hydroisomerisation is the preferred solution as not
only is the oxidative stability not lost as is the case in dehydro-
genation, it requires a lower temperature than hydrocracking and
does not form light ends which reduce yields.
3. Supplementary opportunities
Although the core research is making progress there are still
several key areas, mentioned below, which need to be investigated
further to optimise the performance thereby increasing likelihood
of widespread commercialisation.
Firstly a study into the effects of isomerisation pre and post
trans-esterification could be conducted to determine whether it
is better to isomerise the oil or the resulting biodiesel. Lower con-
versions of the biodiesel would indicate that the presence of the
methyl ester group is likely responsible for the inferior reaction
performance. Studies to a similar effect however not comprehen-
sively have been conducted by Reaume et al., which suggests pos-
sible stearic hindrance [53].
Secondly a series of model compounds with the same carbon
chain length but different functional groups could be tested such
as a C 18 hydrocarbon, fatty acid and fatty acid methyl ester as
shown in Fig. 25. A reduction in conversion going from HC to FAME
would indicate the influence of the functional group however if the
conversions are found to be similar then effects of stearic hin-
drance can be negated [91].

3.1. Effects of stearic hindrance (HC, FA and FAME) 3.2. Catalyst type

The effects of stearic hindrance on the conversion of a given
reaction can be determined in a variety of ways. However two
investigations are sufficient to conclude whether the presence of
functional groups play a major role in the activity of porous hetero-
geneous catalyst.
A major issue when using porous heterogeneous catalysts such
as zeolites is the deposit of unwanted molecules which block the
pores reducing the reaction effectiveness. The reaction kinetics
observed in a fresh catalyst is usually much greater than used
reacted catalysts as it can difficult to remove all the coking upon

Fig. 25. Structure and melting point of C18 hydrocarbon, fatty acid and fatty acid methyl ester [2,92].
206 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208
reactivation which can over the course of time accumulate. Fur-
thermore even though zeolites can be reactivated the number of
acid sites available decline upon each use therefore each catalyst
has a shelf life. Despite numerous heterogeneous catalysts having
been studied for the application on biodiesel there are a lack of
deactivation and regeneration studies to accurately determine
the life or extent of effectiveness on reusability.
Moreover other than the study by Wiedemann et al. there are
still knowledge gaps missing into identifying which molecular
components are exactly responsible for causing the catalyst to
deactivate [79]. Once identified, appropriate course of action could
then be taken to maximise conversion by either isolating those
components if possible or perhaps even utilising zeolites with bet-
ter suited pore structures either in shape or size. One method of
ascertaining clues as to the culprit for deactivation would be to
conduct a TGA analysis on a deactivated sample to obtain the mass
of the adsorbed species.
3.3. Reactor type
Majority of the research carried out on the topic of biodiesel
production and improving the cold flow properties using heteroge-
neous catalysis have been done using a batch process. By switching
to a continuous system, in particular a plug flow reactor, a plethora
of opportunities could be unlocked and investigated as multiple
processes stages can be consolidated into a single process step.
Reaume et al. mention the possibility of combining the esterifica-
tion process with hydroisomerisation as they both require solid
acid catalysts and fatty acids as the feed [53]. Furthermore, contin-
uous systems may also facilitate the use of lower operating condi-
tions by enabling greater contact between the reactant and catalyst
such as in a trickle bed reactor system which is currently being
investigated by the Sustainable and Environmental Technologies
Catalysis Research Group at The University of Manchester. With
such developments, as suggested by Ngo et al., enabling use of
lower temperatures and minimising processing steps considerable
ground could be covered in the development of a more practical
and economical process due to savings in energy and material
but also obtaining a better quality product. There are numerous
Fig. 26. TBR multiple bed process combinations.
possibilities and combinations which could be investigated in an
attempt to improve the cold flow properties of lipid derived bio-
fuel, some of which have been highlighted in Fig. 26.
Determining the optimal reaction conditions for each process
and producing an appropriate reactor design would be the key
challenges to bringing such an idea to life [77].
Other than a trickle bed reactor (TBR) there are several on-going
technological advances in reactor systems such as the use of oscil-
latory flow reactor (OFR), membrane reactor, reactive distillation,
reactive absorption, microwave irradiation, ultrasonic technology
and co-solvents which not only could facilitate the trans-
esterification and dewaxing reactions by improved mixing inten-
sity, heat and mass transfer rate, but are also much more energy
efficient [39]. That said from an engineering perspective when
endeavouring to scale-up the laboratory based processes signifi-
cant challenge can be anticipated due to mass and heat transfer
limitations.
3.4. Precise isomer identification
One of the most challenging aspects of analysing the compo-
nents of the isomerised biodiesel is their identification. Although
by using a GC system adequate separation of the components
can be achieved, the current MS analysis based on the existing
spectral libraries do not sufficiently differentiate between the var-
ious isomers. This is partially due to the extremely complex nature
of the analyte; not only is the biodiesel composed of numerous
components which undergo various reactions including isomerisa-
tion; just a single saturated component alone e.g. C 18:0 can theo-
retically form a staggering 2,946,889 isomer possibilities as shown
in Table 6 [93]. Although all of these isomers may not practically
exist due to physical constraints of overcrowding atoms it is extre-
mely difficult to estimate the amount of isomers which can be
ruled out for this reason.
Further complexity is added when dealing with unsaturated
components where not only are branched chain isomers formed
but also positional isomers related to the double bond position
and geometric isomers (cis or trans) are formed. In such cases it
is extremely difficult to differentiate which type of isomer has been
 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208 207

Table 6 [2] Knothe G, Dunn R. A comprehensive evaluation of the melting points of fatty
Number of theoretical isomer possibilities of saturated fatty acids with 12–21 carbon acids and esters determined by differential scanning calorimetry. J Am Oil
atoms. Adapted from [93]. Chem Soc 2009;86:843–56. http://dx.doi.org/10.1007/s11746-009-1423-2.
[3] Leung DY, Wu X, Leung M. A review on biodiesel production using catalyzed
Number of carbon Stereoisomers Non- Total transesterification. Appl Energy 2010;87:1083–95.
atoms stereoisomers isomers [4] Hoekman SK, Broch A, Robbins C, Ceniceros E, Natarajan M. Review of biodiesel
composition, properties, and specifications. Renew Sustain Energy Rev
12 4326 110 4436 2012;16:143–69. http://dx.doi.org/10.1016/j.rser.2011.07.143.
13 12,648 184 12,832 [5] Talebian-Kiakalaieh A, Amin NAS, Mazaheri H. A review on novel processes of
14 37,186 310 37,496 biodiesel production from waste cooking oil. Appl Energy 2013;104:683–710.
15 109,980 520 110,500 http://dx.doi.org/10.1016/j.apenergy.2012.11.061.
16 327,216 876 328,092 [6] Festel G, Würmseher M, Rammer C, Boles E, Bellof M. Modelling production
17 979,020 1,471 980,491 cost scenarios for biofuels and fossil fuels in Europe. J Clean Prod
18 2,944,414 2475 2,946,889 2014;66:242–53. http://dx.doi.org/10.1016/j.jclepro.2013.10.038.
19 8,897,732 4159 8,901,891 [7] Simbolotti G. Production of liquid biofuels. International Energy Agency (IEA),
20 27,004,290 6996 27,011,286 The Energy Technology Systems Analysis Programme (ETSAP), The
21 82,287,516 11,759 82,299,275 International Renewable Energy Agency (IRENA); 2013. p. 1–28.
[8] Azapagic A, Perdan S. Sustainable development in practice: case studies for
engineers and scientists. Wiley; 2011.
[9] Yue D, You F, Snyder SW. Biomass-to-bioenergy and biofuel supply chain
optimization: overview, key issues and challenges. Comput Chem Eng
formed as the mass spectra of isomers can be very similar for dif- 2014;66:36–56. http://dx.doi.org/10.1016/j.compchemeng.2013.11.016.
ferent isomers of the same carbon length. Perhaps more advance [10] Lin Y-C, Huber GW. The critical role of heterogeneous catalysis in
lignocellulosic biomass conversion. Energy Environ Sci 2009;2:68–80. http://
analytical techniques in the coming future may make identification
dx.doi.org/10.1039/B814955K.
of individual isomers relative less complex. [11] Gunstone FD, Harwood JL, Dijkstra AJ. The lipid handbook with CD-ROM. 3rd
ed. Taylor & Francis; 2007.
[12] Clark James H, Comerford JW, Macquarrie DJ. In: Zimmerman Julie B, Anastas
3.5. MP and OS for multiple branched molecules Paul T, editors. Innovations in green chemistry and green engineering. Springer
Science; 2013. p. 37–81 [chapter 3].
Tremendous work has been carried out by G. Knothe, R. Dunn [13] Weitkamp J. Zeolites and catalysis. Solid State Ionics 2000;131:175–88.
[14] Knothe G. ‘‘Designer” biodiesel: optimizing fatty ester composition to improve
and B. Moser in determining the physical properties, more specif- fuel propertiesy. Energy Fuels 2008;22:1358–64.
ically the melting point and oxidative stability of the FA and FAME [15] Verheyen E et al. Molecular shape-selectivity of MFI zeolite nanosheets in n-
molecules [1,2]. Their work however does not include significant decane isomerization and hydrocracking. J Catal 2013;300:70–80. http://dx.
doi.org/10.1016/j.jcat.2012.12.017.
data for branched or multiple branched species; this perhaps is
[16] Murata K, Liu Y, Inaba M, Takahara I. Production of synthetic diesel by
due to the complexity involved in synthesising, identifying and iso- hydrotreatment of jatropha oils using PtRe/H-ZSM-5 catalyst. Energy Fuels
lating such species. Nevertheless in order to catapult this research 2010;24:2404–9. http://dx.doi.org/10.1021/ef901607t.
area forward attaining such information is vital as it can poten- [17] Morgan T et al. Catalytic deoxygenation of triglycerides to hydrocarbons over
supported nickel catalysts. Chem Eng J 2012;189–190:346–55. http://dx.doi.
tially open up the avenues of theoretically and then practically org/10.1016/j.cej.2012.02.027.
reverse engineering FAME compositional profiles to give ideal [18] Mohammad M, Kandaramath Hari T, Yaakob Z, Chandra Sharma Y, Sopian K.
quality biodiesel. Overview on the production of paraffin based-biofuels via catalytic
hydrodeoxygenation. Renew Sustain Energy Rev 2013;22:121–32. http://dx.
doi.org/10.1016/j.rser.2013.01.026.
4. Conclusion [19] Liu J, Liu C, Zhou G, Shen S, Rong L. Hydrotreatment of Jatropha oil over
NiMoLa/Al2O3 catalyst. Green Chem 2012;14:2499. http://dx.doi.org/10.1039/
c2gc35450k.
By being able to overcome the challenges hindering the utilisa- [20] Liu C, Liu J, Zhou G, Tian W, Rong L. A cleaner process for hydrocracking of
tion of pure biodiesel (B100) without having to blend with crude jatropha oil into green diesel. J Taiwan Inst Chem Eng 2013;44:221–7. http://
dx.doi.org/10.1016/j.jtice.2012.10.006.
oil derived diesel will not only increase the energy security for [21] Knothe G. Biodiesel and renewable diesel: a comparison. Prog Energy Combust
non-OPEC countries but will also contribute massively in resisting Sci 2010;36:364–73. http://dx.doi.org/10.1016/j.pecs.2009.11.004.
climate change by helping reduce the overall greenhouse gas emis- [22] Gong S, Shinozaki A, Shi M, Qian EW. Hydrotreating of Jatropha oil over
alumina based catalysts. Energy Fuels 2012;26:2394–9. http://dx.doi.org/
sions. Following the COP21 Paris 2015 agreement world leaders 10.1021/ef300047a.
have in principal made a bold statement in opting to prioritise [23] Gong S, Chen N, Nakayama S, Qian EW. Isomerization of n-alkanes derived
addressing climate change over sole economic welfare – if imple- from jatropha oil over bifunctional catalysts. J Mol Catal A: Chem
2013;370:14–21. http://dx.doi.org/10.1016/j.molcata.2012.11.019.
mented a huge surge in demand for bio based products such as bio-
[24] Bezergianni S, Dimitriadis A. Comparison between different types of
diesel is predicted. Improving the cold flow properties in the most renewable diesel. Renew Sustain Energy Rev 2013;21:110–6. http://dx.doi.
effective manner is a fundamental step in enabling the realisation org/10.1016/j.rser.2012.12.042.
[25] Bezergianni S. Catalytic hydroprocessing of liquid biomass for biofuels
of such an innovation to global markets. In terms of using hetero-
production; 2013.
geneous catalysis, skeletal-isomerisation and selectively hydroc- [26] No S-Y. Application of hydrotreated vegetable oil from triglyceride based
racking provide the most enticing theoretical solutions. With biomass to CI engines – a review. Fuel 2014;115:88–96. http://dx.doi.org/
rapid continual advancements by endeavouring researchers the 10.1016/j.fuel.2013.07.001.
[27] Gunstone FD. Biodiesel-Oils and fats in the market place; 2013. <http://
gap between theory and application is ever narrowing; with lipidlibrary.aocs.org/OilsFats/content.cfm?ItemNumber=39445>.
optimal advanced economically feasible commercial enhanced [28] Administration, U.S.E.I. Monthly biodiesel production report with data for
biodiesel production processes expected to be widespread in the September 2015. Washington, DC: U.S. Department of Energy; 2015. p. 1–10.
[29] Administration, U.S.E.I. November 2015 monthly energy review. Washington,
very near future. DC: U.S. Department of Energy, 20585; 2015. p. 1–234.
[30] Nations U. In: Framework convention on climate change. Conference of the
parties 21, ed. Paris: UNFCCC; 2015. p. 1–31.
Acknowledgement
[31] Lin L, Cunshan Z, Vittayapadung S, Xiangqian S, Mingdong D. Opportunities
and challenges for biodiesel fuel. Appl Energy 2011;88:1020–31. http://dx.doi.
The authors would like to thank EPSRC (Engineering and org/10.1016/j.apenergy.2010.09.029.
Physical Sciences Research Council) UK for their sponsorship. [32] Keystone. Rudolf Diesel: Wie der Erfinder die Schweiz veränderte; 2013.
<http://www.handelszeitung.ch/vermischtes/wie-rudolf-diesel-die-schweiz-
veraenderte-503689>.
References [33] Gholami Z, Abdullah AZ, Lee K-T. Dealing with the surplus of glycerol
production from biodiesel industry through catalytic upgrading to
polyglycerols and other value-added products. Renew Sustain Energy Rev
[1] Moser BR. Comparative oxidative stability of fatty acid alkyl esters by
2014;39:327–41. http://dx.doi.org/10.1016/j.rser.2014.07.092.
accelerated methods. J Am Oil Chem Soc 2009;86:699–706.
208 A. Anwar, A. Garforth / Fuel 173 (2016) 189–208
[34] Eia. U.S. Energy Information Administration, International Energy Statistics;
2015.
[35] Carels N. Jatropha curcas: a review. Adv Botan Res 2009;50:39–86.
[36] Yaakob Z, Mohammad M, Alherbawi M, Alam Z, Sopian K. Overview of the
production of biodiesel from waste cooking oil. Renew Sustain Energy Rev
2013;18:184–93.
[37] Atabani AE et al. A comprehensive review on biodiesel as an alternative energy
resource and its characteristics. Renew Sustain Energy Rev 2012;16:2070–93.
http://dx.doi.org/10.1016/j.rser.2012.01.003.
[38] Transport, D.D.f. Carbon and sustainability reporting within the renewable
transport fuel obligation. London; 2008.
[39] Lam MK, Lee KT, Mohamed AR. Homogeneous, heterogeneous and enzymatic
catalysis for transesterification of high free fatty acid oil (waste cooking oil) to
biodiesel: a review. Biotechnol Adv 2010;28:500–18. http://dx.doi.org/
10.1016/j.biotechadv.2010.03.002.
[40] Calero J et al. An overview on glycerol-free processes for the production of
renewable liquid biofuels, applicable in diesel engines. Renew Sustain Energy
Rev 2015;42:1437–52. http://dx.doi.org/10.1016/j.rser.2014.11.007.
[41] Bagheri S, Julkapli NM, Yehye WA. Catalytic conversion of biodiesel derived
raw glycerol to value added products. Renew Sustain Energy Rev
2015;41:113–27. http://dx.doi.org/10.1016/j.rser.2014.08.031.
[42] Atadashi IM, Aroua MK, Aziz AA. High quality biodiesel and its diesel engine
application: a review. Renew Sustain Energy Rev 2010;14:1999–2008. http://
dx.doi.org/10.1016/j.rser.2010.03.020.
[43] Oil I. Thermal-oxidative storage stability of bio-diesel fuels. Ontario,
Canada: Natural Resources Canada; 2009. p. 3–4.
[44] Rashed MM et al. Stability of biodiesel, its improvement and the effect of
antioxidant treated blends on engine performance and emission. RSC Adv
2015;5:36240–61. http://dx.doi.org/10.1039/C4RA14977G.
[45] Serrano M, Bouaid A, Martínez M, Aracil J. Oxidation stability of biodiesel from
different feedstocks: influence of commercial additives and purification step.
Fuel 2013;113:50–8. http://dx.doi.org/10.1016/j.fuel.2013.05.078.
[46] Sorate KA, Bhale PV. Biodiesel properties and automotive system compatibility
issues. Renew Sustain Energy Rev 2015;41:777–98. http://dx.doi.org/10.1016/
j.rser.2014.08.079.
[47] Kazancev K, Makareviciene V, Paulauskas V, Janulis P. Cold flow properties of
fuel mixtures containing biodiesel derived from animal fatty waste. Eur J Lipid
Sci Technol 2006;108:753–8. http://dx.doi.org/10.1002/ejlt.200600074.
[48] Dunn RO. Improving the cold flow properties of biodiesel by
fractionation. INTECH Open Access Publisher; 2011.
[49] Knothe G. Dependence of biodiesel fuel properties on the structure of fatty
acid alkyl esters. Fuel Process Technol 2005;86:1059–70. http://dx.doi.org/
10.1016/j.fuproc.2004.11.002.
[50] Yori JC, D’Amato MA, Grau JM, Pieck CL, Vera CR. Depression of the cloud point
of biodiesel by reaction over solid acids. Energy Fuels 2006;20:2721–6. http://
dx.doi.org/10.1021/ef060245i.
[51] Noureddini H. System and process for producing biodiesel fuel with reduced
viscosity and a cloud point below thirty-two (32) degrees fahrenheit. Chem
Biomol Eng Res Publ 2001;2.
[52] Knothe G, Krahl J, Gerpen JV. Biodiesel handbook. 2nd ed. AOCS Press; 2010,
ISBN 978-1-893997-62-2.
[53] Reaume SJ, Ellis N. Use of hydroisomerization to reduce the cloud point of
saturated fatty acids and methyl esters used in biodiesel production. Biomass
Bioenergy 2013;49:188–96. http://dx.doi.org/10.1016/j.
biombioe.2012.12.008.
[54] Smith PC, Ngothai Y, Dzuy Nguyen Q, O’Neill BK. Improving the low-
temperature properties of biodiesel: methods and consequences. Renewable
Energy 2010;35:1145–51. http://dx.doi.org/10.1016/j.renene.2009.12.007.
[55] Bhale PV, Deshpande NV, Thombre SB. Improving the low temperature
properties of biodiesel fuel. Renewable Energy 2009;34:794–800. http://dx.
doi.org/10.1016/j.renene.2008.04.037.
[56] Arbab MI et al. Fuel properties, engine performance and emission
characteristic of common biodiesels as a renewable and sustainable source
of fuel. Renew Sustain Energy Rev 2013;22:133–47. http://dx.doi.org/10.1016/
j.rser.2013.01.046.
[57] Gunstone FD, Harwood JL, Padley FB. The lipid handbook. 2nd ed. Taylor &
Francis; 1994.
[58] Gunstone FD. An introduction to the chemistry of fats and fatty acids. Wiley;
1958.
[59] Gunstone F. The chemistry of oils and fats: sources, composition, properties
and uses. Wiley; 2009.
[60] Edith O, Janius RB, Yunus R. Factors affecting the cold flow behaviour of
biodiesel and methods for improvement – a review. Pertan J Sci Technol
2012;20:1–14.
[61] Macias EE, Deshmane CA, Jasinski JB, Carreon MA, Ratnasamy P. Catalytic
transformations of methyl oleate and biodiesel over mesoporous gallium–
niobium oxides. Catal Commun 2011;12:644–50. http://dx.doi.org/10.1016/
j.catcom.2010.12.005.
[62] Weitkamp J. In: Hydrocracking and hydrotreating. ACS symposium series, vol.
20. American Chemical Society; 1975. p. 1–27 [chapter 1].
[63] Speight J. Handbook of industrial hydrocarbon processes. Elsevier; 2010.
[64] Seo M-G, Lee D-W, Lee K-Y, Moon DJ. Pt/Al-SBA-15 catalysts for hydrocracking
of C21–C34 n-paraffin mixture into gasoline and diesel fractions. Fuel
2015;143:63–71. http://dx.doi.org/10.1016/j.fuel.2014.11.028.
[65] Sahu R, Song BJ, Im JS, Jeon Y-P, Lee CW. A review of recent advances in
catalytic hydrocracking of heavy residues. J Indust Eng Chem. http://dx.doi.
org/10.1016/j.jiec.2015.01.011.
[66] Ishihara A, Fukui N, Nasu H, Hashimoto T. Hydrocracking of soybean oil using
zeolite–alumina composite supported NiMo catalysts. Fuel 2014;134:611–7.
http://dx.doi.org/10.1016/j.fuel.2014.06.004.
[67] Weitkamp J. Catalytic hydrocracking—mechanisms and versatility of the
process. ChemCatChem 2012;4:292–306. http://dx.doi.org/10.1002/
cctc.201100315.
[68] Shuguang Zhang ZZ. Dale Steichen. Skeletal isomerization of fatty acids. The
United States of America patent 01/00780; 2002.
[69] Shuguang Zhang ZZ. Fatty acid isomerization with mesoporous zeolites. The
United States of America patent US 6723862 B2; 2004.
[70] Zongchao Conrad Zhang, Dery Maurice, Zhang Shuguang, Steichen Dale. New
process for the production of branched-chain fatty acids. J Surfactants Deterg
2004;7:211–5.
[71] Yori Juan C, D’Amato MA, Grau Javier M, Pieck Carlos L, Vera Carlos R.
Depression of the cloud point of biodiesel by reaction over solid acids. Energy
Fuels 2006:2721.
[72] Shuguang Zhang ZCZ. Skeletal isomerisation of unsaturated fatty acids: the
role of mesopores in HBeta zeolites. Catal Lett 2007;115:114–21.
[73] Ngo HL, Nuñez A, Lin W, Foglia TA. Zeolite-catalyzed isomerization of oleic
acid to branched-chain isomers. Eur J Lipid Sci Technol 2007;109:214–24.
http://dx.doi.org/10.1002/ejlt.200600246.
[74] Ngo HL, Dunn RO, Sharma B, Foglia TA. Synthesis and physical properties of
isostearic acids and their esters. Eur J Lipid Sci Technol 2011;113:180–8.
http://dx.doi.org/10.1002/ejlt.201000335.
[75] Ngo HL, Dunn RO, Hoh E. C18-unsaturated branched-chain fatty acid isomers:
characterization and physical properties. Eur J Lipid Sci Technol
2013;115:676–83.
[76] Ngo HL. Improved zeolite regeneration processes for preparing saturated
branched-chain fatty acids. Eur J Lipid Sci Technol 2014;116:645–52.
[77] Ngo HL, Yee WC, McAloon AJ, Haas MJ. Process and cost modeling of saturated
branched-chain fatty acid isomer production. Ind Eng Chem Res
2012;51:12041–5.
[78] Ngo HL, Yee WC, McAloon AJ, Haas MJ. Techno-economic analysis of an
improved process for producing saturated branched-chain fatty acids. J Agric
Sci 2014;6:158–68. http://dx.doi.org/10.5539/jas.v6n10p158.
[79] Wiedemann SCC et al. Skeletal isomerisation of oleic acid over ferrierite in the
presence and absence of triphenylphosphine: pore mouth catalysis and related
deactivation mechanisms. J Catal 2014;316:24–35. http://dx.doi.org/10.1016/
j.jcat.2014.04.018.
[80] Wiedemann SCC et al. Skeletal isomerisation of oleic acid over ferrierite:
Influence of acid site number, accessibility and strength on activity and
selectivity. J Catal 2015;329:195–205. http://dx.doi.org/10.1016/j.
jcat.2015.05.013.
[81] Cason J, Winans WR. Branched-Chain Fatty Acids. XIV. Location of branching
methyl groups by study of the melting points of binary mixtures of branched
and normal acids or amides. J Org Chem 1950;15:148–58. http://dx.doi.org/
10.1021/jo01147a024.
[82] Cason J, Winans WR. Branched-Chain Fatty Acids. XIII. Preparation of branched
and normal acids for use in the study of melting points of binary mixtures.
Completion of the methyloctadecanoic acid series. J Org Chem
1950;15:139–47. http://dx.doi.org/10.1021/jo01147a023.
[83] Cason J, Pippen EL, Taylor PB, Winans WR. Branched-Chain Fatty Acids. XII.
Synthesis in the methyloctadecanoic acid series. J Org Chem 1950;15:135–8.
http://dx.doi.org/10.1021/jo01147a022.
[84] Paul Upham PT, Tomei Julia, Boucher Philip. Substitutable biodiesel feedstocks
for the UK: a review of sustainability issues with reference to the UK RTFO. J
Clean Prod 2009:S37–45.
[85] Macias Eugenia E, Deshmane CA, Jasinski Jacek B, Carreon Moises A,
Ratnasamy Paul. Catalytic transformations of methyl oleate and biodiesel
over mesoporous gallium–niobium oxides. Catal Commun 2011;12:644–50.
[86] Reaume Stephen J, Ellis N. Optimizing reaction conditions for the
isomerization of fatty acids and fatty acid methyl esters to their branch
chain products. J Am Oil Chem Soc 2011;88:661–71.
[87] Reaume Stephen J, Ellis N. Synergistic effects of skeletal isomerization on oleic
and palmitic acid mixtures for the reduction in cloud point of their methyl
esters. Energy Fuels 2012;26:4514–20.
[88] Reaume Stephen J, Ellis N. Use of hydroisomerization to reduce the cloud point
of saturated fatty acids and methyl esters used in biodiesel production.
Biomass Bioenergy 2013;49:188–96.
[89] Reaume Stephen J, Ellis N. Use of isomerization and hydroisomerization
reactions to improve the cold flow properties of vegetable oil based biodiesel.
Energies 2013;6:619–33.
[90] Reaume S, Ellis N. Use of isomerization and hydroisomerization reactions to
improve the cold flow properties of vegetable oil based biodiesel. Energies
2013;6:619–33. http://dx.doi.org/10.3390/en6020619.
[91] Mjøs SA, Pettersen J. Improved methods for analysis of fatty acid isomers. SSF;
2001.
[92] Schobert H. Chemistry of fossil fuels and biofuels. Cambridge University Press;
2012, ISBN 978-0-521-11400-4.
[93] Abrahamsson S, Ställberg-Stenhagen S, Stenhagen E. The higher saturated
branched chain fatty acids. Prog Chem Fats Other Lipids 1964;7:1–157.