Article

pubs.acs.org/IECR

Regeneration of Waste Lubricant Oil by Extraction−Flocculation

Composite Refining
Xin Yang, Ligong Chen,* Shuo Xiang, Liang Li, and Di Xia
Department of Military Oil Application & Management Engineering, Logistical Engineering University of PLA, Chongqing 401311,
China

ABSTRACT: Waste lubricant oil can be regenerated by solvent extraction and flocculating agent to get rid of impurities. In this
paper, butanol was selected as an extractant, and monoethanolamine (MEA) as a flocculating agent. The aim of the work was to
identify the best refining conditions to separate base oil suitable for the formulation of new lubricants and avoid the coextraction
of impurities. The effects of time, temperature, solvent/waste oil mass ratios, and flocculant concentration on the separation
efficiency and yields were investigated. Optimization conditions of the refining were obtained by experiments: refining time = 20
min, temperature = 30 °C, solvent/waste oil mass ratios = 5g/g, flocculant concentration = 2g/kg solvent. When comparing i-
butanol/MEA and n-butanol/MEA, it was found that yields obtained with composites of i-butanol/MEA were slightly higher
than those obtained with n-butanol/MEA. On the other hand, because MEA not only agglomerates waste oil contaminants but
also reduces the dosage of solvent, the solvent/waste oil mass ratios at which stabilization occurs with composite butanol/MEA
were smaller than that obtained with single pure solvent (n-butanol, i-butanol). Under composite refining, metallic and oxidation
compound removal was very efficient, and the properties of the recovered oil were notably improved than that in the original
waste oil. The recovered oil was almost similar to a HVI150 virgin oil and, therefore, suitable for the formulation of new lubricant
oil.

1. INTRODUCTION In this study, alcohols with four carbon atoms were selected
Waste lubricating oils are byproducts of oil used in vehicles and as extractants that dissolve the base oil, and monoethanolamine
machinery. Oil must be replaced because of the degradation of (MEA) was chosen as a flocculating agent that removes the
the fresh lubricant components and contamination from dirt, contaminants from waste oil for the purpose of reusing it as the
water, salts, metals, incomplete products of combustion, lubricant base oil. First, this process seems to overcome the
additives, and other asphaltic compounds coming from overlay major problem of the sulfuric acid treatment: the production of
on bearing surfaces. Once replaced by new lubricants, the used an acid sludge, which is a source of pollution and causes very
oil is a matter of greater environmental concern: polluting difficult disposal problems. Then, for the process solvent can be
water and earth and causing air pollution if burnt. Therefore, to recovered and recycled, since the organic sludge produced may
prevent environmental pollution and to conserve petroleum be used as an asphalt component or may be mixed with liquid
sources, waste oil should be regenerated. fuels and burned. The best use for this sludge is, according to
Several technological processes have been proposed to Alves dos Reis et al., an offset ink component.7,8 More
recycle waste oil. The most familiar technology is the acid- specifically, the effect of the refining time, refining temperature,
clay process, but the process was prohibited because the acid solvent/waste oil mass ratios, flocculant concentration on both
sludge is a source of pollution that causes serious environ- the yield and quality of the recovered oil were studied. The
mental problems.1 In the acid-free clay process the waste oil is quality was evaluated through the determination of phys-
pretreated using atmospheric distillation to remove water and icochemical properties, metallic compounds, and oxidation
light hydrocarbons followed by vacuum distillation and products. The major purpose of this paper is to offer a feasible
hydrogenation treatment.2 However, this process is not highly choice for waste oil recycling by extraction−flocculation, and
efficient in removal of contaminants, and low quality recovered select the best process conditions leading to both high yield and
oil is obtained with unpleasant smells and poor stability.2−4 As quality of the recovered oil.
mentioned above, persistent efforts should be made to improve
waste oil regeneration technologies. Using a single solvent or 2. EXPERIMENTAL SECTION
composite solvents to extract base oil from waste oil was a
feasible alternative before vacuum distillation, which is 2.1. Materials. Organic solvents (n-butanol, i-butanol) were
economical, efficient, and safe. Unfortunately waste oil supplied by Chongqing Reagent Company. Waste lubricant oil
constituents are kept in a stable dispersion as a result of was supplied by Swift Horse Engine Repair Factory. First, the
electric repulsion between heteroatoms and asphaltic com- waste oil was treated in a rotary evaporator at 70 °C under
pounds in the solvent system, a result of which is that the
impurities cannot be separated from the base oil. 5,6 Received: May 12, 2013
Accordingly, the addition of a flocculating agent exerts an Revised: July 31, 2013
action to break this stability and promotes a fast flocculation of Accepted: August 19, 2013
impurity particles. Published: September 3, 2013

© 2013 American Chemical Society 12763 dx.doi.org/10.1021/ie4015099 | Ind. Eng. Chem. Res. 2013, 52, 12763−12770
Industrial & Engineering Chemistry Research Article

vacuum (500 mmHg) to eliminate water and light hydro- As noted before, a large number of pure organic liquids, such as
carbons. Both types of compounds are undesirable for the hydrocarbons, alcohols, and ketones exhibited the properties
formulation of new lubricants and may modify the property of mentioned above and were applied to the extract waste
base oil components in the solvent. Waste oil properties after oil.2−6,9−12
this treatment are shown in Table 1. At the industrial scale this The following points were considered in the selection of an
step usually performed at atmospheric pressure. We made it optimal solvent: Hildebrand solubility theory of the solvents,
under vacuum to accelerate the separation process. the existing background on the regeneration of waste oil by
solvent extraction, and the selectivity or capacity of the solvent
Table 1. Properties of Waste Lubricating Oil to selectively extract base oil from waste oil.
First of all, it should be noted that impurity particles usually
waste oil have a high molecular weight. According to the Hildebrand
appearance black opaque solubility theory,13 they should dissolve more easily in the
w (water) (%) trace heavier solvents because the solubility parameters of the heavier
v40 (mm2.s 1) 104.65 solvents are close to those of the impurity particles. Therefore,
v100 (mm2.s 1) 11.58 in this study, low-molecular weight solvent was used, which
viscosity index 92.5 inhibits the dissolution of the impurity particles.
flash point (°C) 184 Then by previous works, when mixed with waste oil, liquid
pour point (°C) −7 hydrocarbons not only extract the base oil fractions, but also
w(sulphur) (%) 0.29 the solution of macromolecular and other additives as well as
w(sulfate ash) (%) 0.77 dispersion of carbonaceous and other particles.5,6 Conse-
oxidation stability 150 °C/min 148 quently, the liquid hydrocarbons were excluded in this research.
acid number (mgKOH·g−1) 1.53 According to Rincon et al.2−4 and Alves dos Reis et al.,5,6 the
w (metallic content) (mg/kg) ketones and alcohols, between three and five carbon atoms,
Zn 713 have been studied to recycle used oil. As a result of the
Ca 1090 molecular weight, alcohols with three carbon atoms can
Mg 332 dissolve little base oil, while those with five carbon atoms not
Fe 238 only extract base oil but also dissolved impurities.2−6 In regards
to the ketones, the only ketone with less than four carbon
2.2. Extraction Procedure. Mixtures of about 20 g of atoms, propanone, was almost immiscible with base oil at room
pretreatment used oil and solvent in solvent/waste oil mass temperature. On the other hand, both base oil and impurities
ratios ranging from 1/1 to 11/1, and MEA in weight were dissolved in ketones with five carbon atoms, which act the
proportions ranging from 1g/kg solvent to 3g/kg solvent same as the alcohols with five carbon atoms. Besides, ketones
were agitated to ensure adequate mixing (see section 4.1). were strictly controlled in China and availability was difficult.
Then, the mixtures were poured into centrifuge tubes that were Therefore, it was found that alcohols with four carbon atoms
later introduced into the support of a centrifuge. After performed as the efficient extraction solvent.
centrifugation at 600 rpm for 8 min, a sludge phase (additives, Finally, the dipole of butanol was close to that of the
metallic and oxidation compounds, and carbonaceous particles) lubricating oil (naphthenic distillate) (see Table 2), which is
was segregated from the mixture of solvent and oil. The solvent important for support of the solvent to selectively extract base
was separated from the mixture solution by distillation in a oil from waste oil.
rotary evaporator and the recovered oil was weighed. The
extraction yield was calculated as the mass of oil, expressed in Table 2. Dipole of the Solvent and Lubricating Results
grams, separated from 100 g of waste oil. The sludge phase was
also weighed in all experiments. lubricatingb
solvent n- i- s- t- (naphthenic
2.3. Analysis of the Metallic Content. The oil samples property butanolb butanolb butanolb butanolb distillate)
were heated at 200 °C for 4 h and calcined at 650 °C for 8 h μ, D 1.60 1.64 1.66 1.44a 1.23−1.64
prior to the analyses. The noncombustible residue obtained a b
3.26 cm wavelength. Reference 14.
(ashes) was further treated with hydrochloric acid, filtered, and
diluted with deionized water to determine the metallic content.
These analyses were performed by atomic emission spectros- In conclusion, solvent properties suggest that alcohols with
copy using an inductively coupled plasma Spectra HK-9600 four carbon atoms were the best candidates to become the basic
spectrometer. component of the extract solvent. In the study, s-butanol and t-
2.4. Other Analyses. Sulfur content was measured with a butanol were excluded because s-butanol yields two liquid
fluorescence measuring apparatus of ZDS-2000. Acid number, phases at room temperature, and the solid of t-butanol forms
pour point, flash point, sulfate ash, oxidation stability, and below 25.5 °C.
viscosities at 40 and 100 °C were determined according to GB/ 3.2. Choice of Flocculant. The impurities of waste oil are
T264, GB/T3535, GB/T3536, GB/T2433, SH/T0193, GB/ mostly composed of (a) oxidation products, (b) metallic
T265 procedures, respectively. compound, (c) degenerative additives, and (d) carbonaceous
particles. The bulk of these particles are kept in stable
3. EXTRACTANT AND FLOCCULANT dispersion by the dispersant additives.
3.1. Selection of Extractant. The extractant that is Author: Please verify that the changes made to improve the
selected must have two important properties: it should be English still retain your original meaning.At room temperature,
miscible with the base oil contained in the waste oil being when mixed with waste oil, the alcohol with four carbon atoms
processed and it should reject from a solution with impurities. dissolved the desired base oil. But because of the electric
12764 dx.doi.org/10.1021/ie4015099 | Ind. Eng. Chem. Res. 2013, 52, 12763−12770
Industrial & Engineering Chemistry Research Article

repulsions, it could only segregate a small part of the one-factor experimental design was adopted in order to select
undesirable materials which formed flakes that settled by the best process conditions leading to both high yield and
gravity.3,5 To avoid the formation of stable dispersions, quality of the base oil obtained.
flocculating agents can be used and added into the solution. Effect of Refining Time. To determine the mixing time for
By the process of flocculation, particles in suspension form systems to attain equilibrium, a group of experiments was made.
larger agglomerates, or small agglomerates already formed grow This group of experiments was performed at temperature of 30
as a result of coagulation through high molecular weight °C, solvent/waste oil mass ratios of 5g/g, and flocculate
polymeric materials.15,16 concentration of 2g/kg solvent.
Attention should be paid to an important finding of a suitable Results were obtained as follows (Figure 1). It can be
substance as a flocculating agent for recycling waste oil. In this observed that the extraction yields increase with time until a
paper, the flocculating agent was chosen by two steps. The first
step was the existing background on the regeneration of waste
oil by flocculation, and the second step was the difference in the
solubility parameters of the solvent and of a typical polyolefin,
polyisobutylene. Polyisobutylene is generally introduced in
lubricant oils as a viscosity index improver, where it acts to
remove the macromolecules of polymeric additives. Moreover,
according to Alves dos Reis et al,5,6 when flocculation occurs,
macromolecules and carbonaceous particles coflocculate and
settle together. When the solubility parameter difference is
high, one should expect to obtain high sludge removal values.5,6
Some materials were proposed for this operation, namely, l-
butanol by Brownawell and Renard,17 and 2-propanol by
Rincon et al.,3 Alves dos Reis et al.5 and Whisman et al.,18 a
substance with special amidogen groups by Zhang,19 and a
sorbitol oleic acid ester with special amidogen groups by
Xiong.20 All of the substances mentioned above exhibit
flocculant properties and have potential utility to coagulate Figure 1. Effect of refining time on yield. Composite, i-butanol/MEA,
n-butanol/MEA; solvent/waste oil mass ratios, 5g/g; MEA concen-
the additives and other undesirable impurities in waste oil. tration, 2g MEA/kg solvent; T = 30 °C.
According to Alves dos Reis et al.,5,6 the flocculating action
for the segregation of sludge from waste oil by polar solvents is
basically an antisolvent effect exerted on some macromolecules. constant value is reached. With butanol/MEA, the constant
In this study, propanol, butanol, and MEA were chosen, and the yield was approximately 81%−83%, which was achieved
individual component was guided by the difference |δ1 − δ2|, approximately after 20 min of refining. This was because the
where δ1 and δ2 are the solubility parameters of the flocculating contact time of the solvent, flocculant, and the waste oil played
agent and polyisobutylene, respectively. When |δ1 − δ2| was an important role for systems to attain equilibrium. On one
high, high sludge removal results are expected. Table 3 suggests hand, it should be long enough to allow the solvent to act on
the base oil contained in the waste oil being processed. On the
Table 3. Difference between the Solubility Parameters of other hand, it should allow flocculant neutralized charges of
Solvent δ1, and Polyisobutylene, δ2 particulate matter to break stabilization by electric repulsion.11
flocculant δ1, δ2(MPa1/2) δ1-δ2 (MPa1/2) Furthermore, flocculant (MEA) maybe adsorb and “bridge”
colloidal particles by permitting their aggregation to particle
polyisobutylenea 15.5
sizes big enough to separate from the liquid phase.15,16 The
1-propanola 23.5 8.0
slightly higher yield and lighter extract coloration attained with
2-propanola 24.5 9.0
i-butanol may be explained by considering the polarity and
1-butanola 23.1 7.6
molecular structure. It is well-known that components exhibit
2-butanola 22.2 6.7
better solubility as their polarities get closer to each other. Since
MEAa 31.5 16.0
a
base oil is a mixture of nonpolar or slightly polar molecular
Reference 21. components, its polarity is closer to that of i-butanol (55.2)
with lower polarity than n-butanol (60.2).22 This difference is
that the solubility parameter of MEA was far different from that probably due to the molecular structure, as i-butanol has a
of the polyisobutylene, which would be a better flocculating branch chain the same as the base oil molecules (ring
agent than butanol and propanol. In summary, to improve the hydrocarbons and alkane hydrocarbons with branch chain),
quality of the base oil extracted from the used lubricant oil and therefore according to the law of affinity has a larger
without renouncing high extraction yields, MEA was deemed to capacity to dissolve base oil components.
be a good flocculating agent. When the mixing time is compared with that obtained by
Rincon et al.,2 it can be observed that the equilibrium for 2-
4. RESULTS AND DISCUSSION butanol is attained at a shorter mixing time, less than 10 min. It
4.1. Effect of Factors. On the basis of the statement of the is probably for that reason that flocculant (MEA) was added in
basic problem, several factors involving base oil yield and this researchit increases the time of flocculation and
quality were investigated: the effect of refining time, the effect sedimentation of the impurities to be removed. Finally,
of refining temperature, the effect of solvent/waste oil mass according to Rincon et al.2 and Alves dos Reis et al.,5,6 as
ratios, and the dosage of flocculating agent. In this research, regards to the longer equilibrium time of the 2-pentanol,
12765 dx.doi.org/10.1021/ie4015099 | Ind. Eng. Chem. Res. 2013, 52, 12763−12770
Industrial & Engineering Chemistry Research Article

approximately 25 min, it may be because the higher viscosity of

2-pentanol increases the time necessary for the contact and
dissolution of base oil in this solvent. To ensure that
equilibrium is reached with the whole system, the refining
time of 20 min was chosen for performing regeneration of
waste oil.
According to Figure 1, the recovered oil attained high yield at
shorter contact time. Maybe some impurities did not have
enough time to aggregate and, as a consequence, they were
coextracted together with the base oil. To test this hypothesis,
the metallic content of the recovered oil with i-butanol/MEA
versus the evolution with time was determined. As expected, it
can be observed in Figure 2 that for the shortest mixing time

Figure 3. Effect of refining temperature on yield.

Figure 2. Effect of time on metallic concentration of the recovered oil.

Composite, i-butanol/MEA; solvent/waste oil mass ratios, 5g/g; MEA
concentration, 2g MEA/kg solvent; T = 30 °C.
Figure 4. Effect of temperature on metallic concentration of the
recovered oil. Composite, i-butanol/MEA; solvent/waste oil mass
the metallic content of the recovered oil was very similar to that ratios, 5g/g; MEA concentration, 2g MEA/kg solvent; t = 20 min.
of the untreated waste oil (Table 1) and the concentrations of
these metallic impurities decreased as mixing times increased
from 5 to 20 min, finally, being approximately constant at 20 the optimum reaction temperature was considered to be
min onward. around 30 °C.
Effect of Refining Temperature. Studies were carried out at Effect of Solvent/Waste Oil Mass Ratios. These experi-
different refining temperatures performed with i-butanol/MEA ments were performed by keeping the time and temperature at
from 10 to 50 °C, refining time of 20 min, solvent/waste oil the best refining conditions selected for recovering base oil
mass ratios of 5g/g, and flocculant concentration of 2g/kg from waste lubricant oil with i-butanol/MEA and n-butanol/
solvent. Of note higher temperatures result in solvent MEA, at 20 min and 30 °C, respectively. The flocculant
vaporization, and consequently, temperatures beyond 50 °C concentration was 2g/kg solvent, and solvent/waste oil mass
were not tested. ratios varied from 1 to 11 g/g.
As shown in Figures 3 and 4, when the extraction Figure 5 shows that, for both solvents, the refining yields
temperature increased, the amount of recovered oil increased, increased with increasing solvent/waste oil mass ratios up to a
and the percentage of base oil dissolved in i-butanol increased point at which they stabilize, and it can also be observed that,
from 78.3% to 83.6%. However, the concentration of impurities once yields stabilize, those obtained with n-butanol/MEA and i-
first decreased and then increased with the increasing reaction butanol/MEA occur at a similar solvent/oil mass ratios of 5g/g.
temperature. The reason why recovered oil yields constantly According to Rincon et al.,2−4 the reason for yields of recovered
increased might be due to two factors: on one side, with oil to increase with increasing solvent/waste oil mass ratios and
refining temperature increasing, the viscosity of organic solvent then to stabilize is probably because of the combined effect of
mixtures be reduced, which results in an increase in the amount the following factors: at the smaller ratios, the solvents saturate
of oil dissolved in i-butanol.2,3 On the other side, the and do not dissolve all base oil in the waste oil. Then, the base
temperature was higher and the solubility of the base oil oil dissolved increases with increasing solvent/waste oil mass
components in the extraction solvent increased more. Mean- ratios, but only up to a ratio at which the base oil fraction that
while, as regards to the content of impurities with increasing the solvent may dissolve is exhausted; thus, from this moment,
temperature of the variable, it may be attributed to the the yield will not increase anymore. It can also be indicated that
increased solubility of solvent, which results in a large number the solvent/waste oil mass ratios at which stabilization occurs
of impurity groups dissolved in solvent. On the whole, the with butanol/MEA is smaller than those obtained with single
optimum temperature should result in, simultaneously, butanol as shown by Rincon et al.2 and Alves dos Reis et al.5,6
maximum sludge removal and minimum oil losses. Therefore, In these cases, we concluded that MEA could not only
12766 dx.doi.org/10.1021/ie4015099 | Ind. Eng. Chem. Res. 2013, 52, 12763−12770
Industrial & Engineering Chemistry Research Article

oil. Figures 7 and 8 show the metallic content of the recovered

oil at different solvent/oil mass ratios with composition of i-

Figure 5. Effect of solvent/waste oil mass ratios on yield. Composite,

i-butanol/MEA, n-butanol/MEA; MEA concentration, 2g MEA/kg Figure 7. Effect of solvent/waste oil mass ratios on metallic
solvent; t = 20 min; T = 30 °C. concentration of the recovered oil. Composite, i-butanol/MEA;
MEA concentration, 2g MEA/kg solvent; t = 20 min.
agglomerate waste oil contaminants but also reduce the dosage
of solvent, to accelerate equilibrium for the system.
To test this hypothesis, the stabilization of solvent/waste oil
mass ratios obtained with a single solvent (n-butanol, i-butanol)
was determined. As expected, it can be observed in Figure 6

Figure 8. Effect of solvent/waste oil mass ratios on metallic

concentration of the recovered oil. Single solvent, i-butanol; t = 20
min; T = 30 °C.

butanol/MEA and i-butanol. These results supported the

explication mentioned above. Lastly, the darker coloration of
Figure 6. Effect of solvent/waste oil mass ratios on yield. Single
solvent, i-butanol, n-butanol; t = 20 min; T = 30 °C. the base oil with single i-butanol recycling was more evidence
to indicate a higher level of impurities in recovered oil.
According to the previously mentioned results, a solvent/oil
that without the addition of MEA, the solvent/waste oil mass mass ratio equal to 5 g/g was selected for performing
ratios up to 9 for the system are stabilized, which is almost 2 experiments.
times higher than that by addition of MEA. Note that for Effect of Flocculant Concentration. This group of experi-
butanol to be miscible with both the base oil and flocculation of ments was performed at a refining time of 20 min, refining
waste oil impurities, the solvent used must be plentiful enough temperature of 30 °C, and solvent/waste oil mass ratios of 5g/
to ensure the system reaches equilibrium, since the addition of g.
MEA keeps the suspended impurity particles separated from It can be seen in Figure 9 that, refining yields at first
waste oil by neutralizing the electric forces, and MEA may also decreased as the concentration of MEA in the solvent increased
may adsorb two or more particles by “bridging”.23−25 As a up to a point, and then the refining yields begin to increase as
result, the consumption of solvent was reduced which exerts the the concentration of MEA in the solvent increased. The fact
dominating action to dissolve base oil molecules. that refining yields decreased as the concentration of MEA
According to Figure 6, it can be seen that, just as in the single increased may be imputed to two different factors. First, with an
pure solvent case, once stabilization was reached, yields of increasing addition of MEA, charged particles were adsorbed by
recovered oil obtained with single pure alcohol were higher the MEA and aggregation occurred as a result of particle−
than the composition of solvent and flocculant. A possible particle collisions, which led to small clusters. Consequently,
explication of this finding might be that, in the case of refining due to particle−cluster and cluster−cluster collisions, small
waste oil by a single alcohol, impurities did not completely clusters formed larger aggregates and sedimentation.15,16 In
agglomerate and were partly coextracted together with the base addition, maybe hydroxyl(−OH) and amido(−NH2) of MEA
12767 dx.doi.org/10.1021/ie4015099 | Ind. Eng. Chem. Res. 2013, 52, 12763−12770
Industrial & Engineering Chemistry Research Article

concentration of MEA over 2g/kg, the metallic concentration

in the recovered oil began to increase. It also should be
considered that the MEA exhibits alkalescence, which was
introduced to neutralize the acid compounds of the waste oil; in
principle, more MEA that is added leads to less acidic
compounds of the recovered oil. To confirm the last
hypothesis, the quality of the oil refined was assessed through
the measurement of the concentrations of oxidation products.
Relative concentrations are shown in Table 4. It can be seen

Table 4. Effect of the MEA concentration on the acid

numbera
(g of MEA)/(kg of solvent) acid number (mg of KOH)/g
0 1.53
0.5 0.56
Figure 9. Effect of flocculant concentration on yield. Composite, i- 1 0.18
butanol/MEA, n-butanol/MEA; solvent/waste oil mass ratios 5g/g; t = 1.5 0.08
20 min; T = 30 °C. 2 0.04
2.5 <0.01
were formed and by hydrogen bonding were adsorbed onto 3 <0.01
a
colloidal particles where individual particles can become Composite, i-butanol/MEA; solvent/waste oil mass ratios, 5g/g; t =
attached to two or more particles, thus “bridging” them 20 min; T = 30°C.
together to sedimentation.23−25 However, after a concentration
up to 2g/kg, refining yields begin to increase as the that the acid number decreased with increasing MEA
concentration of MEA increases. This phenomenon may be concentration and that the concentration of oxidation
attributed to the equilibrium which was destabilized by the compounds was smaller than that required for a lubricant oil,
addition redundant of MEA. In other words, maybe the that is., <0.05 mg/g, when the concentration of MEA in the
hydrogen bonds between the polar impurities and the solvent was 2 g/kg solvent. As pointed out earlier, this should
dispersant-detergent additives of the waste oil were broken by be due to neutralization followed by precipitation of acid
superabundant MEA.23−25 Therefore, contaminations did not compounds.
completely agglomerate and were partly coextracted together Meanwhile, it is important to point out that low quantities of
with the base oil. According to La Mer et al,26−28 the most potassium hydroxide (KOH) were added to the solution to
perfect impact for “bridging” is when the surface of impurity break the stabilization by electric repulsion according to Rincon
particles is covered with about 50% flocculating agent, et al.3 and Alves dos Reis et al.5,6 The addition of 2−3 g/L
otherwise, destabilization occurs because colloidal particles KOH has proved to increase the sludge and/or additive
adsorb the surplus of flocculant and, adversely effect the removal from waste oils.3,5,6 A combination of the above
“bridging”. discussion and the literature,3−6 allows us to reach the
To test this hypothesis, the concentration of a type of conclusion that the −OH groups of the potassium hydroxide
impurities (metallic compound) in the recovered oil was and MEA can neutralize acid compounds and induce their
determined. Results from these analyses are shown in Figure saponification and subsequent precipitation from the waste oils.
10. Confirming the suggested hypothesis, results indicated that 4.2. Properties of Recovered Oil. The recovered oil was
the metallic concentration in the recovered oil began to obtained at optimum conditions of the following variables:
decrease with a larger concentration of MEA, while with a refining time, 20 min; refining temperature, 30 °C; solvent/
waste oil mass ratios, 5g/g; MEA concentration, 2g/kg solvent.
Under certain conditions, the quality of the recovered oil was
tested (Table 5).
As shown in Table 5, waste oil characteristics notably
improved after the extraction/flocculation treatment, the
quality of recovered oil becoming very similar to that of HVI
150 virgin oil. First of all, the lighter coloration of the recovered
oil after being regenerated at the optimal conditions indicated
that the extraction solvent had dissolved the base oil fraction
and the flocculant agglomerated waste oil impurities. Mean-
while, according to Rincon et al.,2 the lighter coloration of the
base oil extract indicated higher solvent/oil mass ratios. Then,
the viscosity of the butanol/MEA treated oil was lower than
that of the crude waste oil. Because the flocculating agent
neutralizes the acid compounds of the oil, coagulation occurs
and large particles separate from the liquid phase. The acid
Figure 10. Effect of flocculant concentration on metallic concentration number being smaller than that required for HVI150 lubricant
of the recovered oil. Flocculant, MEA; solvent/waste oil mass ratios oil and the oxidation stability being higher than that required
5g/g; t = 20 min; T = 30 °C. for HVI150 lubricant oil indicates that both acid compounds
12768 dx.doi.org/10.1021/ie4015099 | Ind. Eng. Chem. Res. 2013, 52, 12763−12770
Industrial & Engineering Chemistry Research Article

Table 5. Properties of Recovered Oil

w(water) v40 v100 viscosity
characteristic appearance (%) (mm2.s−1) (mm2.s−1) index flash point (°C) pour point (°C)
recovered oil (i-butanol/MEA) salmon pink transparent trace 47.12 7.38 134.4 212 −11
recovered oil (n-butanol/MEA) salmon pink transparent trace 46.43 7.22 132.0 212 −11
HVI150 base oil transparent trace 28−32 report 100 (min) 200 (min) −9 (max)
w (metallic content) (mg/kg)
w(sulfur) w(sulfate ash) Oxidation stability
characteristic (%) (%) 150 °C/min acid number (mgKOH·g−1) Zn Ca Mg Fe
recovered oil 0.15 0.009 206 0.04 306 452 76 73
(i-butanol/MEA)
recovered oil 0.19 0.01 204 0.04 311 449 74 78
(n-butanol/MEA)
HVI150 base oil report report 200 (min) 0.05 report report report report

and other oxidation products can be removed by the addition but also reduces the dosage of solvent needed for the system to
of MEA. Furthermore, it is well-known that sulfate ash attain equilibrium.
originates from additives and bearing surfaces, which form With the composite, refining and metallic and oxidation
macromolecular compounds. The distinct decrease of the compound removal are very efficient, and the characteristics of
sulfate ash may be explained by the fact that MEA may make the recovered oil are obviously improved. Furthermore,
these macromolecular compounds attach to two or more oxidation stability, viscosity, pour point, and flash point are
particles thus “bridging ” them together to form large flakes, similar to those of HVI-150 virgin oil.

■
which settled by gravity or centrifugation. As regards to the
decrease of the metallic concentration, it may be explained by AUTHOR INFORMATION
the following points: on one hand, metallic impurities formed
Corresponding Author
larger aggregates and sedimentation occurred with addition of
flocculant (MEA). On the other hand, according to Rincon et *E-mail: clg818@qq.com.
al.2 and the Hildebrand solubility theory,13 metallic impurities Notes
were removed because the solvent molecular weight is much The authors declare no competing financial interest.
smaller than that of metallic compounds, which made the
dissolution of the metallic impurities difficult in the lighter
solvents. Finally, evident improvement of both pour point and
■ REFERENCES
(1) Kamal, A.; Khan, F. Effect of extraction and adsorption on re-
flash point may be also imputed to the following factors: refining of used lubricating oil. Oil Gas Sci. Technol. 2009, 64, 191−
oxidation compounds and the fact that carbonaceous particles 197.
flocculated with the addition of MEA and were removed by (2) Rincon, J.; Canizares, P.; Garcia, M. T. Regeneration of used
sedimentation. lubricant oil by polar solvent extraction. Ind. Eng. Chem. Res. 2005, 44,
The quality of the oil refined with i-butanol/MEA was very 4373−4377.
close to those obtained with n-butanol/MEA. It was probably (3) Rincon, J.; Canizares, P.; Garcia, M. T. Waste oil recycling using
because the alcohols of equal numbers of carbon atoms exhibit mixtures of polar solvents. Ind. Eng. Chem. Res. 2005, 44, 7854−7859.
(4) Rincon, J.; Canizares, P.; Garcia, M. T.; Gracia, I. Regeneration of
the property to dissolve the same base oil fraction. Besides,
used lubricant oil by propane extraction. Ind. Eng. Chem. Res. 2003, 42,
since the same flocculant (MEA) with the same concentration 4867−4873.
was used in the recycling process, the parallel impurities (5) Alves dos Reis, M.; Silva, J. M. Waste lubricating oil rerefining by
(metallic compounds, oxidation products, and carbonaceous extraction-flocculation. 1. A scientific basis to design efficient solvents.
particles) could be removed. Ind. Eng. Chem. Res. 1988, 27, 1222−1228.
(6) Alves dos Reis, M.; Silva, J. M. Waste lubricating oil rerefining by
5. CONCLUSIONS extraction-flocculation. 2. A method to formulate efficient composite
By using the proper selection of components, a composite solvents. Ind. Eng. Chem. Res. 1990, 29, 432−436.
(7) Reis, M. A.; Jerónimo, M. S. “Regeneracao̅ de óleos lubrificantes
solvent and flocculant recovering base oil from waste lubricant usados por extraccao-floculaca
oil were formulated. The alcohol with four carbon atoms (n- ̅ ̅
o”. Addition to Portuguese Patent
69392, October 1982a.
butanol, i-butanol) was an extractant which was miscible with (8) Reis, M. A.; Jerónimo, M. S. Fabrico de tinta de base para fabrico
the base oil; the MEA was a flocculant which improved the de tinta de impressao̅ a partir de óleos usados. Portuguese Patent
flocculation and segregation of the waste oil impurities. The 75702, October 1982b.
best regeneration conditions were obtained as follows: refining (9) Alves dos Reis, M.; Silva, J. M. Waste lubricating oil re-refining by
time, 20 min; refining temperature, 30 °C; solvent/waste oil extraction-flocculation. 3. A pilot plant study. Ind. Eng. Chem. Res.
mass ratios, 5g/g; flocculant concentration, 2g/kg. 1991, 30, 2449−2456.
When i-butanol/MEA and n-butanol/MEA were compared, (10) Mo, Y. N.; Guo, D. G.; Zhang, Y. X. Lubricating oil recovery and
yields obtained with a composite of i-butanol/MEA are slightly regeneration by solvent refining. Chem. Eng. Oil Gas 2007, 36, 124−
134.
higher than those obtained with n-butanol/MEA. Maybe the (11) Martins, J. P. The extraction−flocculation re-refining lubricating
polarity and molecular structure of i-butanol is similar to those oil process using ternary organic solvents. Ind. Eng. Chem. Res. 1997,
of the base oil molecules. The solvent/oil mass ratios at which 36, 3854−3858.
stabilization occurs with butanol/MEA are smaller than that (12) Hamad, A.; Al-Zubaidya, E.; Fayed, M. E. Used lubricating oil
obtained with single pure solvent (n-butanol, i-butanol), which recycling using hydrocarbon solvents. Environ. Manage. 2005, 74,
indicates that MEA not only flocculates waste oil contaminants 153−159.

12769 dx.doi.org/10.1021/ie4015099 | Ind. Eng. Chem. Res. 2013, 52, 12763−12770

Industrial & Engineering Chemistry Research Article

(13) Hildebrand, J. H.; Scott, R. L. The Solubility of Nonelectrolytes,

3rd ed.; Dover: New York, 1964.
(14) McClellan, A. L. Tables of Experimental Dipole Moments; W. H.
Freeman and Company: San Francisco and London, 1963.
(15) Tripathy, T.; Ranjan, De, B. Flocculation: A new way to treat
the waste water. J. Phys. Sci. 2006, 10, 93−127.
(16) Brostow, W.; Hagg Lobland, H. E.; Pal, S.; Singh, R. P.
Polymeric flocculants for wastewater and industrial effluent treatment.
J. Mater. Educ. 2009, 31 (3−4), 157−166.
(17) Brownawell, D.; Renard, R. Re-refining of used oil. US Patent
3,639,229, 1972.
(18) Whisman, M. L.; Reynolds, J. W.; Goetzinger, J. W.; Cotton, F.
O.; Brinkman, D. W. Method for reclaiming waste lubricating oil. U.S.
Patent 4,073,720, 1978.
(19) Zhang, X. M.; Jiao, Z. J.; Li, C.; Jia, Y. M.; Zhang, C. M.
Regeneration of waste lubricating oil by flocculation-clay composite.
Environ. Eng. 2008, 26, 47−49.
(20) Xiong, W.; Xiong, D. L. Research of recovering conditions of
waste lubricating with new flocculants. JiangXi Sci. 2009, 27, 356−359.
(21) Si, T. Y.; Luo, J. Y. Basic principles and application on solubility
parameter; Academy of Art, Shanghai University: Shanghai, 2007.
(22) Smallwood, I. A. Handbook of Organic Solvent Properties; Arnold:
London, 1996.
(23) Manfredi, L.; Hill, R. J.; van de Ven, T. G. M. Bridging
flocculation of PEI-functionalized latex particles using nanocrystalline
cellulose. J. Colloid Interface Sci. 2011, 360, 117−123.
(24) Yasarla, L. R.; Ramarao, B. V. Dynamics of flocculation of
lignocellulosic hydrolyzates by polymers. Ind. Eng. Chem. Res. 2012, 51,
6847−6861.
(25) Furusawa, K.; Ueda, M.; Nashima, T. Bridging and depletion
flocculation of synthetic lattices induced by polyelectrolytes. Colloids
Surf., A 1999, 153, 575−581.
(26) La Mer, V. K.; Healy, T. W. Adsorption−flocculation reactions
of macromolecules at the solid liquid interface. Rev. Pure Appl. Chem.
1963, 13, 112−133.
(27) La Mer, V. K.; Healy, T. W. Specification of materials for
retarding water evaporation-spreading hexadecanol monolayers. Science
1964, 144, 564.
(28) Dixon, J. K.; La Mer, V. K.; Linford, H. B. Factors affecting
filtration rates of flocculated silica. J. Water Pollut. Control Fed. 1967,
39, 647−658.

12770 dx.doi.org/10.1021/ie4015099 | Ind. Eng. Chem. Res. 2013, 52, 12763−12770