Polymerization of sunflower oil diesel fuel
by Joan Patricia French Keller
A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in
Chemical Engineering
Montana State University
© Copyright by Joan Patricia French Keller (1986)
Abstract:
The mechanism of insoluble gel formation in hydrocarbon basestock lubricating oil contaminated with
sunflower oil was studied in a laboratory apparatus simulating the conditions of a diesel engine
crankcase. Two distinct and separate phases formed within the system when using basestock oil as the
lubricating substrate - a solid insoluble gel phase and a supernatant liquid phase.
The research was conducted to understand and characterize the physical and chemical differences
between polymer species contributing to viscosity and those contributing to insoluble gel. Addition
polymerization was known to yield viscosity rise at conditions of this work. A theory was developed
which hypothesized simultaneous oxidation of addition polymers in basestock oil to yield more polar
compounds which formed the separate gel phase.
Experiments supported the polar gel theory. Attempts to homogenize or disperse the gel in basestock or
commercial lube oils failed to show similarity to the physical behavior of non-gel addition polymers.
Infrared spectroscopy also showed that gel contained more carbonyl groups than pure sunflower oil or
addition polymerized sunflower oil.
Antioxidant and free radical initiator trials indicated gel was chemically different from addition
polymerized sunflower oil, with the presence of oxygen being key to gel formation. A long chain
amine was successful in preventing gel formation. When the acidic addition polymers were converted
to less polar amides, the oil mixture remained a single phase. These results generally confirm that the
polymers resulting from addition polymerization are polarized by oxidation to form the separate gel
phase. POLYMERIZATION
OF SUNFLOWER OIL DIESEL FUEL
by
Joan Patricia French Keller
A thesis submitted in partial fulfillment
of the requirements for the degree
of
Master of Science
in
Chemical Engineering
MONTANA STATE UNIVERSITY
Bozeman, Montana
December 1986
MAIN LIB.
Lop - X
ii
APPROVAL
of a thesis submitted by
Joan French Keller
This thesis has been read by each member of the thesis
committee and has been found to be satisfactory regarding
content, English usage, format, citation, bibliographic
style and consistency, and is ready for submission to the
College of Graduate Studies.
-hereDate
%
mSC,
Chairperson, GraduateTfommittee
Approved for the Major Department
. IZJMt
Date
CuX. T" xuiasui.
Head(y Major Department
Approved for the College of Graduate Studies
Date
Graduate Dean
iii
STATEMENT OF PERMISSION TO USE
In presenting this thesis
requirements
for
a
in partial fulfillment of the
master's
degree
at
Montana
State
University, I agree that the Library shall make it available
to borrowers under rules
from this thesis
are
of
the Library.
Brief quotations
allowable without special permission,
provided that accurate acknowledgment of source is made.
Permission for extensive
of this thesis may be
quotation from or reproduction
granted
his absence, by the Dean
of
by my major professor, or in
Libraries when, in the opinion
of either, the proposed use of the material is for scholarly
purposes.
for
Any copying or use of the material in this thesis
financial
gain
permission.
Signature.
Date
. m n
QiCjLmlo^J
shall
not
be
allowed
without
my
iv
ACKNOWLEDGMENTS
The author would like to
the
Chemical
Engineering
University for their
and
encouragement
Department
guidance
given
research by my advisor.
appreciated.
thank the faculty and staff of
Dr.
Daniel
the
State
The advice
course
of
this
L. Shaffer, is greatly
The author also wishes to thank Mr. Sid Schiff
Montana State
University's
recomoendations.
for assistance
and
with
the
is
infrared
support
Natural
Dr.
Chemistry
Special thanks
financial
Department of
Montana
assistance.
throughout
of Phillips Petroleum Company
generous
and
at
Department for their
extended to Tom Mendes
spectroscopy.
received
Resources
Paul W. Jennings of
and
from
the
Finally,
Montana
Conservation and the
National Science Foundation is gratefully acknowledged.
V
TABLE OF CONTENTS
Page
APPROVAL..................
ii
STATEMENT OF PERMISSION TO USE.............
iii
ACKNOWLEDGMENTS............................
iv
TABLE OF CONTENTS..... .......................
LIST OF TABLES..........
LIST OF FIGURES......
v
vi
vii
ABSTRACT...............
x
INTRODUCTION.... ..........
I
RESEARCH OBJECTIVES... .....................
6
THEORY............
7
Oxidative Polymerization...............
Other Oxidation Reactions......
7
15
EXPERIMENTAL.................... . ...j_____
22
Equipment.......
Materials.......... ........ ....... t...
22
28
RESULTS AND DISCUSSION,..... ...............
30
SUMMARY....... .......................... ...
77
CONCLUSIONS................................
79
SUGGESTIONS FOR FUTURE RESEARCH. ............
81
LITERATURE CITED........ . ..................
82
APPENDIX..... ............... ...___ ..._____
86
vi
LIST OF TABLES
Table
Page
1.
Fatty Acid Distribution in Sunflower Oil....
9
2.
Summary of Experiments with Additives.......
49
3.
Atomic Emission Data.......................
52
4.
Operating Parameters for Oil Bath Runs......
86
vii
LIST OF FIGURES
Figure
'
Page
1.
Reaction Kettle...........................
24
2.
Oil Bath and Reaction Kettle..............
24
3.
Oil Bath and Operating Diagram........ .
26
4.
Viscosity of commercial oil and 5.0
sunflower oil vs. time for standard
conditions.................................
32
Viscosity of basestock oil and 5.0%
sunflower oil vs. time for standard
conditions........................
33
Viscosity of basestock oil and 25.0%
sunflower oil vs. time for standard
conditions at 150 C and 135 C.............
37
Viscosity of supernatant/gel mixture vs.
time of homogenization....................
40
Viscosity of basestock oil and 5.0%
- sunflower oil vs. time for standard
conditions with initial addition of
1.0% ZDTC.................... ............
45
Viscosity of basestock oil and 5.0%
sunflower oil vs. time for standard
conditions with nitrogen and additions of
0.5% Lupersol 130 every 4hours............
47
5.
6.
7.
8.
9.
viii
LIST OF FIGURES— Continued
Figure
Page
10. Viscosity comparison of old and new
basestock oils and 5.0 % sunflower oil
vs. time for standard conditions.........
53
11. Viscosity comparison of old arid new
basestock oils and 5.0% sunflower oil
vs. time for standard conditions with
nitrogen and additions of 0.5% Lupersol
130 every 4 hours..... ...................
57
12. Viscosity comparison of old and new
basestock oils and 5.0% sunflower oil
vs. time for standard conditions with
initial addition of 1.0% ZDTP.... ........
61
13. Viscosity of new basestock oil and 5.0%
sunflower oil vs. time for standard
conditions with nitrogen and addition
of 0.5% Lupersol 130. Nitrogen
switched to oxygen at times indicated
by arrows..................
............
68
14. Viscosity of new basestock oil and 5.0%
sunflower oil vs. time for standard
conditions with 15 g ODA added
initially......... .......................
70
15. Viscosity of new basestock oil and 5.0%
sunflower oil vs. time for standard
conditions with additions of 0.7 g ODA
every 4 hours....... .....................
72
16. Total base riumber of new basestock oil
and 5.0% sunflower oil vs. time for
standard conditions with additions
of 0.7 g ODA every 4 hours...... .
73
ix
LIST OF FIGURES— Continued
Figure
17. Infrared spectroscopy of equal
concentrations of pure sunflower oil,
addition polymerized sunflower oil and
insoluble gel from sunflower oil.........
Page
76
X
ABSTRACT
The mechanism of insoluble gel formation in hydrocarbon
basestock lubricating oil contaminated with sunflower oil
was studied in a
laboratory apparatus simulating the
conditions of a diesel engine crankcase. Two distinct and
separate phases formed
within
the system when using
basestock oil as the lubricating substrate - a solid
insoluble gel phase and a supernatant liquid phase.
The
research
was
conducted
to
understand
and
characterize the physical and chemical differences between
polymer species
contributing
to
viscosity and those
contributing to insoluble gel. Addition polymerization was
known to yield viscosity rise at conditions of this work. A
theory was
developed
which
hypothesized simultaneous
oxidation of addition polymers in basestock oil to yield
more polar compounds which formed the separate gel phase.
Experiments supported the polar gel theory. Attempts to
homogenize or disperse the gel in basestock or commercial
lube oils failed to show similarity to the physical behavior
of non-gel addition polymers.
Infrared spectroscopy also
showed that gel contained more carbonyl groups than pure
sunflower oil or addition polymerized sunflower oil.
Antioxidant and free radical initiator trials indicated
gel was chemically different from addition polymerized
sunflower oil, with the presence of oxygen being key to gel
formation. A long chain amine was successful in preventing
gel formation.
When the acidic addition polymers were
converted to less polar amides, the oil mixture remained a
single phase.
These results generally confirm that the
polymers
resulting
from
addition
polymerization are
polarized by oxidation to form the separate gel phase.
I
. I M rRQDUCTIOM
Recently, fuel
costs
have
declined,
but an increased
awareness that the supply of petroleum-based fuels is finite
has sparked interest in finding
new sources of motor fuels.
Vegetable
diesel
oils
as
alternate
engine
fuels
received modest interest for several decades Zll .
economic factors
have
favored
the
have
However,
use of petroleum-based
fuels C2] .
The use of vegetable oils as fuels for diesel engines is
not a new concept.
As
far back as 1912, Rudolf Diesel, the
' I
'
inventor of the diesel engine, tried using vegetable oils as
diesel fuels, but economics
and
design never favored their
use C33 .
The development of the
the availability of
turn has been
engines.
engine has been based on
petroleum-derived
tailored
During
diesel
to
this
meet
the
period,
knowledge has been developed
a
that
diesel fuel which in
needs of the current
wealth
serves
of
empirical
as the data base
for the current diesel fuel specifications E43 .
Two principle problems
have
been identified with using
vegetable oils directly as diesel fuels:
form
carbon
deposits
inside
the
I)
vegetable oils
combustion
chambers of
2
direct injection engines and 2)
the
crankcase
polymerize
in
vegetable oils carried into
the
thickened oil mixture plugs the
sticking and plugs
crankcase.
orifices
These
failure.
The
problems
oil.
The
oil filter, causes oil ring
leading
problems
above
lubricating
into
and
out of the
can
cause
eventual
engine
are
related to the chemical
structure differences between vegetable oils and diesel fuel
C53 .
Positive aspects of
natural state is
heat content is
vegetable
liquid
and
oils
hence
comparable
to
widespread availability and
4)
as
fuels are:
I)
easily transported, 2)
diesel
fuel, 3)
potential
renewability as resources
[63 .
Studies have been
decarboxylation of
either of which
performed
the
vegetable
increases
processed vegetable
on transesterification and
oils
oils
for fuel purposes,
cost.
Direct use of minimally
should
permit on-farm processing
and minimize costs £5,7,8,93 .
This
research
laboratory.
thickens
to
may
£10,11,123 .
conditions
part
of
continuing
work
at
this
Previous workers confirmed that lubrication oil
due
thickening
is
vegetable
cause
These
an
workers
consisting
of
oil
contamination.
This
unacceptable
viscosity
developed
set of standard
variables
a
known
to
rise
strongly
influence the thickening of lubricating oil due to vegetable
3
oil contamination.
The
factors
considered
conditions
were
catalysts.
Rewolinski
temperature,
temperature because
CIO]
150
encountered by the oil
and
engine
viscosity
strongly
in
is
areas.
vegetable
influenced
by
the
As
oxygen
oxygen
was
flow
viscosity rise increased.
affect
viscosity
oxygen
flow
investigated
concentration.
2.0
the
effects
As
vegetable
the rate of viscosity rise
viscosity rise
in
a
oil
and
as the standard
through test oil
increased,
the
rate of
presence of nitrogen did not
conditions
ml/sec.
of
include an
Rewolinski
varying
also
vegetable
oil
oil concentration increased,
increased.
reasonable
conditions include 5.0 weight
also showed
polymerization was
%
of oxygen.
In a
percolated
Standard
of
C
presence
rate
The
rise.
rate
environment
Rewolinski
to
mixtures.
standard
a rough average temperature
due
standard exposure
150
the
it travels through the crankcase
combustion
rise
chemical
chose
C
as
developing
To get a measurable
period
percent
of time, standard
sunflower oil in the
lubricating oil.
Jette's research fill
catalyst in the
metal, and
system.
Rewolinski
focussed
Copper
had
on
is
determined
important polymerization catalyst than
the
role of copper
a common engine wear
copper
iron.
was
a more
Jette went on
4
to determine
that
catalyst form.
soluble
He
copper
used
viscosity rise increased
copper
with
was
foil
is
present
control the amount of
in
the
metal
and
increased
As a result of this combined research,
foil strip
the most important
observed that
copper foil area.
a 2 cm x 5 cm copper
standard
present
conditions.
To
in the system, all of
the equipment in contact with the oil mixture is glass.
Lubrication oil thickening
the engine design or perhaps
may
be
reduced by changing
by changing the chemical make­
up of the lubricating oil.
Engine design modifications are
costly; therefore, alteration of the lubrication oil is more
feasible [13] .
oils and
Extensive work has been done on lubricating
conventional
diesel
specific antioxidants,
have been developed.
fuel
dispersants
systems
and
where system-
metal deactivators
There is a need to explore these areas
with vegetable oil fuels.
The equipment
environment of a
copper,
oxygen
used
at
this
crankcase.
flow
laboratory
Variables
rate,
amount
of
simulates the
such as amount of
sunflower
oil and
temperature can be controlled to a greater degree than in an
actual engine.
Simulation also allows for repeated tests in
the same apparatus as
well
as avoiding the recurring costs
of replacing expensive engines upon their failure.
The use of hydrocarbon
substrate is desirable
if
basestock as the lubricating oil
a
complete understanding of the
5
contaminated system chemistry is sought. This is due to the
unknown chemical
nature
of
the
additive
package
in the
commercial Iuhe oil.
Dutta
CI23
contaminated
attempted
with
5.0
as
the
precipitated
sunflower
out
measurements to
as
a
quantify
for this two-phase system.
only problem
Dutta
was
swollen
of
was
with
hydrocarbon
percent
two
oil
oil and
phases.
polymerized,
separate
basestock
sunflower
distinct
phase.
and
A gel
this gel
Viscosity
polymerization became meaningless
Measuring viscosity was not the
encountered.
amount of gel formed
gel
use
weight
encountered the formation
formed
to
He
also discovered the
difficult to quantify because the
lubricating
oil.
Until
the
mechanism(s) of gel formation in the current lubricating oil
system is understood, viscosity
quantify polymerization of
are meaningless.
measurements as a method to
sunflower
oil
in basestock oil
6
RESEARCH OBJECTIVES
This
research
was
conducted
mechanism(s) of contaminant
a lubrication oil system.
the chemical nature
formation as
to
understand
the
vegetable oil polymerization in
A primary goal was to understand
of
gel
precipitation versus viscosity
vegetable
oil
polymerization
occurs
in the
given system.
A further
objective
was
to
hydrocarbon basestock lubricating
the impact of unknown chemical
accomplish this objective,
the
must be sharply minimized.
As
is understood,
gel
formation
be
able
oil
to
utilize the
in future studies so
additives is eliminated.
formation
To
of insoluble gel
the chemistry of the system
and
viscosity
eliminated by future research findings.
rise
may be
7
THEORY
The presence of
oxygen
in
contaminated with vegetable
chemical reactions.
One
a diesel lubrication system
oil
of
may
produce
these
a variety of
reactions is oxidative
polymerization where oxygen interacts
with the double bonds
of vegetable
the
formation of addition
reactions
may occur when oxygen
polymers.
oils
to
Other oxidation
cleaves a vegetable
acids.
initiate
oil
Ketones
can
The
possible
cleavage.
double
also
be
bond
to form aldehydes or
formed
without double bond
reactions
of
vegetable
oils and
oxygen will be discussed in the following sections.
Oxidative Polymerization
Current
chemically
diesel
fuels
different
contains hydrocarbons
are
from
which
petroleum-derived
vegetable
are
oils.
arranged
and
are
Diesel fuel
in
straight or
\
branched chains.
It
is
usually
may contain some aromatics.
hand,
are
(glycerol
water-insoluble,
esters
of
fatty
paraffinic in nature, but
Vegetable
oils, on the other
hydrophobic
acids).
A
triglycerides
vegetable
oil
8
triglyceride is
approximately
three
times
larger
than a
typical diesel fuel component E143 .
Vegetable oil can be thought of as a reaction product of
glycerol and fatty acids.
CH0-OH
I 2
CH0-OOCR1
HOOC-R1
+
HOOC-R0
I
--- >
Water
scheme, R^,
even numbered hydrocarbon
chains
upon
They are typically different
bonds.
The
triglyceride molecule
of
bonds.
The
molecular
acids that are
The size of R^, R2 and
particular vegetable oil.
in
chain length and number of
of
vary
weight
molecule is 750 to 1000.
fatty
the
degree
can
h Triglyceride
R2 and R3 symbolize the
usually 16 to 22 carbons in length.
depending
2
CH2-OOCR3
Fatty Acids
In the above reaction
1
CH-OOCR0
HOOC-R3
Glycerol
R3 may vary
+
3H0H
2
CH2-OH
double
I 2
I
1
CH-OH
unsaturation
from
of
a
zero
of
one
to nine double
typical triglyceride
The fatty acids contribute roughly
95% of the total weight
of
the physical and chemical
the molecule and influence both
properties
of the vegetable oils
[143 .
The current
contaminant
system.
research
vegetable
Sunflower
oil
oil's
constituents are oleic,
is
using
in
the
primary
linoleic
sunflower
oil
simulated
lubricating
unsaturated
and
as the
fatty
linolenic.
acid
An oleic
9
fatty acid is an eighteen-carbon
bond while linoleic has
two
three double bonds E143 .
fatty acid with one double
double bonds and linolenic has
Compositions of typical sunflower
molecules are shown in the following table.
Table Is
Fatty Acid Distribution in Sunflower Oil
Sunflower 2a *
Sunflower I*
Fatty Acid
.
Palmitic
6.0
6.4
Stearic
4.2
4.2
Oleic
18.7
23.9
Lirioleic
69.3
61.4
Linolenic
0.3
3.0
Eicosenoic
0.1
———
AKaufman and Ziejewski C153
AAPeterson, Wagner and Auld C133
The double bonds in the sunflower oil may be attacked by
oxygen.
This
autoxidation
autocatalytic.
process
is
because
the
When
vegetable
result is addition
sometimes
oxidation
polymerization
oils
to
mechanism
as
is
are autoxidized, the
which
radical, hydroperoxide mechanism [143 .
referred
occurs
by a free
10
Oxidative
polymerization
of
vegetable
oil
occurs as
described below E163 .
1.
The initiation
preceded
to
the
by an
of
the
induction
presence
of
oxidative chain reaction is
period which has been attributed
natural
antioxidants.
There
are no
detectable changes in the vegetable oil physical or chemical
properties.
The
induction
period
may
be
eliminated by
adding a small quantity of a hydroperoxide.
2.
The double
bonds
are
directly attacked by oxygen,
and hydroperoxides are formed.
As oxygen is consumed, the
polymerization reaction may be detected.
3.
The hydroperoxides decompose
to free radicals.
The
decomposition of these hydroperoxides causes the reaction to
become autocatalytic.
4.
High
molecular
weight,
formed by polymerization,
and
cross-linked polymers are
scission reactions yield low
molecular weight compounds such as carbonyls and hydroxys.
The
initiation
controversial topic.
of
the
of
the
unsaturated
peroxides E173 .
chain
reaction
an initial attack on the double
fatty
acids
to
form
This reaction is depicted below.
-CH2-CH=CH-
is a
The autoxidation of vegetable oils was
first thought to consist of
bonds
radical
+
O2
--- >
-CH^-CH-CH-
2
I
I
0— 0
cyclic
11
Later
work
showed
the
noncyclic, alpha-methylenic
bonds still intact C183
intact,
this
would
Initial
products
hydroperoxides
.
If
imply
the
to
contain
with the double
the double bond were still
alpha-methylenic
hydrogen bond was broken.
This
kcal/mole.
little available energy to break
Since there is
this relatively
strong
the oxygen directly
bond,
bond
carbon-
has a strength of 80
investigators have concluded
attacks
the
carbon-carbon double bond
€19,20,21,223 .
Only a few
of
attacked to produce
the
carbon-carbon
hydroperoxides.
double bonds need be
Once formed, even in
trace amounts, hydroperoxides can act as catalysts.
point,
it
should
be
mentioned
autoxidation may be due to
that
the
At this
initiation
of
metal catalysts since most fatty
esters contain metal 123,243 .
Hydroperoxides are
formed
carbon-carbon double bond.
bond, the electrons
by
attack at the
As the oxygen attacks the double
rearrange
double bond is shifted.
oxidative
in
such
a
manner that the
This shifting is often referred to
as conjugation [20,213 .
02
-CH2-CH=CH-
--- >-CH=CH-CH
OOH
Hydroperoxides may decompose
hydroperoxides
decompose,
by
the
several
following
mechanisms.
groups
Mhen
can
be
12
produced:
and
a
tertiary
radicals,
carbon-carbon
decompose to
form
hydroxyl radicals, carbonyls
cleavage.
free
radicals
The
in
hydroperoxides
may
the following manner
[19] .
R O O H --- > RO * + HO 0
ROO + HOOR
> R00---H 0 0 R --- > HOH + RO ° + R00 °
I
I
H
H
The mechanism of chain
investigators
C253
.
propagation is agreed upon among
The
oxidized or may combine
initiation
products
may
be
with another hydrocarbon to produce
the following reactions.
R° + © 2 --- y R00 0
R000 + R H ---> R00H + Ro
Termination reactions
and often yield polymers.
generally
consume
free radicals
Some of these reactions are shown
below C19,253 .
R00« + R00 ° --- > R00R + O2
R00 ° + *OH
> ROH + O2
Ro + R o ---> R-R
ROOo +
R o --- > ROOR
Free radicals may also attack carbon-carbon double bonds
and produce larger hydrocarbon free radicals [19,253 .
HH
R0 +
I I
—G=CH H
--- y
R
I 0
-CH-CH00R
I O
13
The following general
of
the
double
bond,
scheme describes oxidative attack
formation
of
hydroperoxides,
hydroperoxide decomposition, chain reactions and termination
reactions [19,24] .
-CH2-CH=CH-
2
-CH2 -CH-CH-
I
00'
-CH2-CH=CHY
-CH--CH-CH
2
+
I
CH-CH=CH-
OOH
-CH--CH=CH-
-CH0-CH0-CH- + -CH-CH=CH-
2
-CH=CH-CH- + H»
-CH-CH=CHI
00"
(I)
I
chain
reaction
(4)
I
chain
reaction
chain
reaction
14
Each of the
[19,253 .
radicals
may
react
with
a variety compounds
These are detailed below.
Radical (2) may react
with
a carbon-carbon double bond
and polymerize to form another radical.
This is the primary
polymerization pathway.
-CH=CH-CH- + -CH0-CH=CH----- > -CH0-CH-CH2
2
j
-CH=CH-CHRadical (4) has the capability to attack a carbon-carbon
double bond to
produce
polymer
products
and continue the
chain reaction.
-CH=CH-CH- + -CH0-CH=CH----- >
2
00»
Radical
(2)
may
also
-CH=CH-CHi
o
I
0
0 I
-CH2-CH-CH-
combine
with
Radical
(4) and
polymerize to a nonradical polymer.
> -CH=CH-CHI
0
-CH-CH=CH- + -CH=CH-CHI
OOo
I
I
0
-CH=CH-CHTwo
Radical
(2)'s
may
polymerize
polymer.
-CH-CH=CH- + -CH-CH=CH- .--- > -CH=CH-CHI
-CH=CH-CH-
to
a
nonradical
15
Two Radical (4)'s may polymerize.
-CH=CH-CH- + -CH=CH-CH- — — > -CH=CH-CH-
I
I
OO0
J
00 0
0
I
-CH=CH-CHAnother way to visualize Radical
(2) or (4) attacking a
carbon-carbon double bond is shown below [25] .
HH
R° +
I I
I I
HH
I I
I I
—C=C-
---- ) R-C-C °
R1R2
HH
R1R2
HH
I I
R00 ° +. -C-C-
> Higher polymers
I I
---- ) R00—C—C°---- > Higher polymers
I I
I I
r Ir2
R1R2
This predominant polymerization
pathway
to higher polymers
is known as vinyl polymerization.
Other Oxidation Reactions
Besides
oxidative
oxidation mechanisms.
polymerization,
there
exist
other
The double bonds in the sunflower oil
can also be homolytically
cleaved
cleaves the double
the
bond,
by
oxygen.
When oxygen
alkene molecule is converted
into two smaller molecules C26] .
The products of
cleavage
each
contain a carbonroxygen
double bond with the oxygens attached to the carbons present
in the original carbon-carbon double bond [273 .
16
-CH0-C=C-CH0
2 I I 2
-CH0-C=O
2
H H
alkene
O=C-CH0-
I
I
H
2
H
aldehydes
Aldehydes may also
carbon-carbon
+
double
be
formed
bond.
without
If
contains a terminal double bond,
a
cleavage of the
hydrocarbon
molecule
this bond may be attacked;
and an aldehyde may be formed C263 .
R-CH=CH2
Oxygen may open the
of the double bond) in
--- >
R-CH2-C-H
hydrocarbon chain (without cleavage
the following manner to form ketones
[26,273 .
°2
-- >
-C=C-CH0-C=C-
I I
H
H
Aldehydes can
ease.
They
are
2
H
-CH0-C-CH0-C=C-
I I
2
H
N
O
undergo
readily
further
converted
2
H
I I
H
oxidation
with extreme
to carboxylic acids by
copper and heat E263 .
RCHO
RCOOH
aldehyde
Another mechanism by
carboxylic acid
which
aldehydes
carboxylic acids and alcohols may be
are converted to
[26,273
17
O
OH
Il
0
I
H
R-C-H + R-CH=CH-CH----- > R-C-O-OH--- > R-C-OH + ROH
I
I
OOH
R
Aldehydes may undergo
autoxidation
to that of hydrocarbons.
in a manner similar
Hydroperoxide radicals act as the
chain carriers, and the products can be acids C283 .
RCHO +R'
---- > RCO
+R'H
RCO + © 2 --- > RCOg0
RCOg0+ RCHO --- > RCOgH + RCO
RCOgH + RCOgH --- > RCO2H + RCOgH + Og
The mechanism for these reactions may be as follows:
0
Il
0
.
R-C-H +
Il
R ' ---> R-C° + R'H
4 O2
0
0
0
0
Il
Il
H
Il
R-C-O-O 0 + R C H -- -> RC0 + RC-00H
f
.
-
0
0
0
0
Il
Il
Il
Il
RC-00H + RC-00H --- > RC-0H + RC-0H + 0g
A molecule
produced.
of
oxygen
Acids
are
is
regenerated,
generally
the
and
two acids are
terminal
oxidation
pathway products.
Oxidation
of
bonds, and from a
ketones
requires
thermodynamic
under severe conditions.
can take place, ketones
breaking carbon-carbon
viewpoint takes place only
If conditions exist where cleavage
are
cleaved
on either side of the
carbonyl group to yield a mixture of carboxylic acids E263 .
18
Both aldehydes and
ketones
contain the carbonyl group,
C=O and are referred to as carbonyl compounds.
group plays an important
role
The carbonyl
in determining the chemistry
of aldehydes and ketones.
The carbonyl
group
addition and increases
attached to
the
provides
the
alpha
a
site
acidity
carbon.
of
Both
for nucleophilic
the hydrogen atoms
these
effects are
consistent with the structure of the carbonyl group 1126,273.
The carbonyl group contains a carbon-oxygen double bond.
The
pi
carbonyl
electrons
carbon
pull
strongly
toward
electron-deficient
electron-rich.
Because the
susceptible to
unhindered
carbonyl
approach
Approach is perpendicular to the
the polarized carbonyl
group
oxygen
and
and make
carbonyl
group
from
oxygen
is flat, it is
above
or below.
plane of the group.
is
accessible,
Since
it is highly
reactive £26,273 .
Because aldehydes and ketones
group, they resemble
each
properties.
An aldehyde
attached to
the carbonyl
other
has
a
their
with
reactive
difficulty;
than
ketones
in most of their
carbon and a hydrogen atom
(b)
of ketones.
properties
aldehydes are easily oxidized,
only
closely
group while there are two carbons
attached to the carbonyl group
in structure affects
both contain the carbonyl
This difference
in
two
ways:
(a)
whereas ketones are oxidized
aldehydes
toward
are
usually
nucleophilic
more
addition.
19
Nucleophilic
addition
is
the
characteristic
reaction of
carbonyl compounds [26,27] .
Aldehydes, ketones
nature.
Once formed,
solvents
because
and
carboxylic
they
polar
may
acids
not be soluble in nonpolar
groups
tend
to
associate
themselves more readily than with the solvent.
hydrogen bond,
and
are polar in
precipitation
or
with
These groups
phase separation may
occur as they form in a system.
For example, carboxylic acid molecules are polar and can
form hydrogen bonds with
each
other.
Two carboxylic acids
can strongly hydrogen bond in the following manner:
C263
O-H-- 0
R-C
^
In
this
case,
the
together by two
hydrogen
in
/
0-- H-O
carboxylic
hydrogen
bond
C-R
bonds.
hydrocarbon
bonded, they are less
acid
likely
molecules
are
held
Carboxylic acids readily
solvents.
to
Once
hydrogen
react with other chemical
species in the system.
Carboxylic acids
were
given
their
name because their
most characteristic property is acidity. The hydroxyl group
of an acid can be replaced
by
NHR to yield amides.
Amides
are functional derivatives of acids and contain the carbonyl
group.
Amides
undergo
amines [26,27] .
hydrolysis
to
revert to acids and
20
Formation of amides involves
of the acid:
cleavage
of the C-OH bond
[26]
0
/
R-C
"
0
„ + BJH0R
/
R-C
— >
2
A'
OH
+ H2°
\
EHR
Anhydrides and esters are also functional derivatives of
carboxylic acids.
The hydroxyl group is replaced by OOCR or
OR' respectively [26,273 .
0
//
R-C
\
0
OR'
/
R-C
0
anhydride
ester
The presence of the carbonyl group makes these latter groups
polar [273 .
As the number of
increases, they may
phase.
If the acidic
carbonyl
species in a nonpolar system
precipitate
species
chain amine, the resulting
or
could be reacted with a long
amide
should
remain soluble in a nonpolar system.
became
long
chain
amides,
form a separate liquid
oxidized
be less polar and
If enough of the acids
triglycerides
might
remain in solution in a nonpolar oil solvent.
Amides can be derived from acids, esters and anhydrides.
Some examples are shown on the following page [273 .
21
H
\\
R-C-OH + RWH2 — > R-C-NHR + H3O
R-C-OR' + RNH2 --- > R-C-NHR + R'OH
0
0
0
0
Il
Il
Il
Il
R-C-0—C—R + RNH3 — — > R-C-NHR + HO-C-R
Sunflower
weights.
oil
consists
Making one portion
polar may not make the
of
a
variety
of
molecular
of an extremely large molecule
entire molecule polar enough to form
a separate phase in a
nonpolar
Hii^ht have to contain
a
number
oil solvent.
of
Each molecule
polar groups before it
separates.
The resulting phase may be a solid (or gel) with
entrapment
of
solvent.
other
molecules
such
as
a
nonpolar
oil
22
EXPERIMENTAL
Equipment
The environment of the crankcase
of a diesel engine was
simulated in the laboratory in the form of a reaction kettle
placed in an
oil
bath
conducted in a pair of
heater.
500
All the experiments were
ml reaction kettles fitted with
four post entrance lids (Figure I).
Two of the openings
(the
kettle lid were fitted with
center
and one side) on each
Ace threads to provide airtight
seals for entering and exiting gas tubes, respectively.
tight
seals
environment
were
as
necessary
well
as
to
to
provide
measure
the
a
Air
controllable
gas
flow rate.
Silicon grease insured gas-tight seals between each entrance
and its ground glass
stopper.
As
entering gas tube was connected
to
shown in Figure I, the
a 30 mm glass frit that
provided gas percolation through the oil mixture.
frit was
accurately
positioned
consistent positioning from
in
a
experiment
percolation could be observed
The exiting tube was connected
by
The glass
fixed location with
to experiment.
Gas
removing a glass stopper.
to tygon tubing leading to a
soap film flow meter which measured
normally adjusted to 2.0 ml/sec.
the gas flow.
Flow was
23
Copper
Copper foil
cylinder and
was
used
with
as
an
placed
copper foil was 5 cm
a
catalyst
area
of
over
the
long,
2
cylinder,
20
cm
gas
cm
all experiments.
was
rolled into a
dispersion
tube.
The
wide and 0.125 mm thick.
When forming
the
overlapping.
When resting on the fritted glass surface, the
copper was in intimate
the
in
O
contact
ends
with
were
both
touching, not
the gas and oil
(Figure I).
The
reaction
(Figure 2).
The
kettle(s)
oil
was
bath
placed
contained
in
the
oil bath
paraffin oil which
reached a higher level on the reaction kettle than the level
of the oil mixture within
the
kettle.
A Polyscience Model
73 immersion circulator was utilized to maintain an oil bath
temperature of slightly above 150
C.
The Polyscience Model
73 has automatic temperature control with a precision of 0.2
C and circulates approximately 13
minute.
liters of heating oil per
The automatic temperature control was adjusted to a
setting where the oil mixture
bath) was maintained at 150
within the kettle (not in the
C.
The temperature within the
reaction kettle was checked periodically with a thermometer.
The oil bath was well
insulated with approximately 2 inches
of vermiculite insulation
between
the sides and bottom of the bath.
covered the vapor space
above
the
steel plates that formed
A tight fitting steel lid
kettle(s) and bath oil.
The oil bath was placed under a venting hood.
24
Gas Dispersion Tube
Cos Exit Tube
Gloss Stopper
Ace Thread
Reaction K ettle L id
Reaction K e ttle
Copper F o il —
F ritted Disc
Cos Dispersion
H ead
Figure I.
A. immersion
Circulator
Reaction Kettle
D. Thermometer
B.Insulated
Gas Line
Line to
Preheating Coil
C. Gas
E . Reaction
K ettle
F. Gas Preheating
C oil
Figure 2.
Oil Bath and Reaction Kettle
25
High pressure
nitrogen) to the
exited from the
cylinders
reaction
gas
supplied
kettle.
cylinder
position gas header mounted
the oil bath.
the
(oxygen or
Stainless steel tubing
regulator
on
gas
to
enter a four-
a steel frame placed beside
Two precision needle valves were connected to
the headers via
tygon
reaction kettles.
tubing
to
control
gas flow to the
Gas was preheated by passing it through a
one-quarter inch coil
of
immersed in the oil bath.
stainless
steel
tubing that was
Insulated teflon tubing connected
the preheating coil to the glass stem of the frit.
was adjusted to 2.0 ml/ sec.
Gas flow
An operating diagram is shown
in Figure 3.
Viscosity of the
using
calibrated
viscometers
oil
mixture was periodically measured
Cannon-Fenske
were
used
for
viscometers.
specific
viscosity
Specific
ranges.
Viscometers were placed in a constant temperature water bath
that was
maintained
immersion
at
circulator.
40
C
by
The
a
Polyscience Model 73
Polyscience
Model
73
automatic temperature control with a precision of 0.2 C.
take a sample,
one
of
removed and 8 ml of
reactor.
The
viscometer.
8
In
the
ml
an
the
oil
sample
attempt
kettle
was
of
the
lid
on
lid glass stoppers was
then
to
placed
steel
To
mixture was pipetted from the
procedure, the pipet was
the
has
transferred to the
standardize the pipetting
8 inches below the surface
oil
bath.
Two
viscosity
26
A. Immersion Circulator
B. Insulated Gas Line
C. Gas Line to Heating Coil
D. Thermometer
E. Tygon Tubing
Fi Soap Film Flow Meter
G. Oil Both
H. Gas Lines
I . Stainless Steel Tubing
J. Needle Valve
K. Tygon Tubing
L. Four Position Header
M. Stainless Steel Tubing
N. Shut~off Valve
O. Pressure Regulator
P. Gas Cylinder
Figure 3.
Oil
Bath and
Operating
Diagram
27
measurements were
taken
on
each
sample,
and the average
value was used as the data
point.
If necessary, the sample
was then saved for a Total
Base Number (TBN) titration.
If
not, it was returned to the reaction kettle.
In an attempt
care was
taken
particles
in
measurement.
to
to
quantify
note
the
when
the
rate of gel formation,
gel
was
viscometer
Once gel
was
frit and copper foil were
the copper surface
taking
a
viscosity
observed in the viscometer, the
checked
for gel formation.
the frit were
covered with gel, the experiment was stopped.
At the end of
the
dispersing
gel
was
area
When
of
each experiment,
and
when
first observed as
allowed
to
drain on paper
towels in an attempt to remove as much of the supernatant as
possible.
The
gel
was
approximate comparison of
then
gel
weighed.
This
allowed
formation from experiment to
experiment.
Total Base Number (TBN)
oil mixture.
2896,
TBN values were determined according to ASTM D
"Total
Base
Number
Potentiometric Perchloric
method suggested
sharp end
indicates the alkalinity of the
points
using
when
of
Acid
the
Petroleum
Titration".
back
working
titration
with
used
Products
This standard
method to get
oils.
Excess
standard HClO^ solution was added to a prepared sample.
excess was then back
solution.
An
Orion
titrated
by
The
with standard sodium acetate
Research
Model
901
Microprocessor
28
Ionalyzer fitted with a
Corning sleeve-type saturated glass
electrode was used to detect the endpoints C293 .
The iodine value of the
sunflower oil was determined to
provide a relative indication
present.
This
value
was
Standard 1959-69 which is
of the amount of unsaturation
determined
applicable
according
to ASTH
to vegetable oils and
their fatty acids [30] .
Fourier transform infrared
spectroscopy (FTIR) was used
to determine relative amounts of carbonyl groups in selected
samples.
FTIR
is
digitization of spectra
information in a
matter
a
low-cost,
computer-controlled
enables
the user to extract
which
of . seconds.
spectroscopic instruments, shelves
on chart paper.
[313 .
A
With older infrared
of spectra were recorded
FTIR spectra is stored in the computer
Nicolet 5DX computer system with a helium-neon
laser and a sodium chloride sample chamber was employed.
Materials
The vegetable oil used at
was sunflower mill
oil
the beginning of the research
from
Continental
Grain Company of
Culbertson, Montana.
It had an
in the research, the
Culbertson oil supply became depleted.
New
sunflower
mill
oil
was
iodine value of 140.
obtained
Incorporated in Fargo, North Dakota.
from
Early
Cargill
It had an iodine value
29
of 144.
by
Hydrocarbon
Phillips
basestock lubricating oil was provided
Petroleum
in
Bartlesville,
Oklahoma.
Two
different batches of basestock oil were obtained.
Lupersol
130
Corporation
of
was
provided
Buffalo,
New
dithiocarbamate (ZDTC)
Vanlube AZ, by R.
Connecticut.
Chemical
T.
Co.
obtained
supplied
Zinc
the
dialkyl
tradename,
under
dialkyl
the tradename,
Company, Inc. of Norwalk,
from
Lubrizol
Bartlesville,
Exxon
Chemical Products, Inc.
stearate.
Pennwalt
Zinc
dithiophosphate
TBHQ
of
of
Food-Grade
Stamford,
(ZDTP)
was
1395, by Phillips
Oklahoma.
Chemicals
Tertiary butyIhydroquinone (TBHQ)
Company
York.
Vanderbilt
Petroleum Company of
name Tenox
Lucidol
Octadecylamine (ODA) was obtained from Aldrich
supplied under
was
was
by
of
Paranox 107
Houston,
Texas.
was received from Eastman
Kingsport,
Tennessee under the
Antioxidant.
Connecticut
Sattva Chemical
provided
the
copper
All other chemicals were reagent grade.
From the standpoint
of
safety, inherent problems arise
when working with hot oils.
The experimental apparatus was
placed under a venting hood
to remove noxious vapors.
Care
was taken when working with the high pressure gas cylinders.
All waste oils and cleaning agents were treated as hazardous
wastes and disposed
Chemical and
of
Hazardous
through
Waste
Montana State University's
Department.
Gloves, safety
glasses and aprons were worn when handling hot oils.
I
RESULTS AMD DISCUSSION
Any
given
commercial
additive package
that
particular oil.
lubricating
is
specifically
contains, an
designed
for that
The oil additive package contains chemicals
that maximize engine performance.
are
oil
dispersants,
detergents,
Some of these chemicals
rust
inhibitors,
oxidation
inhibitors, viscosity modifiers and friction reducers.
the presence
of
such
a
wide
variety
is
extremely
additive package chemistry
of
With
chemicals, the
complex.
Each of
these additives are chemicals and may react with one another
to form new compounds when placed in the lubricating oil.
The
showed
previous
workers
commercial
CIO,11,121
lubricating
weight percent sunflower oil
was
quantified
by
polymerization of
oil
oil
this
mixture
laboratory
contaminated
degraded rapidly.
viscosity
the
in
with 5.0
Degradation
measurements,
measured
with
by viscosity
rise.
To begin the current
with
Super
commercial
HD
II
research,
low
lubricating
percent sunflower oil.
ash
oil
a standard run was made
MIL-L-2104C
contaminated
API
with
CD
SAE
30
5.0 weight
Standard conditions consisted of the
oil mixture being exposed to 20 cm2 copper foil, 150 C and 2
31
ml/sec oxygen percolation.
rise for this case.
Figure 4 presents the viscosity
Sunflower oil polymerization was rapid,
and simulated engine failure occurred within 20 hours.
commercial lubricating oil
experiment,
In a
failure takes place
when the viscosity of the oil mixture reaches or exceeds 500
centistokes.
Due to
This result was consistent with previous work.
the
unknown
chemical
package in commercial oil,
Dutta
30 hydrocarbon basestock oil
basestock oil
as
the
nature
in
his
lubricating
sunflower
oil
standard oxygen flow, copper
severe degradation of
the
The
viscosity
When using
basestock oil with 5.0
exposed
foil
and
the mixture to
150 C.
He observed
mixture within twenty hours
and noted formation of insolubles
heavy sludge.
research.
substrate, new problems
and
oil
the additive
[12] attempted to use SAE
were encountered. Dutta contaminated
weight percent
of
which he referred to as a
data
of
Dutta,
shown as a
dotted line in Figure 5,
are viscosities taken of the clear
liquid above the sludge.
This clear liquid will be referred
to as the supernatant phase.
To familiarize the
basestock oil,
verified.
Dutta's
current
standard
Hydrocarbon
basestock
weight percent sunflower
oxygen percolation
supernatant
can
be
and
researcher
oil
150
seen
was
C.
in
with the use of
conditions experiment was
contaminated
exposed
Viscosity
Figure
5.
with
5.0
to copper foil,
data of the
Formation
of
32
H-
300
(A 200
TIME, hrs
Figure 4.
Viscosity of commercial oil and 5.0%
sunflower oil vs. time for standard
conditions.
33
Dutta: o
this work:a
TIME, hrs
Figure 5.
Viscosity of basestock oil and 5.0%
sunflower oil vs. time for standard
conditions.
34
appreciable
insolubles
occurred within 8
to
(referred
10
to
hours.
as
gel
or
sludge)
Both
gel and viscosity
results are consistent with Dutta's work.
The gel blanketed
the copper foil and was 1/8
to
1/4 inch thick on the sides
and bottom of the reaction vessel.
The gel was sticky with
an
with
irregular
surface
structure
nipples
of
gel
protruding into the supernatant.
The proportion of
nine to one.
This
supernatant
result
current researcher felt
contributing to gel
was
that
to gel was approximately
not unexpected because the
polymerized sunflower oil was
formation.
The original concentration
of sunflower oil in the
latter experiment was nine and one-
half to one.
is highly swollen with supernatant.
This may
The
account
gel
for
the
small
difference
in the above
proportions.
The gel was
were made
to
difficult
to
gravimetricalIy
methods of quantification were
phase was swollen with
quantify.
measure
Initial attempts
the
gel, but these
not reliable because the gel
lubricating
oil.
Since two phases
are formed when using hydrocarbon basestock oil, the current
researcher
felt
little regarding
the
viscosity
degradation.
measurements
It
was
alone
meant
also difficult to
avoid gel particles when pipetting the supernatant, and even
fine gel particles
lodged
viscosity measurements.
in
the viscometer and distorted
35
When
using
the
commercial
lube
oil,
sunflower
oil
polymerization resulted in viscosity rise with only a slight
amount of gel at the end of
research
in
the
a
basestock
run CIO,113 .
oil
indicated
polymerization contributed to gel
must be eliminated
going to be
from
measured
unquantifiable
and
the
by
C123
sunflower
oil
formation.
system
viscosity
adds
Dutta's
The gel phase
if oil degradation is
rise.
another
Gel is largely
complication
to
the
research.
Preliminary experiments (Runs I
attempt to find
milder
conditions
polymers might contribute to
formation.
have a
Since
lower
the
and
2) were done in an
where the sunflower oil
viscosity rise rather than gel
addition polymerization reaction may
activation
energy
than
the other oxidation
reactions that form aldehydes, ketones and acids, decreasing
the temperature might
produce
reduced gel C273
From
.
conditions of 5.0 weight
Dutta's
percent
ml/sec oxygen flow and presence
within
10
hours;
and
and
get
In
A control
of
standard
of copper foil produced gel
viscosity
rose
only
order to avoid gel at a lower
measurable
reasonable time frame, more
C123 results, standard
sunflower oil, 150 C, 2.0
supernatant
slightly within 60 hours.
temperature
increased viscosity rise and
viscosity
rise
in
a
sunflower oil may be necessary.
conditions
(basestock
oil, oxygen
percolation, copper foil and 150 C) with 25.0 weight percent
36
sunflower oil (Run I) was run to establish a baseline for an
increased sunflower oil trial at reduced temperature.
Viscosity
showed
measurements
little
of
viscosity
rise
conditions) as shown in Figure
observed at 8 hours.
55 hours, the
Upon
mixture
was
allowed to cool. Twenty
formation
with
observed.
As
ridges
the
(similar
6.
conditions
sunflower oil.
to
Run I
standard
completion of the experiment at
removed
to
from
the oil bath and
thirty hours later, uneven gel
protruding
anticipated,
in
Gel formation was first
more
above
gel
experiment with 25.0 weight percent
standard
supernatant
experiment
the
surface was
was observed in this
sunflower oil than in a
with
5.0
weight
percent
This was consistent with prevailing thinking
that sunflower oil polymerizes and forms gel simultaneously.
Increasing
the
amount
of
sunflower
oil
in
the
system
resulted in an increase in the amount of gel formed.
Lowering the temperature from 150
allow
proceed
oxidation
as
to
rapidly
polymerization.
This
the
species
as
those
might
to 135 C might not
contributing
contributing
limit
experiment (Run 2) was conducted
C
at
gel
to
to
gel to
addition
formation.
This
135 C with 25.0 weight
percent sunflower oil, basestock oil, copper foil and oxygen
percolation and gave viscosity
the higher temperature was used.
in Figure 6.
Significant
rise
similar to Run I where
This result is also shown
gel formation was observed in the
37
I—
300
CO 200
135 C
TIME, hrs
Figure 6.
Viscosity of basestock oil and 25.0%
sunflower oil vs. time for standard
conditions at 150 C and 135 C .
38
same time frame as
Run
I
(8
slightly more gel seemed to
to 10 hours).
be
present
than in Run I at 150 C.
It
was
alleviate
that
problems
Future work would
Run 2 at 135 C
1
concluded
the
in
Unexpectedly,
lower
associated
emphasize
temperature
with
gel
did
not
formation.
understanding the formation of
the gel phase and its relation to viscosity rise.
When working with
distinguish
and
polymers,
classify
the
it
gel and the species contributing
be
slight,
might show them to
be
physically
and vigorous homogenization
gel could be physically dispersed
to gel
and
viscosity
between solids
The difference between the
to viscosity might
viscosity rise, one might
often difficult to
differences
(especially gels) and liquids.
species that contribute to
is
similar species.
If the
and made to contribute to
conclude the species contributing
rise
are
physically and chemically
similar.
A standard
basestock
oil/sunflower
oil experiment was
run for twenty hours.
The viscosity was measured and found
to be 103 centistokes.
Gel from this experiment was swollen
with the supernatant.
Most
of the supernatant was removed
from the gel by draining and "patting" dry with an absorbent
cloth.
The "dried" gel was then weighed.
The proportion of
gel to supernatant was determined to be approximately ten to
ninety.
The
gel,
in
proper
proportions,
was
then
39
homogenized at room
temperature
Homogenization of the
in
the supernatant phase.
gel/supernatant
mixture
in a Waring
high speed laboratory blender (Model 700B) at 20,000 rpra for
five minutes produced a
H O centistokes.
viscosity
produced
was
At
slight
twenty
124
viscosity
minutes of homogenization, the
centistokes.
increased
rise from 103 to
viscosity
homogenization, the viscosity
Further
rise.
At
appeared
homogenization
75
minutes
to decrease.
of
These
results can be seen in Figure 7.
The reliability of the
7 was questionned.
mixture through
detected.
Upon
the
Once
measurements
closely
gel
riot
the
be
mixture
distinct
The
particles
gel
supernatant phase.
fine
particles
lighting where a
fine
examining the flow of the
viscometer,
may
homogenization,
viscosity measurements in Figure
are
gel
particles were
detected,
reliable.
was
two
Upon
inspected
phase
ceasing
under
bright
system was observed.
gradually
Homogenization
viscosity
had
settled
from
the
produced a finely
dispersed two phase system
where the above viscosities have
little meaning.
particles probably distorted the
Fine
gel
viscosity measurements when
passing through the viscometer.
This led to the
that physical agitation did not
cause the
gel
conclusion
to
revert,
phase, viscous material.
even
temporarily,
to a single
40
TIME, min
Figure 7.
Viscosity of supernatant/gel mixture vs.
time of homogenization.
41
Dispersants
additive
and
package
surfactants
are
in
intended
to
substances such as dirt/grime, but
preventing the gel phase from
the
commercial
disperse
they
oil
inorganic
may be a factor in
precipitating
C323 .
It was
believed the commercial oil dispersant might be able to keep
the gel in solution
rise.
Some of the
and
thus
"dried"
gel from a basestock experiment
was heated in the commercial
ten to ninety
for
24
checked every four
nature of the gel
the lube
oil
yield a meaningful viscosity
lubricating
hours
at
hours.
were
did
not
150
No
oil at a ratio of
C.
The mixture was
physical
changes in the
observed.
The additive package in
appear
change
to
the physical or
chemical nature of the gel.
The amount of gel present did
not appear to
the
decrease,
and
gel particles remained a
separate phase from the commercial oil.
This result was not
surprising because the
dispersants are not
commercial
oil
designed to disperse organic species.
the gel-forming species
were
This result indicated
chemically different from the
species contributing to viscosity.
A
commercial
obtained from Exxon
organic
dispersant,
Chemicals
of
Paranox
Houston, Texas.
107 is a suecinamide-based, ashless dispersant.
that this strong dispersant might
forming species dispersed
as
be
they
dispersant in the commercial oil
107,
may
was
Paranox
It was felt
able to keep the gel­
are
being formed.
The
not have been able to
42
disperse the gel that
was
already highly associated in the
homogenized gel/commercial oil experiment.
The amide part
of the dispersant is basic in nature and might interact with
acidic species to keep
them
reasons, the exact chemical
not be obtained.
Paranox 107 may
A
be
dispersed.
structure
Due to proprietary
of Paranox 107 could
Phillips 66 representative speculated
capable
species in the present
of dispersing organic chemical
system
and recommended using 1.0 to
5.2 weight percent [32] .
Paranox 107
was
added
at
basestock oil/sunflower oil
3.0
mixture
percolation, copper foil and 150
was observed by 14
hours.
the
gel
again
made
have
formation
worked
still
for
a
Gel formation
to 10 hours.
meaningful
Presence
supernatant
The commercial dispersant
short
viscosity-forming species and
exposed to oxygen
that gel appeared in a
8
taking
occurred.
percent to the
(Run 3).
Recall
viscosity measurements difficult.
may
and
C
standard conditions experiment in
of
weight
period
This
the
of
again
time, but gel
indicated
the
gel-forming species were
chemically different.
The homogenization trials
experiment indicated
and the commercial dispersant
formation
simply a physical separation.
the commercial oil might
act
of
the
gel
phase was not
It was felt the additives in
in
chemically inhibit gel formation or
either
2)
of
two ways:
I)
keep gel suspended.
43
If the gel were merely suspended, it might be chemically the
same as viscosity-forming species.
Dutta
[123
found
two
little gel was produced
experimental
where
in experiments using basestock oil.
Both these experiments were
present work because they
future research.
situations
These
reproduced and confirmed in the
were
pivotal to the direction of
experiments
are
discussed in the
following paragraphs.
An antioxidant, zinc
dialkyl dithiocarbamate (ZDTC), is
sometimes used in commercial
oil
and
ZDTC
an
anti-wear
agent.
as an oxidation inhibitor
is
corrosion by inhibiting oxidation of
species as well as
surfaces.
understood,
The
by
forming
mechanism
but
by
ZDTC
a
weight
inhibit
protective film on metal
ZDTC
believed
hydroperoxides. The recommended usage
lubricating system is 1.0
to
the lube oil to acidic
which
is
thought
acts is not well
to
decompose
level in a diesel oil
percent. The structure of
ZDTC is given below.
c ShII
[j
N-C-S
C5H11
Zn
2
In Run 4 where 1.0 weight percent ZDTC was added at time
zero to the 5.0
weight
oil and exposed to
percent sunflower oil and basestock
copper foil, standard oxygen percolation
44
and 150
C,
no
viscosity
rise
and
no
significance were observed (Figure 8).
block
polymerization
formation.
This
and
work
hence
gel
The ZDTC appeared to
viscosity
confirms
formation of
the
rise
and
gel
finding of Dutta C123
with ZDTC.
Rewolinski
[10]
showed
commercial oil proceeded by
oxidative
a
free
Rewolinski's work, a commercial
periodically added
polymerization
of
radical mechanism.
In
peroxide (Lupersol 130) was
to
the
commercial
oil
nitrogen
environment,
and
viscosity
rise
produced
with
oxygen
percolation
was
system under a
matching
observed.
that
The
structure of Lupersol 130 is shown below.
CHL
CH-,
3
3
I
I
(CH3)3C00-C-C=C-C-00C(CH3 )3
I
I
CH3
Lupersol 130 is known
to
to produce free radicals
CH3
homolytically cleave at 0-0 bonds
which
of oxidative polymerization.
catalyze the chain reaction
It
is
widely used as a free
radical initiator in vinyl polymerization.
Dutta [123 used Lupersol
oil in the
nitrogen
basestock
environment,
Lupersol 130 was
hours.
oil
He
added
observed
130 with 5.0 percent sunflower
and
copper
at
0.5
subjected
catalyst
weight
significant
the mixture to a
and
150
C.
The
percent every four
viscosity
rise
with
45
600
DuttaJ o
this work:a
500-
CO
400-
O
>
I—
30 0
CO
O
U
#
200
100
-
20
30
40
50
TIME, hrs
Figure 8.
Viscosity of basestock oil and 5.0%
sunflower oil vs. time for standard
conditions with initial addition of
1.0% ZDTC.
60
46
negligible gel formation.
as that encountered
Viscosity
when
using
rise was not as rapid
commercial oil and oxygen,
but still occurred;
In this work, the
free
radical initiator at 0.5 weight
percent was added every four hours to the 5.0 weight percent
sunflower
oil/basestock
oil
mixture
nitrogen percolation, copper
This trial
resulted
in
catalyst
and
essentially
viscosity rise presented in
Figure
with
no
9.
150
gel
exposure
to
C (Run 5).
and gave the
Dutta's results are
shown on Figure 9 as a dotted line.
Zinc dialkyl
rise and gel
while the
dithiocarbamate,
formation
Lupersol
without oxygen.
No
oxygen atmosphere.
that
gel
is
a
in
the
ZDTC, blocked viscosity
basestock
130
experiment
gel
was
These
separate
formed
yielded viscosity rise
in
experiments
chemical
oil with oxygen
the absence of an
seemed to indicate
species
produced
by
oxidation.
If
oxygen produced gel
and
Lupersol 130 without oxygen
yielded only viscosity rise, what
used simultaneously?
would happen if they were
One might expect to get both viscosity
rise and gel formation.
make species which could
Lupersol
130 free radicals might
contribute to viscosity rise while
simultaneous oxidation might result in gel formation.
47
Dutt a: a
this work:o
TIME, hrs
Figure 9.
Viscosity of basestock oil and 5.0%
sunflower oil vs. time for standard
conditions with nitrogen and additions of
0.5% Lupersol 130 every 4 hours.
48
An experiment with 0.5 weight percent Lupersol 130 added
every four hours at
standard conditions with oxygen present
(Run 6) produced significant gel
standard conditions without
Only one viscosity
in a time frame similar to
Lupersol
measurement
was
130
(8 to 10 hours).
taken
because the gel
formation by 8 hours was so great that gel particles plugged
the viscometer. It was
because
the
speculated that this result occurred
triglycerides
oxidized with the
simultaneously
resulting
polymerized and
polymers perhaps precipitating
due to their "carbonyl polarity."
Previous
results
indicated
inhibiting polymerization that
gel formation.
ZDTC
was
produced
an
antioxidant
viscosity rise and
What if more free radicals as derivatives of
Lupersol 130 were added to the earlier experiment using 1.0%
ZDTC with oxygen in
basestock oil?
viscosity rise
an
Dutta
when
exposure
[12]
used
Lupersol 130 present.
of
5.0% sunflower oil in
showed
with
Current
a
ZDTC
does not prevent
nitrogen atmosphere and
and past research indicated
ZDTC does not inhibit the initiator role of Lupersol 130 but
does seem to work as
an
Lupersol 130
were
system,
results
might
be
viscosity
rise
because
the
significant
all
antioxidant.
present
oxidation of polymers to gel
produce viscosity rise.
If ZDTC, .oxygen and
in the sunflower/basestock
and
no
gel
ZDTC
formation
may
block
but
the
yet allow Lupersol 130 to
49
An experiment using
oil, 5.0 weight
1.0
percent
weight percent ZDTC, basestock
sunflower
oil, 0.5 weight percent
Lupersol 130 added every four hours, copper foil, oxygen and
150 C (Run 7)
produced
again difficult to
heavy
take
such large quantities
gel.
because
by
8
Viscosity results were
gel formation occurred in
hours
that
the gel particles
plugged the viscometer.
The
experiments
with
ZDTC
and/or
Lupersol
130
are
summarized below.
Table 2;
Summary of Experiments with Additives
ZDTC
ZDTC
ZDTC
Lupersol 130
oxygen
Lupersol 130
Lupersol 130
oxygen
nitrogen
oxygen
no viscosity
viscosity
gel
little gel
little gel
gel
A strong hypothesis may be made regarding Table 2.
dialkyl dithiocarbamate
(ZDTC)
polymerization, but hot
the
appears
to block oxidative
oxidation reactions that cause
polymerized material to become increasingly polar.
may
be
blocking
Zinc
hydroperoxide
formation,
The ZDTC
but
not
50
hydroperoxide decomposition.
When
Lupersol 130 is present,
hydroperoxides are already present
and decomposition of the
hydroperoxides occurs.
In the ZDTC, Lupersol 130 experiment
with nitrogen present, the
rise.
In
the
same
polymers contribute to viscosity
experiment
polymers may undergo oxidation
with
and
oxygen present, the
form
gel.
ZDTC cannot
stop the oxidation reactions that cause polarity.
is present,
occurs,
the
polymer
is
polymer
still
being
becomes
oxidation reactions appear to
made.
increasingly
be
fast
When ZDTC
As oxidation
polar.
The
enough to knock the
polymer out of solution to form gel before it contributes to
viscosity rise.
When
Lupersol
simultaneously, oxidation
of
130
the
and
polymers
oxygen are used
results in gel
formation.
These foregoing results
research approach.
It
led
was
to
a re-evaluation of the
decided
that pursuing a system
with two additives, Lupersol
best interest of
future
130
and
research.
ZDTC, was not in the
Dealing with a simple
chemical system might be the best approach.
At this point in
the
research,
hydrocarbon basestock oil was
the original supply of
exhausted.
The new basestock
oil was obtained from Phillips 66 and is known as Baltic Oil
ISO UG 68, Grade
supplies
of
315,
commercial
81550.
Jette C113 found different
lubricating
oil
had
different
51
additive packages which
their source of
crude
different sources
were
oil.
with
formulated
These
their
for
and based on
lubricating oils from
customized additive packages
gave viscosity rise in different time frames.
Due to Jette's Clll findings,
exposed
to
standard
conditions
sunflower oil, oxygen
(Run 8).
the new basestock oil was
with
percolation,
Gel was still
produced
5.0
copper
weight percent
foil
and 150 C
but at a later time.
The
old basestock oil and standard
conditions produced gel in 8
to 10
basestock
hours,
while
the
conditions did not produce
hours.
Supernatant
significant
viscosity
oil was similar to
basestock oil.
new
supernatant
and the new basestock oil
spectroscopy
between these oils.
sent to
gel
and standard
until 15 to 20
with the new basestock
viscosity rise with the old
This comparison is shown in Figure 10.
Why did the old basestock
emission
rise
oil
Lubricon
at
would
Samples
Laboratory
oil
produce gel at ten hours
twenty hours?
identify
of
in
a
Perhaps atomic
key
difference
both basestock oils were
Indianapolis, Indiana and
analyzed for trace metal content.
Atomic emission data indicated differences in trace metals
as shown in Table 3.
52
Table 3:
Atomic Emission Data
New Basestock
Old Basestock
ppm
ppm
Iron
I
I
Aluminum
I
I
Copper
I
2
Tin
3
O
Silicon
5
5
Sodium
2
8
Maanesium
3
6
Zinc
3
9
Barium
O
4
It is difficult to
pinpoint any significant differences
that may be causing the
basestock oil.
a
catalyst
of
gel formation in the new
Some possibilities
are tin may be acting as
while
delay
other
metals
magnesium and barium may
be
tests from Lubricon
not
parts per million;
are
therefore,
such
sodium,
zinc,
acting as deactivators.
These
extremely
the
as
accurate below ten
small parts per million
numbers shown in Table 3 may not really indicate significant
differences between the two basestock oils.
On visual inspection, the
be more iridescent
speculated
than
aromatics
the
were
new basestock oil appeared to
old
basestock
involved
and
oil.
It was
might
somehow
53
Basestock
H
I 1^
300
C/) 200
TIM E, hrs
Figure 10.
V i s c o s i t y c o m p a r i s o n of o l d a n d n e w
b a s e s t o c k o i l s a n d 5.0 % s u n f l o w e r o i l
vs. t i m e f o r s t a n d a r d c o n d i t i o n s .
54
influence
the
sunflower
oil
polymerization
C333
spectroscopy (IR)
or
.
on
other
Fourier
the
two
differences in aromatics.
oxidation
transform
samples
E343
were
of the
infrared
analyzed for
The aromatic region of IR
showed no distinct differences between the two basestocks.
Differences between the two
definable using
data.
Up
infrared
to
this
researcher assumed
spectroscopy
point
the
basestocks were not clearly
in
the
basestock
or
atomic emission
research,
oil
the current
acted
as
an inert
diluent for
the
sunflower
oil
because
Dutta [123 showed
that,
the
basestock
oil
without
the
when
sunflower oil
was
viscosity did not
exposed
rise
to
and
oxygen,
no
presence
copper
gel formed.
of
and 150 C,
Surprisingly,
different batches of hydrocarbon basestock appeared to cause
sunflower
oil
to
react
differently
at
the experimental
conditions of this work.
In attempting to
different
answer
basestocks,
there
explanations. Perhaps a
refinery chemical
sources
of
crude
was
may
the
question
are
a
trace
agent
causing
the
contain
of influence of
variety
of
possible
such as a homogeneous
difference.
varying
Different
amounts
elements due to geographical differences C333 .
of trace
These trace
elements may not have been detectable in the atomic emission
study because variations in results occur when attempting to
detect metals in amounts
of
10
parts per million or less.
• V
V-.
55
Further
speculation
basestocks was
as
judged
to
to
differences
be
between
unproductive
the
two
to the current
work.
Because the
new
basestock
from those found previously,
to be extended would need
ZDTC
and
Lupersol
130
oil
any
to
gave different results
earlier research that was
be reproduced.
experiments
discussed were reproduced
(Runs
4-7)
that
The series of
were previously
in the new basestock
oil.
The first experiment with
the new hydrocarbon basestock
oil (Run 9) involved
5.0
weight percent sunflower oil, 1.0
weight percent
and
0.5
ZDTC
added every four hours.
copper foil and 150
weight
percent Lupersol 130
The mixture was exposed to oxygen,
C.
It
was hypothesized that because
.
sunflower oil in the
new
basestock
later time than in the old
This was
of
believed
the polymers which contribute
because
slower oxidation of polymers which
new basestock-'(gel at
20
with the old basestock).
make polymers
produced gel at a
basestock oil that ZDTC might be
able to prevent oxidation
to gel.
oil
hours
there appeared to be a
contribute to gel in the
instead
of at 10 hours as
Lupersol 130 might then be able to
contributing
to
viscosity
rise because the
polymers might not be oxidized
and become polar.
experiment was performed,
gel
formation occurred within 20
that
ZDTC did not act against the
hours.
It was concluded
When this
56
oxidation reactions that result in polar polymers.
did not perform any better
in
basestock oil than it
with
did
The ZDTC
this experiment with the new
the same conditions in the
old basestock oil.
Next, the
every
four
new
basestock
hours
conditions of
at
oxygen
0.5
oil
with
weight
percolation,
without ZDTC present (Run 10)
Lupersol 130 added
percent
copper
were tested.
also produced heavy gel in the
and
foil
standard
and 150 C
This experiment
twenty hour time frame.
The
polymers were again being oxidized and becoming polar.
The two key experiments that
old basestock oil were
The first experiment
added every four
weight percent
repeated
used
hours,
was
oil.
observed,
Comparison of the
Figure
11.
5.0
copper,
This
in
and
rises
weight
exposed to Lupersol 130 and
C and 5.0
experiment (Run 11)
no
gel
formation took
results can be seen in
of
the
percent
nitrogen
nitrogen environment, viscosity
150
the old basestock oil.
viscosity
Viscosity
with
the new basestock oil.
weight percent Lupersol 130
obtained
place.
contaminated
0.5
in
nitrogen,
sunflower
reproduced the results
Viscosity rise
did not produce gel in the
rise
two
basestocks
sunflower
oil and
are similar.
Under a
occurred
in the same
time frame.
The second experiment that
produced
no
gel in the old
basestock was 1.0 weight percent ZDTC, oxygen, copper, 150 C
57
Basestock I*.A
Basestock
2 :o
T IM E , hrs
Figure 11.
Viscosity comparison of old and new
basestock oils and 5.0% sunflower oil
vs. time for standard conditions with
nitrogen and additions of 0.5% Lupersol 130
every 4 hours.
58
and 5.0 weight percent sunflower oil.
were tested with the new
These same conditions
basestock
basestock, gel formation occurred
(Run 12).
between
With the new
15 and 20 hours.
The amount of gel formation at 20 hours in the new basestock
oil with ZDTC present
was
formation at 10 hours in
present.
The
amount
comparable
the old basestock oil without ZDTC
of
gel
formed
the new basestock oil was also
formed in the
new
occurred in
the
to
be
with ZDTC present in
similar to the amount of gel
basestock
ZDTC. ZDTC appeared
to the amount of gel
oil
without
the presence of
ineffective since gel formation
same
time
was
so
frame
as
standard conditions
(around 20 hours).
Because
ZDTC
formation in the
old
effective
basestock
oil
in
minimizing
gel
(exposed to an oxygen
atmosphere) and was considered a key to future research, the
exact same
ZDTC
experiment
(Run
13)
was
repeated.
Gel
formation again occurred in significant amounts at 20 hours.
There appeared
to
be
no
improvement
when
using the new
basestock oil in conjunction with ZDTC.
The new basestock oil may have some metal or contaminant
compound
that
ineffective.
is
deactivating
the
Some difference(s)
ZDTC
and
making
it
between the two basestocks
causes ZDTC to perform differently in each.
From the beginning of
this
laboratory,
it
was
the
sunflower oil experiments in
known
that
ZDTP,
zinc dialkyl
59
dithiophosphate, was the
most
commonly used antioxidant in
the lubricant industry C323 .
ZDTP supposedly inhibits the
initiation stage of autoxidation by decomposing intermediate
hydroperoxides to nonradical products.
the propagation step
by
reacting
It may also inhibit
with the peroxy radical.
The structural formula of ZDTP is given below.
RO
S
>
Zn
RO
Dutta [123 showed
1.0
formation and viscosity
weight
percent
2
weight
rise
sunflower
percent ZDTC blocked gel
in
oil
the old basestock oil/5.0
mixture
percolation, copper foil and 150 C.
was ineffective when used
current
research
antioxidant.
indicates
However,
ZDTP decomposes.
it
Jette
zinc (presumably
from
oil/5.0 weight percent
time.
under
Based on these
may
same conditions.
may
remain
[113
showed
ZDTP)
in
sunflower
results,
to oxygen
One weight percent ZDTP
the
ZDTC
exposed
a
not
be
The
a superior
in the system while
the concentration of
commercial lubricating
oil system decreased with
a
decision to test ZDTP in
the new basestock was made.
Dutta's results with 1.0 weight
basestock
oil/5.0
reproduced with
sunflower
oil
weight
the
under
new
percent ZDTP in the old
percent
basestock
standard
sunflower
oil
were
oil/5.0 weight percent
conditions
(Run
14).
60
Significant
viscosity
rise
was
formation occurred within 20
are shown in Figure
dotted line.
over
12
There
hours.
The viscosity results
improvement with ZDTP present
conditions
formation
occurred
at
a
case
later
time
a
gel
no
was
in
and
Dutta's results depicted by a
standard
and
observed,
with
the
experiment
not
with
time
no
than
similar
frame
ZDTP.
in
to
Gel
Dutta's
a standard
conditions experiment.
Jette
Since
declined with
[113
time
showed
the
concentration
ZDTP
may
be
and
of
zinc
decomposing in this
laboratory system, progressive additions every four hours of
1.0 weight
oil/5.0
oil
sunflower
were
ZDTP
percent
foil
this experiment was run,
the new basestock
to
Again ,
mixture
exposed to oxygen, copper
within 15-20 hours.
made
and
the
system was
C (Run 15).
150
When
significant gel formation occurred
The
ZDTP
added
either initially or
periodically was not effective in the given system.
Neither ZDTC nor ZDTP
had
worked in the new basestock,
yet one of the earlier speculations was that antioxidants in
the commercial
lube
Perhaps there was
an
oil
were
stopping
interaction
oxidation to gel.
between the antioxidants
and some other material in the commercial lube oil.
time, an ongoing
review
stearate
have
might
of
a
At this
the literature revealed copper
synergistic
effect
with
the
61
Basestock
Basestock
K
300
200
T IM E , hrs
Figure 12.
Viscosity comparison of old and new
basestock oils and 5.0% sunflower oil
vs. time for standard conditions with
initial addition of 1.0% ZDTP.
62
antioxidants C353 . The
structure
of copper stearate is as
follows:
(ci7H35C00~)2 Cu+2
Perhaps the
activated
the
old
ZDTC
Copper was a known
reaction and may
reactions.
basestock
whereas
the
new
a
material that
basestock
did not.
catalyst for the addition polymerization
be
Dutta's
a
catalyst
for gel-forming oxidation
copper
and no-copper experiments
E123
could not distinguish
formation.
contained
any
differences
He concluded copper
was
The
current
formed so rapidly in
old
basestock
copper played was
undefinable.
where gel formation occurs
at
In
a
the rate of gel
not a catalyst for the
gel-forming reaction(s).
the
in
researcher felt gel
oil that the role
the new basestock oil
slower
rate, the role of
copper with respect to gel may become clearer.
To establish a control,
10 ppm copper stearate replaced
copper foil in an experiment of new basestock oil/5.0 weight
percent sunflower oil exposed to
It was speculated that
yield Cu
9+ ions
in
oxygen and 150 C (Run 16).
copper stearate might dissociate and
the
system.
However, significant gel
formation occurred in Run 16 within 5 hours which was faster
than with
standard
conditions.
This
was not unexpected
because Jette [113 showed soluble copper was the most active
form of catalyst in
the copper first
had
this
to
system.
When using copper foil,
dissolve
to form active species.
63
With direct
addition
of
copper
stearate,
the copper was
already in a soluble form.
Next copper stearate
ZDTC were both added
17).
ZDTC appeared to
at
(10
ppm)
and
1.0 weight percent
otherwise standard conditions (Run
be
effective
time, but significant gel formed
stearate experiments seemed to
for a short period of
by
16 hours.
produce
Both copper
gel faster than the
standard conditions case.
The ongoing
literature
search
suggested phenols might
work as antioxidants in the present system.
Quinones which
are oxidized phenols might destroy free radicals and thereby
terminate radical reactions C363
might increase the induction
process.
Dutta
butylphenol) in
C123
his
phenol might not have
may
be
the
action C373 .
active.
High levels of phenols
period and delay the oxidation
tried 4,4'-methylenebis (2,6-di-tertresearch.
been
functional
It
converted
form
was
believed Duttas
to the quinone which
responsible
for antioxidant
The tertiary butyl groups might have hindered
the hydroxyl group and the
been
.
Tertiary
quinone structure might not have
butyl
hydroquinone
(TBHQ)
is a
simpler molecule and might be easily oxidized to the quinone
form. This conversion is shown on the following page.
64
When 0.5 weight percent
system with new basestock
TBHQ
was
used in the current
oil, 5.0 weight percent sunflower
oil, copper foil, oxygen and 150 C (Run 18), significant gel
was formed within 20 hours.
There was no improvement with
the addition of TBHQ.
The
complex
additives such as
chemistry
associated
antioxidants
was decided that some
and more fundamental
simpler
oil
resulted
rise
viscosity
oil
again reviewed, and it
variables
examined.
At
removal of the copper foil from
the commercial lubricating
a
using
chemistry must be understood
experimental
this time, Raman C383 found
in
was
with
conditions where the copper was
system
similar
at four hours still
to
the
standard
left in the mixture for the
duration of the experiment.
In the
new
basestock
chemistry might be obtained
system,
more
insight
into the
by manipulating the environment
to which the oil mixture was exposed.
From Raman's results.
65
it was hypothesized that long-lived radicals might be formed
early
in
the
experiment
polymerization.
oxygen for
keep
producing
addition
If the new basestock system were exposed to
four
hours
environment, viscosity
formation.
and
and
then
subjected
rise
might
take
to a nitrogen
place without gel
The initial oxygen environment might produce the
long-lived radicals
that
might not form because
lead
the
to
viscosity,
oxygen
and the gel
would not be present to
polarize the addition polymers.
A standard experiment
of
new
basestock oil/5.0 weight
percent sunflower oil exposed to
copper
run with oxygen for
four
and
remainder (Run 19).
No
hours
viscosity
foil and 150 C was
then nitrogen for the
rise was detected in 48
hours (4 hours of oxygen and 44 hours of nitrogen).
amount of gel formed due
Because viscosity
rise
formed, the theory of
It appears more
to the initial exposure to oxygen.
was
not
detected
and minimal gel
long-lived radicals seems improbable.
likely
copper after four
A small
that
hours
is
Raman's
due
result with removing
to soluble copper species
that remain in the system after the copper is removed.
The body of experimental
oxygen
polymers
was
attacking
and
making
addition polymers
are
lubricating oil.
If
the
these
then
the
data gathered so far indicated
double
bonds
polymers
not
in
polar.
soluble
carbon-carbon
in
the addition
The
polar
the nonpolar
double bonds were
66
converted to addition polymers by Lupersol 130 in a nitrogen
atmosphere, the large
materials
that
dissolved in neutral
solvent might produce viscosity.
Next,
three
experiments
basestock/sunflower
Lupersol
130
atmosphere.
oil
added
were
performed
mixture
every
to
two
1.0
exposing the
weight
hours
and
a
percent
nitrogen
The first (Run 20) subjected the oil mixture to
Lupersol 130, nitrogen, copper foil and 150 C until reaching
a viscosity of
500
centistokes.
then switched to oxygen
(28
Nitrogen percolation was
hours)
to produce a viscosity
rise to 1278 centistokes in
an
additional 24 hours.
viscosity results are shown
in
Figure
ran for another 44
hours
formation was occurring.
No
However, the oil
was a solid at
room
temperature;
conclusion was drawn that
gel.
The
it
the
most
were made to see if gel
apparent
mixture
have been detected even if
convert to
The experiment
where viscosity measurements were
not taken and frequent observations
place.
13.
These
gel formation took
became so viscous that it
therefore, gel might not
were present.
The tentative
polymerized material did not
accessible
converted to addition polymers by
double bonds were
the Lupersol 130, and few
were left for oxygen to attack to form gel.
The
second
experiment
conditions as Run 20, and
(Run
21)
involved
the
same
a viscosity of 215 centistokes at
20 hours was measured before switching to oxygen.
Viscosity
67
continued to rise for the
centistokes.
These
next
results
20 hours until reaching 500
are
also
shown in Figure 13.
The experiment ran for an additional 22 hours while checking
for gel formation.
hours, but
it
supernatant
An apparent
could
hot
phase.
temperature,
be
When
attempts
gel phase was present at 62
separated
the
to
from
mixture
separate
the
the viscous
cooled
to
two
phases
room
by
decanting were unsuccessful.
The result of this experiment
leads
that
to
material
the
conclusion
converts
to
gel
only
accessible double bonds were
moderately
very
converted
polymerized
slowly.
The most
by the Lupersol 130
to addition polymers, and few were left for oxygen to attack
to form gel.
In the third experiment
H O centistokes at 12
environment.
hours
before
switching to an oxygen
Viscosity results are also shown in Figure 13.
Significant gel was formed
to oxygen.
(Run 22), the viscosity reached
The
amount
of
within
15 hours after switching
double
from 60 to H O centistokes may
bonds consumed in going
not have been high enough to
prevent gel formation after oxygen introduction.
The
double
theory
bonds
that
to
polymerization was
further test this
oxygen
yield
was
polar
proceeding
hypothesis,
simultaneously
compounds
appears
a
to
while
be
cleaving
addition
valid.
To
compound that might react
with carboxylic acids and prevent gel formation was tested.
68
Run 20:
to O7 0 500 cSt: o
Run 21: N7 to O7 0 215 cSt: A
T IM E , hrs
Figure 13.
Viscosity of new basestock oil and 5.0%
sunflower oil vs. time for standard
conditions with nitrogen and additions of
0.5% Lupersol 130. Nitrogen switched to
oxygen at times indicated by arrows.
69
Octadecylamine (ODA) is
an
eighteen
carbon amine with
the following formula.
CH3(CH2 )17MH2
The amine should react
with
amides which would be less
soluble in the
nonpolar
the
acids
polar
to form long chain
than the acids and remain
lubrication
oil.
As the carbon-
carbon double bond is being cleaved to form acid groups, the
ODA might react with the acids.
The product molecule should
be a triglyceride with an eighteen carbon amide which should
be somewhat larger
and might remain
than
in
the original triglyceride molecule
solution
due
to its overall nonpolar
character.
Fifteen grams octadecyIamine was
beginning of a
basestock
initially added at the
oil/sunflower
oil experiment and
exposed to copper foil, oxygen and 150 C (Run 23). Viscosity
at 20 hours was 147 centistokes
it
was
309
experiment.
centistokes
gel
experiment
a
with
as
a
can
did not appear
to
oil
occurred and gel particles
seen
commercial
in
formation was detected at twenty
for an additional 16
commercial
Viscosity rise occurred at a slower
standard
be
standard
were difficult to take after
formation
plugged the viscometer.
than
in
Viscosity results
20 hours because
rate
in Run 23 while at 20 hours
hours,
increase.
Figure
hours.
and
lubricating
14.
oil
Slight gel
The experiment ran
the amount of gel present
The
final
quantity of gel
70
commercial oil:A
ODA initially: o
T IM E , hrs
Figure 14.
Viscosity of new basestock oil and 5.0%
sunflower oil vs. time for standard
conditions with 15 g ODA added initially.
71
formed was substantially less than
the amount of gel formed
in a standard conditions experiment with basestock.
It was hypothesized
ODA
might
be evaporating from the
system; therefore, periodic additions
grams ODA were added
every
C
(Run
24),
gel
hours
hours.
18
gel formation.
The ODA
as they formed.
and
exposed to copper, oxygen
formation
experiment ran for 32
An additional
When 0.7
four hours to basestock oil/5.0
weight percent sunflower oil
and 150
were tried.
with
was
prevented.
The
ODA additions every four
hours
of exposure resulted in no
appeared
to be reacting with acids
Viscosity
rise
was similar to a standard
experiment as shown in Figure 15.
The total base numbers (TEN'S)
Figure 16.
Two
TEN
The first sample was
ODA.
the
The second
ODA.
maintained.
samples
taken
sample
An
for
were
Run 24 are shown in
taken every four hours.
immediately prior to adding the
was
average
taken 25 minutes after adding
TEN
of
approximately
0.3
was
These results indicated the ODA was keeping the
system from becoming increasingly acidic.
To test the
polar
speculated the ODA
been formed.
might
A reaction
oil, 4 grams gel from
grams ODA was heated to
"dissolved",
compound
and
a
hypothesis further, it was
solubilize
gel
that had already
kettle with 36 grams new basestock
a standard basestock experiment and 4
150
C.
viscosity
After six hours, the gel
measurement
was
taken.
72
commercial oil:A
ODA at intervals:o
T IM E , hrs
Figure 15.
Viscosity of new basestock oil and 5.0%
sunflower oil vs. time for standard
conditions with additions of 0.7 g ODA
every 4 hours.
TBN1 mgKOH/g
73
TIME, hrs
Figure 16.
Total base number of new basestock oil
and 5.0% sunflower oil vs. time for standard
conditions with additions of 0.7 g ODA
every 4 hours.
74
Basestock oil viscosity was
the
gel
into
centistokes.
basestock
gave
Another experiment
of gel, ODA
oil.
the
60 centistokes while dissolving
and
oil
The gel also
was
a
viscosity
of
75
using the same proportions
run using commercial lubricating
"dissolved" within six hours.
Viscosity
increased from 95 to 155 centistokes.
The acidic gel
further
hypothesis
confirm
this,
spectroscopy was used.
amounts
of
It
Fourier
This
carbonyl
comparisons.
appeared
was
believed
polar and was thus more
oxidized
valid.
To
infrared
should show relative
and
the
be
transform
method
compounds
to
allow
gel
for
sample
material was more
than the sunflower oil by
itself or the same amount of sunflower oil in a degraded oil
mixture
that
had
been
converted
Sunflower oil in degraded
lube
oxidized than pure sunflower
degree of oxidation might
oil
converted
to
oil
oil.
be
to
addition
polymers.
would probably be more
The order of increasing
pure sunflower oil, sunflower
addition
polymers
from
given
and
sunflower
oil
converted to gel.
Since the
swollen
with
reasonably pure
conditions
gel
any
supernatant,
gel.
experiment
it
The
was
swollen
(Run
16)
experiment
was highly
necessary
to
gel
a standard
was
from
washed
at
obtain
room
temperature with octadecane and hexane and then vacuum dried
at
room
temperature
for
two
or
three
minutes.
This
75
procedure was repeated until
the
gel
was dry and crumbly.
The gel was then subjected to a warm nitrogen atmosphere for
30 minutes to remove
any
occluded
solvents.
The gel was
then dissolved in ortho-chlorophenol (OCP) which is a common
solvent for dissolving complex polymers.
Three samples were analyzed
with each sample containing
the same concentration of sunflower
were
I)
pure
conditions
sunflower
experiment
oil/sunflower
oil
oil,
and
from
a
thickened to 300 centistokes.
OCP.
Next
the
three
oil.
2)
The three samples
gel
3)
from
a
commercial
standard
standard
lubricating
conditions experiment
All samples were dissolved in
samples,
containing
the
same
concentration of sunflower oil were analyzed by FTIR.
As expected, the gel was the most oxidized and contained
more carbonyl
groups
than
the
other
two
samples.
The
degraded lubrication oil contained more carbonyl groups than
the pure sunflower oil,
another anticipated result.
17 gives the FTIR spectra.
was determined.
Areas under each of the curves
The degraded lube oil had 1.7 times as many
carbonyl groups as pure sunflower
times as
many
Figure
carbonyls
results were consistent
as
with
oil while the gel had 4.2
pure
the
sunflower
oil.
These
"polar gel" theory and
other experiments.
A tabulated review of all experiments performed is given
in Table 4 of the Appendix.
100.000
Sunflow er OH ( 5 % )
63.333
56.667
Degraded Lube (5 % S.O .)
^
50.000
H
33.333
16.667
GEL
2002.4
1946.4
1890.3
1834.3
( 5 % S.O.J
1778.2
1722.2
If22.2
1666.1
ibiu.i
1554.0
iooh
.u
1498.0
V/AVENUMBERS ( C M - I )
Figure 17.
Infrared
s p e c t r o s c o p y of
addition polymerized
equal concentrations
sunflower
oil and
of
pure
insoluble gel
s u n f l o w e r oil,
from
s u n f l o w e r oil.
77
SUMMARY
The mechanism of gel
formation in basestock lubricating
oil needed to be clarified
at
The experiments performed in
into the
differences
the conditions of this work.
this research provided insight
between
polymerization of
sunflower
and gel formation
appear
gel
formation
oil.
to
and addition
Addition polymerization
occur simultaneously and only
exlude each other as double bonds are consumed.
Once a
polymer
is
polarized
competing oxidation reactions, it
nonpolar lubricating oil.
a two phase system.
certain degree by
has less affinity for the
Smalll molecules that are polarized may
polymerization
simultaneously with
a
The difference in affinity causes
still remain in solution, so it
addition
to
should be kept in mind that
to
large
polarization
to
molecules
acidic
occurs
species.
The
addition polymers that become polarized appear to contribute
to gel formation.
Attempts to make the
of a long
chain
When
acidic
the
amine
system
were
polymers
amides, the oil mixture
less polar by the addition
successful in dissolving gel.
were
existed
converted
as
to
less polar
a single phase.
Total
78
base number
acidic.
carbonyl
results
the
system
was
no longer as
Infrared spectroscopy showed the gel contained more
groups
sunflower oil.
resulting
showed
from
than
All these
addition
pure
sunflower
oil
or
degraded
results confirm that the polymers
polymerization
are
polarized by
oxidation to form the separate gel phase.
These results now facilitate the way for future research
in hydrocarbon basestock oil without the presence of gel.
/
79
CONCLUSIONS
1.
Insoluble gel formed from sunflower oil in basestock
lubricating oil appears to be due to simultaneous
addition polymerization and other oxidation
reactions.
These other oxidation reactions also
take place at points of unsaturation to yield polar
carbonyl groups, especially acids.
These polymers
then lose affinity for the nonpolar lubricating oil
and form a separate phase.
2.
The formation of insoluble gel requires the presence
of oxygen at the conditions of this work.
Sources
of peroxy free radicals other than oxygen do not
yield insoluble gel.
3.
Insoluble gel formation can be prevented by reaction
with long chain amines to yield amides which reduce
overall molecular polarity by addition of a long
chain polar component.
Other long chain basic
species should show similar gel-retarding behavior.
80
4.
The antioxidant, zinc dialkyl dithiocarbamate
(ZDTC), appears to retard gel formation by blocking
addition polymerization and not by inhibiting the
oxidation reactions that yield polar species.
r
81
SUGGESTIONS FOR FUTURE RESEARCH
1.
A new standard experimental procedure needs to be
devised where gel-forming species are converted to
soluble species which can be quantified by viscosity
rise.
2.
Further confirmation of the theory that acidic
species contribute to gel formation should be gained
by alkalinity studies of gel and gel-forming
systems.
3.
Thin layer or gel permeation chromatography should
be investigated to determine the relative polymeric
natures of insoluble gel and soluble addition
polymers.
4.
Copper is known to promote addition polymerization,
and hence viscosity rise.
also be copper promoted.
The formation of gel may
The role of copper with
respect to gel formation needs to be clarified.
Other metals need to be tested for their ability to
catalyze gel formation.
LITERATURE CITED
Geyer, S.M., Jacobus, M.J., and Lestz, S.S.,
"Comparison of Diesel Engine Performance and
Emissions from Neat and Transesterified Vegetable
Oils," Transactions of the ASAE, Vol. 27, No. 2,
pp. 375-384 (1984).
Hunke, A.L. and Barsion, N.J., "Performance and
Emissions Characteristics of a Naturally Aspirated
Diesel Engine with Vegetable Oil Fuels-(Part 2),"
Society of Automotive Engineers Special Publication
SP-495 (1981).
Ryan III, T.W., Dodge, L.G., and Callahan, T.J.,
"The Effects of Vegetable Oil Properties on
Injection and Combustion in Two Different Diesel
Engines," J. Am. Oil Chem. Soc. 61(10):1610-1619
(1984).
Pestes, M.N. and Stanistas, J., "Piston Ring
Deposits When Using Vegetable Oil as a Fuel,"
Journal of Testing and Evaluation, JTEVA, Vol. 12,
No. 2, pp. 61-68, (1984).
Darcey, C.L., LePori, W.A., Yarbrough, C.M., and
Engler, C.R., "Lubricating Oil Contamination From
Plant Oil Fuels," Transactions of ASAE, Vol. 26,
No. 6, pp. 1626-1632 (1983).
Pryde, E.H., "Vegetable Oils as Diesel Fuels;
Overview," J. Am. Oil Chem. Soc. 60(8);1557-1558
(1983).
Sridharan, R., and Mathai, I.M.,
"Transesterification Reactions," J. Sclent. Ind.
Res., Vol. 33, pp. 178-187, (1974).
Hiebert, D.R., "Diesel Fuels From Decarboxylation
of Vegetable Oils," Thesis, Montana State
University, Bozeman, MT (1984).
83
9.
Peterson, G.R., "The Transesterification by
Heterogeneous Catalysis of Rapeseed Oil
Triglycerides to the Methyl Ester Derivatives,"
Thesis, Montana State University, Bozeman, MT
(1984).
10.
Rewolinski,C., Vegetable Oil Dilution of Diesel
Engine Lubricating Oil," Thesis, Montana State
University, Bozeman, MT (1984).
11.
Jette, S., "Copper Catalysis of Polymerization of
Sunflower Oil Diesel Fuel," Thesis, Montana State
University, Bozeman, MT (1985).
12.
Dutta, A., "Polymerization of Lubrication Oil Base
Stock Contaminated with Sunflower Oil," Thesis,
Montana State University, Bozeman, MT (1986).
13.
Peterson, C.L., Wagner, G .L ., and Auld., D.L.,
"Vegetable Oil Substitutes for Diesel Fuel,"
Transactions of the ASAE, Vol. 26, No. 2, pp. 322327 (1983).
14.
Sonntag, N-O1-V., Bailey's Industrial Oil and Fat
Products, 4C ed., Vol. I, Svern D., Ed., pp. 1-45,
135-159, Wiley-Interscience, New York (1979).
15.
Kaufman, K.R. and Ziejewski, M., "Vegetable Oils as
a Potential Fuel in Direct Injection Diesel
Engines," Soc. of Automotive Engineers, Inc., pp.
1-28 (1983).
16.
Rheineck, A.E. and Austin, R.O., "Treatise on
Coating," Myers, R.R.tand Long, J.S., Eds. Vol. I,
Part 2, Ch. 4, Marcel Dekker, New York (1968).
17.
Stephens, H.N., J. Am. Chem. Soc., 50:568 (1928).
18.
Farmer, E.H. and Sutton, D.A., J. Chem. Soc. 119
(1943).
/
19.
Formo, M.W.». Bailey's Industrial Oil and Fat ^
Products, 4t' Ed., Vol. I, Swern, D., Ed., pp.
678-716, John Wiley & Sons, New York (1979).
20.
Bolland, J.L. and Gee, G., Trans. Faraday Soc.
42:244 (1946).
21.
Farmer, E.H., Trans. Faraday Soc., 42:228 (1946).
84
22.
Gunstone, F.D. and Hilditch, T.P., J. Chem. Soc.
1022 91946).
23.
Swern, D., Fatty Acids, 2nd ed., Part 2, Markley,
K.S., Ed., pp. 1387-1436, Interscience Publishers,
Inc. New York (1961).
24.
Wexler, H., Chem. Reviews 64(6):591 (1964).
25.
Uri, D., Autoxidation and Antioxidants, Lundberg,
W.O., Ed., Vol. I, Ch. 2, Wiley-Interscience, New
York (1961).
26.
Morrison, R.T. and Boyd, R.N., Organic Chemistry,
1st ed., Allyn and Bacon, Inc., Boston, Mass., pp.
152-156 and 628-629 (1959).
27.
Wingrove, A.S. and Caret, R.L., Organic Chemistry,
Is ed., Harper and Row, New York, pp. 339-342 and
867-898 (1981).
28.
Sheldon, R.A. and Kochi, J.K., "Metal Catalyzed
Oxidations," pp. 272-291.
29.
"Total Base Number of Petroleum Products by
Potentiometric Perchloric Acid Titration," ASTM D
2896, 1974 Annual Book of ASTM Standards, part 24,
ASTM, pp. 870-875 (9174).
30.
"Iodine Value of Drying Oils and Fatty Acids," ASTM
D 1959-69, Annual Book of ASTM Standards, part 29,
ASTM, pp. 283-286 (1979).
31.
Fourier Transform Infrared Spectroscopy. Ferraro,
J.R. and Basile, L.J., Eds., Vol. 4, Ch.l, Academic
Press, Inc., Orlando, Florida, pp. 1-2 (1985).
32.
Schiff, S., Personal Communication, Phillips
Petroleum Company, Bartlesville, OK (1986).
33.
Berg, L., Personal Communication, Montana State
University, Bozeman, MT (1986).
34.
Mendesf T., Personal Communication, Montana State
University, Bozeman, MT (19861.
35.
Scott, G., "Some New Concepts in Polymer
Stabilisation," British Polymer Journal, Vol. 3,
pp. 24-27 (1971).
yj 1 1.
85
36.
Sims, R.J., and Fioriti, J.A., "Methional as an
Antioxidant for Vegetable Oils," J. Am. Oil Chem.
Soc. 54(1);4-7 (1976).
37.
Jennings, P.W., Personal Communication, Montana
State University, Bozeman, MT (1986).
38.
Raman, R . , Personal Communication, Montana State
University, Bozeman, MT (1986).
86
APPENDIX
Table 4:
I
2
3
4
5
6
7
8*
9*
10*
11*
12*
13*
14*
15*
16*
17*
18*
19*
20*
21*
22*
23*
24*
Oxygen Nitrogen
yes
yes
yes
yes
no
yes
yes
yes
yes
yes
no
yes
yes
yes
yes
yes
yes
yes
4 hr s
28 hrs
20 hrs
12 hr s
yes
yes
Copper
foil
no
foil
no
no
foil
foil
no
foil
yes
no
foil
foil
no
foil
no
no
foil
no
foil
foil
yes
no
foil
foil
no
foil
no
foil
no
stearate
no
stearate
no
foil
no
44 hr s
foil
foil
68 hrs
foil
42 hrs
21 hrs,
foil
foil
no
foil
no
* indicates new basestock
I
I •
I O
I •
I 02
I
I
Run
Operation Parameters for Oil Bath Runs
25
25
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
Temp.
150
135
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
150
Additives
none
none
Paranox
ZDTC
Lupersol
Lupersol
Lup, ZDTC
none
Lup, ZDTC
Lupersol
Lupersol
ZDTC
ZDTC
ZDTP
ZDTP adds
none
ZDTC
TBHQ
none
Lupersol
Lupersol
Lupersol
ODA
ODA adds
MONTANA STATE UNIVERSITY LIBRARIES
3
762 100 4251 O
DATE DUE
I if /
i I
I 7
AR^irear
-
"
■
2 G'2 » "
^
c ;
APf 1 2 2002
HIGH S M ITH REORDER #45-230
I