Separation and Purication Technology 133 (2014) 227235

Contents lists available at ScienceDirect

Separation and Purication Technology

journal homepage: www.elsevier.com/locate/seppur

Partial purication of iron solutions from ripe table olive processing

using ozone and electro-coagulation
Pedro Garca-Garca , Francisco No Arroyo-Lpez, Francisco Rodrguez-Gmez
Food Biotechnology Department, Instituto de la Grasa (CSIC), Padre Garca Tejero 4, 41012 Sevilla, Spain

a r t i c l e

i n f o

Article history:
Received 3 December 2013
Received in revised form 4 June 2014
Accepted 6 June 2014
Available online 13 June 2014
Keywords:
Contamination removal
Electro-coagulation
Ferrous solution
Ozonation
Ripe olive

a b s t r a c t
This study investigates the application of electro-coagulation and ozonation technologies for the partial
depuration of ferrous solutions deriving from the color xation stage of ripe olive processing. Different
operational conditions were investigated along with the possibilities of combining the two techniques
sequentially. In electro-coagulation the best depuration conditions were obtained using a current density
of 25 mA cm2 for 40 min; COD elimination reached 30%, the phenols and color of the solutions were also
considerably reduced and the dissolved Fe was completely removed. On the contrary, the microbial population mainly composed of lactic acid bacteria and yeasts, hardly decreased. The application of ozonation also reduced the total phenols, color and COD of solutions, but did not decrease the iron
concentration. However, it considerably affected the microbial population in a higher proportion than
electro-coagulation. In light of the results obtained, the best working conditions consist of applying
40 min electro-coagulation with aluminum in the anode and iron in the cathode at 25 mA cm2 current
density, followed by the storage of a partially puried solution bubbling ozone to reduce the microorganism population and even eliminate it completely when 6 g O3 L1 were added. This would allow for a possible reuse of these solutions for preparing new xing color solutions or as cover brine in packaging.
2014 Elsevier B.V. All rights reserved.

1. Introduction
Olives darkened by oxidation [1] are commonly known by their
original American name: ripe olives. The production of this preparation was reached around 630,000 tons in the 2011/2012 season,
30% of the worlds table olive production [2].
The darkening process consists of successive treatments of the
fruits with a dilute solution of NaOH (lye); during the intervals
between lye treatments the fruits are suspended in water through
which air is bubbled [3]. Nowadays, a preservation solution is
added to the tap water in this phase of oxidation [4]. Throughout
this operation the fruits darken progressively, but the color formed
is not stable permanent and fades progressively after oxidation; to
prevent this deterioration, the fruits are immersed in ferrous lactate or gluconate for several hours [5]. The product has a nal pH
above 4.6 and its preservation is only achieved by sterilization [3].
Electro-coagulation (EC) is an electrochemical method consisting in treating polluted water whereby sacricial anodes dissolve
to produce active coagulant precursors (usually aluminum or iron
Corresponding author. Address: Food Biotechnology Department, Instituto de la
Grasa (CSIC), Avda. Padre Garca Tejero 4, 41012 Sevilla, Spain. Tel.: +34 954 690
850; fax: +34 954 691 262.
E-mail address: pedrog@cica.es (P. Garca-Garca).
http://dx.doi.org/10.1016/j.seppur.2014.06.011
1383-5866/ 2014 Elsevier B.V. All rights reserved.

cations) in the solution. The generation of coagulants in situ means

that EC process does not require the addition of any chemicals. The
gases produced at the cathode during the electrolysis of water and
metal dissolution allow the resulting ocks to oat [6].
EC has been successfully used in the treatment of wastewaters
from food industries such as alcohol distillery [7], dairy [8], olive
oil mill [9] and oil renery [10]. In green table olive processing,
the combined biological and EC treatments of the debittering solution (lye) and washing water reduce the chemical oxygen demand
(COD) up to 97% [11]. Also, EC application to wastewaters from the
packaging industry results in a practically colorless solution of lactic and acetic salts which could be reused in packing [12]. In natural black table olives, the EC of wastewaters by boron-doped
diamond electrodes led to 73% COD removal [13]. However, this
technology has not yet been studied for the treatment of efuents
from the darkening process of ripe olive industries.
Ozone has been declared Generally Recognized as Safe (GRAS)
and it was admitted as a direct food additive [14,15]. The treatment of wastewaters with ozone from the table olive industry
[16] and olive oil extraction process [17] lightened the nal color
of the solutions due to destruction of the brownish polyphenol
polymers.
The green olive alkaline solutions (lye and washing waters)
treated with ozone may be reused as fermentation brines after

228

P. Garca-Garca et al. / Separation and Purication Technology 133 (2014) 227235

adding the appropriate NaCl quantity [18,19]. Also, Spanish green

olives may be packed with diluted (1:1) ozonated fermentation
brines with similar organoleptic characteristics to those using fresh
brine [20]. In ripe olives, preservation solutions can be reused in a
new preservation cycle after ozone treatment, the resulting quality
of the nal product being similar to the ones stored in a fresh acid
solution [4].
Color xation solutions in the darkening process could be
reused after the addition of the required quantity of ferrous gluconate if the organic matter, mainly phenols, is removed. These compounds can be chemically bonded to the ferrous ions when adding
more ferrous gluconate and preventing the cations penetrating
into the olives, forming the Fephenol complexes in the liquid
and the color of olive not being xed [5].
In the present work, the use of EC and ozonation to treat ferrous
gluconate solutions from ripe olive processing was studied. Particularly, the effect of various operating conditions (current intensity
and contact time of solutions in EC or amount of O3 bubbled) on
the elimination of COD, iron, phenols, color and microorganism
survival was investigated. This information is essential both to
assess the possibility of reusing these solutions by the industry
and also for scaling up the treatment at the industrial level.

between the lowest part of the electrode and the cell was 0.8 cm
and between electrodes was 1 cm; the total effective surface area
of the electrodes was 91 cm2. Prior to each experiment; the electrodes were immersed in 1% HCl for 12 h and rinsed with distilled
water.
The volume of liquid treated each time was 1.6 L. The liquid was
magnetically stirred by two bars, allowing a correct homogenization of the wastewater. A direct current was imposed by a stabilized power supply (Quasar 500, CRS Industrial Power
Equipment, Calco, Lecco, Italy) for 1 h and the range of current density variations was 1075 mA cm2. The temperature was monitored with a thermometer and maintained stable (2025 C)
during the experiments by cooling the solution with a cooling
jacket. All experiments were performed in duplicate and samples
(10 mL) were withdrawn at 5 min intervals. Before their analysis,
the samples were centrifuged to remove any suspended solid.
The electrical energy consumption per unit mass of organic load
removed (SEEC) was calculated using the following equation:

SEEC VIt=C 0  CQ

where V, voltage in volts; I, current in amperes; t, time in hours; Q,

volume of wastewater in liters; and (C0  C), pollutant load
removed in g O2/l.

2. Experimental
2.3. Treatment with ozone
2.1. Ferrous solutions
Ferrous gluconate solution samples were collected for the darkening processes of ripe olives carried out at the Instituto de la
Grasa pilot plant (CSIC, Spain) similarly to what industrial processes are [3]: relationship olives/liquid, 1/1; lye treatment (2.5%
NaOH, 5 h to reach stone); rst washing only tap water (20 h); second washing (tap water/preservation solution, 1/1, 24 h), ferrous
gluconate solution (0.1%, 20 h).
The initial physico-chemical and microbiological characteristics
of the solutions are shown in Table 1. The high population of
microorganisms, especially lactic acid bacteria (LAB) and yeasts
were found in these solutions due to the preservative liquid reuse
in washing during oxidation [4]. This is the rst time that this fact
is mentioned in a publication.

The ozone was produced by TODOZONO1, Mod TD ZN equipment (Colmenar Viejo, Madrid, Spain). Aliquots of 2.0 L of solution
were put in a cylindrical test tube (52 cm  8.5 cm diam.). The
solutions were then treated by bubbling ozonated air with an
ozone concentration of 3.63 mg L1 at a ow rate of 200 L/h
(3.63 g O3 h1 L1 of solution) through a synthetic glass diffuser
introduced at the bottom of the tube. The pH of the solution was
allowed to evolve freely. All experiments were performed in duplicate and samples (10 mL) were withdrawn at 1 h intervals.
The electrical energy consumption was estimated from the
power consumption indicated by the manufacturer of the ozoneproducing equipment (50 w) and the time (h) to produce a given
amount of ozone in g (Q) per unit volume of solution (L).

EkW h=L 50  103  Q  1=3:63 0:0137  Q

2.2. Electro-coagulation procedure

2.4. Experimental design
The EC was carried out in a plexiglas parallelepiped reactor
(2.0 L). Two pairs of Al/Fe (anode/cathode) electrodes connected
in a bipolar mode were placed in the electrochemical reactor, each
one with dimensions of 13 cm  6.5 cm  0.2 cm; the distance

The effect of various EC operating conditions (current intensity

and contact time) and bubbled O3 quantity during ozonation treatment on the elimination of COD, iron, phenols and color, was

Table 1
Initial characteristics of ferrous gluconate solutions and after tested treatments to remove pollution. Electric power and aluminum consumptions.
Treatments

Initial
ECb,c (40 min)
Ozoneb (5 g O3 L1)
Ozone + ECb,c
(5 g O3 L1) + (30 min)d
ECc + Ozoneb
(40 min) + (6 g O3 L1)d
a
b
c
d
e
f

Total
phenols
(mg L1)

Color
(A440A700)

4700(88) a
3300 (94) c
4200 (97) b
3190 (99) c

171 (9) a
80 (7) b
55 (8) c
30 (12) d

1.05 (0.08) a
0.27 (0.05)c
0.46 (0.04) b
0 d

3200 (95) c

15 (5) d

COD
(mg O2 L1)

Iron
(mg L1)

Microorganism population
(log10 CFU mL1)
LAB

0 d

51 (5)a
0b
51 (4)a
0b
0b

6.5
6.3
5.2
4.5
0d

Yeast
(0.3)
(0.2)
(0.3)
(0.2)

a
a
b
c

5.5
5.3
4.9
4.4

(0.3)
(0.1)
(0.1)
(0.2)

a
ab
b
c

0d

Averages of samples collected from 6 different darkening process.

Averages of two experiments.
EC at 25 mA cm2 current density.
Time (min) by EC and bubbled O3 (g L1) by ozonation.
Standard deviation in parenthesis.
Column values followed by the same letter do not differ at the 5% level of signicance according to Duncans multiple-range test.

Electric
consumption
(kW h m3)

Consumed
aluminum
(g m3)

6.5
68.5
68.5 + 7

0.89

2.5

6.5 + 82.2

0.89

229

P. Garca-Garca et al. / Separation and Purication Technology 133 (2014) 227235

The COD analysis was carried out according to the procedure

described in the Standard Methods of APHAAWWAWPCF [21].
The appropriate amount of sample was placed with potassium
dichromate and sulfuric acid with mercuric sulfate solutions in a
sealed test tube and incubated for 2 h at 150 C in a COD reactor
(Selecta, Barcelona, Spain). COD was measured by titration with a
ferrous sulfate solution.
Total polyphenols were determined by the FolinCiocalteau
method [22] and the iron concentration in liquid was determined
by ame atomic absorption spectrometry [23].
The color of solutions was calculated as the difference in absorbance at 440 and 700 nm (A440A700), using a ber optics probe in a
Varian Cary 60 UVVis Spectrophotometer (Agilent Technologies,
Malgrave, Vi, Australia). Previously, the liquids were centrifuged
at 12000g for 5 min [24].
All physico-chemical analyses were carried out at least in
duplicate.

10

COD (mg O2 L-1 )

2.5. Physico-chemical analysis

4800

2.7. Statistics
Statistica version 6.0 (StatSoft, Tulsa, USA) for windows was
used for data analysis. Comparisons of the average values were
performed by one-way analysis of variance (ANOVA) followed by
the multiple Duncans multiple range test (p < 0.05).
3. Results and discussion
3.1. Effects of EC on ferrous solutions
3.1.1. Inuence on physico-chemical parameters and chemistry of the
process
Fig. 1 shows the inuence of different current densities on the
evolution of COD, polyphenols and iron dissolved in the ferrous
gluconate solutions during the application of EC. A decrease in
the COD content occurred in all EC processes at the different current densities assayed (Fig. 1A). Increasing current density led to
an increase in the COD reduction rate and less nal COD values
after 1 h of treatment were obtained at the highest current density
values.

50

75

4000
3600
3200
2800

10

20

30

40

50

60

TIME (min)
10

160

25

50

140

75

120
100
80
60

10

20

30

40

50

60

TIME (min)

2.6. Microbiological analysis

50

10

25

50

75

40

Fe (mg L-1 )

Ferrous gluconate solutions or their decimal dilutions were plated using a Spiral System model dwScientic (Don Whitley Scientic Limited, England) on the appropriate medium described
below.
Lactic acid bacteria (LAB) were plated on de Man, Rogosa and
Sharpe (MRS) agar media (Merck KGaA, Darmstadt, Germany) supplemented with 0.02% (w/v) sodium azide (Sigma, St. Louis, MO,
USA), and yeasts on a yeastmaltpeptoneglucose medium (YM
agar, DifcoTM, Becton and Dickinson Company, Sparks, MD, USA)
supplemented with 0.005% (w/v) gentamicin sulfate and oxytetracycline antibiotics (Oxoid Ltd., Basingstoke, Hampshire, England).
The plates were incubated at 30 C for 48 h, counted using a CounterMat v.3.10 (IUL, Barcelona, Spain) image analysis system and
expressed as log10 CFU/mL.

25

4400

TOTAL PHENOLS (mg L-1)

investigated in a rst step, in order to know the best individual

treatment conditions. Then, using the best conditions obtained
for both treatments (ozonation and EC), additional experiments
were performed alternating the order of operations, rst EC and
then ozonation and vice versa.

30

20
10
0

10

20

30

40

50

60

TIME (min)
Fig. 1. COD, total phenols and iron removal in the ferrous gluconate solution from
ripe olive processing during EC treatment, using different current densities with an
electrode sets of Al (anode)/Fe (cathode).

The concentration of total phenols in the solutions followed a

similar trend to that observed for the COD. When current density
increased the phenol contents in the solutions decreased. The nal
concentrations obtained in the gluconate solutions ranged from 62
to 88 mg L1 (Fig. 1B).
The iron concentration also decreased with time and increasing
current density led to an increase in the iron reduction rate and
less time was required to achieve dissolved metal disappearance.
With the lower tested current density (10 mA cm2) all dissolved
iron disappeared in 40 min (Fig. 1C).
Thus, removal efciency increased signicantly as the current
density increased. The highest current densities produced the fastest treatment with the lowest values of COD, total polyphenol concentration and dissolved iron. This behavior is ascribed to the fact
that, at a high current density, the amount of oxidized metal
increases, resulting in a greater amount of precipitate for the
removal of more pollutants [10]. In addition, bubble size decreases
with increasing current densities and bubble density increases
[25], resulting in a greater upwards ux and a faster removal of
pollutants and sludge otation.

P. Garca-Garca et al. / Separation and Purication Technology 133 (2014) 227235

the initial formation of more colored polymers, as was also noticed

during the oxidation phase of ripe olives [29] and in the ozonation
of green olive brines [20]. Later on, the polymers and phenols were
apparently oxidized and their disappearance (Fig. 1B) produced the
subsequent brine discoloration, which led to a decrease in the difference in absorbance (A440A700) (Fig. 2B).

pH
8.5

8
7.5
7
6.5
6

10

10

20

25

30

50

40

3.1.2. Electrode consumption

Information on electrode consumption is essential for estimating the cost of the EC process because the dissolution rate affects
the life of the electrodes. Theoretically, according to Faradays
law, whenever one Faraday of charge passes through the circuit,
9.0 g of aluminum is dissolved at the anode. However, the effective
electrode consumption may be reduced or increased from this theoretical value depending on the wastewater characteristics and
operational conditions.
The aluminum consumption (y) was linearly related (R2 > 0.95)
to the increase in current density (x) and it was higher (y = 0.079 x)
than theoretically estimated values (y = 0.061 x) and slightly
greater than in green table olive wastewaters EC (y = 0.074 x) [12].
The higher than expected dissolution of aluminum on the anode
is due to the electrochemical process; a reduction in water can modify the pH on the anode with respect to the pH in the rest of the solution [9]. In the cathode, a decrease in iron weight was not observed.

75

50

60

TIME (min)

COLOR (A440 -A700 )

1.2

10

25

50

75

1.0

0.8
0.6
0.4
0.2
0.0

10

20

30

40

50

60

TIME (min)
Fig. 2. Changes in pH and color in the ferrous gluconate solution from ripe olive
processing during EC treatment, using different current densities with an electrode
set of Al (anode)/Fe (cathode).

During the EC treatment, a gradual increase in the pH of the

solution was observed (Fig. 2A). The values of pH after 1 h ranged
from 8.0 to 8.9, with higher nal values with increasing current
density.
A complete examination of what happens in the EC process
shows the following electrochemical and chemical reactions at
anodes (3), (4), both in the bulk solution (5) and in the cathode
(4) [8,10,26]:
Al-anode:
Ze

Z0 e

OC ! OCx ! Final
3

Al  3e Al

3
4

3.1.3. Efciency and cost of process

The COD removal per theoretical unit of electrode consumption
decreases with treatment time (Fig. 3A). The best COD removal
versus electrode consumption was obtained at the lowest current
density (10 mA cm2). Using higher current densities, efciency
values were lower.
On the contrary, the electrical energy consumption per unit
mass of organic load removed (SEEC) increased, as the current

Solution:
3

Al

3H2 O AlOH3 3H

10

5.0

Kg DQO g -1 Al cons.

230

25

50

4.0

75

3.0
2.0
1.0
0.0

10

20

30

40

50

60

Time (min)

Cathode:

The anodic oxidation of organic compounds (3) is the most

important reaction that should be produced in the EC of the ferrous
gluconate solution [28]. The indirect oxidation of organic matter
can hardly occur, mainly phenols by the action of the hypochlorite
eventually formed from the NaCl dissolved in the wastewater
[27,28] because today in the industrial process NaCl is not used,
the preservation of the olives is carried out with organic acid solutions [3,4].
The reactions in the cathode (6) produce OH and other anions
such as Cl, and HCO
3 . which are dissolved in the wastewater can
exchange partly with OH- in Al(OH)3, to liberate free OH, which
causes an increase in pH.
At the beginning of the EC treatment, the liquid became progressively darker, as evidenced by the increase in the (A440A700)
parameter (Fig. 2B). This fact has also been observed in the wastewater EC of green olive industries [12]. This effect may be due to

10

35

6
SEEC (kWh kg -1 COD)

3H2 O 3e 3=2H2 g 3OH

25

50

75

30
25

20
15
10
5
0

10

20

30

40

50

60

TIME (min)
Fig. 3. Changes in COD removal per unit of aluminum electrode consumption
(anode) (A) and specic electrical energy consumption (SEEC) (B) during the EC
treatment of ferrous gluconate solution from ripe olive processing, using different
current densities with electrode sets of Al (anode)/Fe (cathode).

P. Garca-Garca et al. / Separation and Purication Technology 133 (2014) 227235

density was higher (Fig. 3B), due to a proportional increase in

ohmic voltage loss in the cell. The energy efciency was higher
(lower SEEC) when the current density was 10 mA cm2. Fig. 3B
shows that the specic power consumption is very similar when
working at 10 and 25 mA cm2. However, the nal value of COD
was signicantly lower when operating at 25 mA cm2 (Fig. 1A).
Thus, it can be concluded that after the 40 min treatment, the
COD and the color of the solution (Fig. 2B) remained unchanged
and the dissolved Fe was completely removed (Fig. 1C).
Then, with a 25 mA cm2 current density and 40 min treatment,
the power consumption expected would be around 5 kW h/kg COD
(Fig. 3B), which means that for one cubic meter of wastewater and
a COD reduction of around 1300 mg O2 L1 (Fig. 1A), the total consumed electricity should be 6.5 kW h and the theoretical sacricial
electrode loss 0.89 g of aluminum.
3.1.4. Inuence on microbial population
The LAB and yeast population slightly decreased during EC at
25 mA cm2 current density (Fig. 4A). After the 40 min treatment,
the LAB population only decreased about a third (0.25 log10
cycles). There is a slightly greater inuence on yeasts and the population was reduced by almost 60% in the same period of time
(0.5 log10 cycles). Continuous EC involves the reduction of the
microbial population.
The inactivation of LAB and yeast cells by electrochemical
means has been well documented [30]. The cells were probably
inactivated during the EC process by the chemical oxidizing products and the short-lived germicidal agents produced, such as free
radicals [31], but the performance of the process strongly depends
on the characteristics of the initial efuent (microorganism population, organic load, etc.) and the type of electrodes used [32].
The low inactivation of bacteria and yeast observed in this study
may be related to the high organic loading presented by the ferrous
solution and the fact that EC is more effective at inactivating

log10 CFU mL-1

6.5

LAB

Yeast

EC
0

10

20

30

40

50

60

TIME (min)
5

log10 CFU mL-1

3.2. Effects of ozonation on ferrous solutions

3.2.1. Inuence on physico-chemical parameters and treatment cost
Ozone treatment decreased the concentration of phenols and
the color of the solutions (Fig. 5A and B). After bubbling 6 g O3 L1,
both parameters had lower values than under the best conditions
in EC (Figs. 1B, and 2B).
The ozonation of phenols implies diverse breakdown reactions
that give rise to short chain carboxylic acids [34] producing a gradual decrease in pH values (Fig. 5C). A faster pH decrease which
dropped the pH to about 5.3 was observed after the added ozone
reached 4.0 g L1, to then decrease slowly to reach a value of 5.2
when 10 g O3 L1 were added.
Bubbling ozone produced a decrease in COD until 5 g O3 L1
were added. Adding more ozone does not involve a signicant
decrease in the pollutant load (Fig. 5D). Finally, a decrease in the
iron concentration due to the effect of bubbling ozonated air was
not observed (data not shown).
The cost of the process is only due to the electrical energy consumption. Fig. 5A, B and D show that after adding 5 g O3 L1, the
total phenols, color of the solution and COD were practically
unchanged. Therefore, for one cubic meter of solution, the total
consumed electricity should be 68.5 kW h.

3.2.2. Effect on microbial population

The ozone also produced a decrease in the microbiological population, as reported by other authors using different matrices
[15,20]. After adding 6 g O3 L1. the population was reduced by
about 1.01.5 log10 cycles for both LAB and yeasts (Fig. 6A).
The residual free ozone not used in phenol oxidation could have
reacted earlier with the cellular membrane of the microorganisms
and cause their initial death [15]. This was also observed when
ozone is bubbled into the fermentation brines of green olives [20].

3.3. Ozonation of ferrous solutions followed by EC

5.5

LAB

YEAST

3
2

OZONATION
1
0

microbes at lower population densities than at higher ones [33],

as occurred in this work.

6.0

5.0

231

O3 (g/L)
Fig. 4. Lactic acid bacteria (LAB) and yeast population counts during the EC
treatment of ferrous gluconate solutions using 25 mA cm2 current density and Al
(anode)/Fe (cathode) electrode sets (Fig. 4A), and after the subsequent treatment of
solutions with ozone (Fig. 4B).

The previous experiments allowed us to determine the best conditions for the individual application of EC and ozonation to the ferrous solutions. Consequently, the following step consisted in
assessing the inuence of the sequential application of both techniques. Thus, solutions previously treated with ozone (5 g O3 L1)
were then submitted to an EC procedure with 25 mA cm2 current
density. The main results obtained from EC are shown below.

3.3.1. Physico-chemical changes

A decrease in pollutant load was observed during the rst
30 min of EC treatment (Fig. 7A); later, unchanged COD and the
nal value was slightly lower than when no pretreatment with
ozone was carried out (Fig. 1A). The iron concentration decrease
rate (Fig. 7B) was higher than when no pretreatment with ozone
was carried out (Fig. 1C) and the dissolved metal disappeared completely in 30 min. An increase in the pH of the solutions was also
observed (Fig. 7C); after 30 min of EC, the pH reached about 7.0
units.
The total phenol concentration was reduced from 5060 mg L1
(Fig. 5A) to about 30 mg L1 in 30 min; also, a nearly complete discoloration of the solution occurred (data not shown).
In accordance with the above mentioned results, it can be concluded that the purication process nished after 30 min EC in previously ozonated iron solutions with 5 g O3 L1.

P. Garca-Garca et al. / Separation and Purication Technology 133 (2014) 227235

-1

TOTAL PHENOLS (mg L )

232

160

1.2

140
100
80
60

0.8
0.6
0.4
0.2

40
20

1.0

(A440 -A700 )

120

0.0

10

O3 (g/L)

5.6
5.4
5.2
0

COD (mg O2 L-1)

5.8

5.0

10

4800

6.0

pH

O3 (g/L)

4400
4200
4000

10

O3 (g/L)

4600

10

O3 (g/L)

Fig. 5. Total phenols, color and COD removal as pH changes in the ferrous gluconate solution from ripe olive processing during ozonation. (For interpretation of the references
to color in this gure legend, the reader is referred to the web version of this article.)

LAB

log10 CFU mL-1

6.5

YEAST

6
5.5
5

OZONATION

4.5
4

O3 (g/L)
5.0

log10 CFU mL-1

LAB

Yeast

4.5

4.0

EC
3.5

10

20

30

40

50

60

TIME (min)
Fig. 6. Lactic acid bacteria (LAB) and yeast population counts during the treatment
with ozone of ferrous gluconate solutions (Fig. 6A) and after the subsequent EC
treatment of solutions using 25 mA cm2 current density and Al (anode)/Fe
(cathode) electrode sets (Fig. 6B).

When EC extended for over 30 min, a greater inactivation rate

of microorganisms was produced than in both the rst 30 min
(Fig. 6B) and also when EC was applied to the solution without
any pretreatment (Fig. 4A). This may be related to both lower
organic load and lower population density, which makes EC more
effective in killing the microorganisms [32,33].
3.3.3. Efciency and cost of process
After ozonation, the initial COD removal per theoretical unit of
electrode consumption (Fig. 8A) was higher (about three times)
than when no pretreatment was performed (Fig. 3A) with the same
current density (25 mA cm2). After 30 min of EC treatment COD
removal per g of consumed aluminum was also higher (around
2.5 kg DQO g1 Al consumed) than when no pretreatment was carried out (1.7 kg DQO g1 Al consumed).
Also, at the same current density (25 mA cm2), the electrical
energy consumption per unit mass of organic load removed (SEEC)
was higher in solutions with a previous ozonated treatment
(Fig. 8B) than when no pretreatment was performed (Fig. 3B).
Therefore, the power consumption expected after 30 min of EC
would be around 7 kW h/kg COD (Fig. 8B), which means that, for
one cubic meter of wastewater and a COD reduction of around
1000 mg O2 L1 (Fig. 7A), the total consumed electricity should
be 7 kW h. The expected sacricial electrode loss is around 2.5 g
of aluminum m3.
The previous cost of the electricity needed for producing
5 g O3 L1, 68.5 kW h/m3 of solution, must be added.
3.4. EC of ferrous solutions followed by ozonation
Solutions previously treated with EC (40 min at 25 mA cm2
current density) were ltered and then submitted to ozonation.

3.3.2. Effect on microbial population

EC at 25 mA cm2 current density after ozonation of solutions
also produced a decrease in the microbial population. After
30 min the LAB population was reduced by about 0.5 log10 cycles
and the yeast population hardly decreased (Fig. 6B).

3.4.1. Effect on physico-chemical parameters

The application of ozone (2 g L1) produced a phenol concentration reduction from about 80 mg L1 to one fourth of the amount
(data not shown); it is necessary to bubble 7 g O3 L1 to reach

233

P. Garca-Garca et al. / Separation and Purication Technology 133 (2014) 227235

kg DQO g-1 Al cons.

4300

COD (mg O2 L-1 )

4000

3700
3400
3100
2800

4
2

A
0

10

20

30

40

50

60

40

50

60

TIME (min)
0

10

20

30

40

50

60
12

SEEC (kWh kg-1 COD)

50
40

Fe (mg L-1 )

TIME (min)

30

20

10
8
6
4
2

10

20

30

TIME (min)

10
0

10

20

30

40

50

60

TIME (min)

Fig. 8. Changes in COD removal per unit of aluminum electrode consumption

(anode) (A) and Specic electrical energy consumption (SEEC) (B) during the EC
treatment of the ferrous gluconate solution from ripe olive processing after
previous treatment with 5 g ozone per liter. Electrode sets of Al (anode)/Fe
(cathode) and current density 25 mA cm2.

pH
The reduction in the microorganism population, and even its
complete elimination, may prevent secondary fermentations that
could lead to the formation of both strange smells and tastes that
would make the reuse of these liquids unviable in ripe table olive
processing.

C
6

5
0

10

20

30

40

50

60

TIME (min)
Fig. 7. COD and iron removal as pH changes in the ferrous gluconate solution from
ripe olive processing during EC treatment after previous treatment with 5 g ozone
per liter. Electrode sets of Al (anode)/Fe (cathode) and current density 25 mA cm2.

values below 10 mg L1. At the same time, a nearly complete discoloration of the solution occurs in addition to a slight decrease
in the pH with a nal value near neutrality (7.0). The COD of the
solution decreases only 100 mg L1 after bubbling 10 g O3 L1 (data
not shown).
Therefore, the ozonation of the gluconate solutions after EC
does not signicantly decrease the pollution load and only the
removal phenols and discoloration occurred.
3.4.2. Effect on microbial population
Ozonated air bubbling after EC treatment reduced the microorganism population and a complete destruction of the LAB was
achieved after the total bubbled ozone reached 4.0 g O3 L1
(Fig. 4B). Yeasts were more resistant to the applied ozone than
LAB and at least 5.8 g O3 L1 must be applied to achieve total elimination. This different effect of ozone on microbial inactivation was
also observed in the ozonation of green table olive brine [20].

3.4.3. Efciency and cost of process

Previously it has been shown that the power consumption of
the EC for 40 min was 6.5 kW h m3 of solution and the theoretical
sacricial electrode loss was 0.89 g of aluminum. To this previous
cost, the electricity need for bubbling 6 g O3 L1 of ferrous solution
to kill all microorganisms must be added, 82.2 kW h m3.
Alternatively, the solution could be used directly without ozone,
reducing this way the expenditure. When the solution is stored
temporarily for later reuse, bubbling ozone would be convenient
in order to limit the possible development of microorganisms
and even their complete elimination.
3.5. Choice of the purication procedure
When the solutions were treated with ozone alone, the worst
results in removing pollution were obtained: highest values of
COD and color, without removing any ferrous ion and maintaining
a high population of microorganisms. On top of this, the energy
cost was very high (Table 1).
Statistically (p < 0.05) same nal COD values (around 3200
3300 mg O2 L1) were obtained when performing only EC with
Al(anode)/Fe(cathode) electrodes at 25 mA cm2 current density
for 40 min and combined with ozonation regardless the order in
which the treatments were performed (Table 1). However, the
use of ozone results in removing color and also the lowest nal
concentration of phenols.
The order of application of treatments inuences the nal population of microorganisms and, also, in aluminum and energetic
consumptions. Ozonation performed rst and then the EC

234

P. Garca-Garca et al. / Separation and Purication Technology 133 (2014) 227235

(30 min at 25 mA cm2) reduced the population of LAB in 2 log10

cycles while the yeast in 1.1 log10 cycles (Fig. 6, and Table 1); however, in the reverse order, it is possible to eliminate all microorganisms (Table 1, and Fig. 4). In the latter case the aluminum
consumed at the anode was lower (0.89 g/m3) due to the increased
effectiveness of the EC when the liquid to be treated had a higher
organic load, although the energy consumption was greater
(88.7 kW h m3) than when ozone was rst bubbled and the EC
was performed after (75.5 kW h m3).
The rst application of the EC (40 min at 25 mA cm2) eliminated the dissolved iron. Additionally, organic load, phenol concentration and color of the solution, decreased. A priori, this partially
puried solution could be reused after the addition of the required
quantity of ferrous gluconate in new xation treatments, since the
concentration of phenols and organic load have decreased, thereby
partly preventing ions ferrous bonded to these compounds from
penetrating into the olives, not properly xing the color [5].
In the industries, the partially puried solutions must be stored
for a range time (224 h) until possible reuse. During storage after
EC, because of the population of microorganisms is present in these
solutions, fermentations can occur producing unusual smells and
tastes [3]. To prevent this, the bubbling of ozone can control the
population of microorganisms [15] and even eliminate it completely (Fig. 4B), same as in the ozonation of Spanish green olive
fermentation brines [20].
According to the results presented here, the best method for
partially purifying color xing solutions in ripe olive processing
consists in performing an EC with Al(anode)/Fe(cathode) electrodes set at 25 mA cm2 current density for 40 min and storing
the obtained partially puried solution bubbling ozone until further reuse.

4. Conclusions
In this study, the application of EC and ozonation for the regeneration of the iron solutions from the ripe table olive processing
industry was assessed and the following results were obtained:
(1) The EC with aluminum in the anode and iron in the cathode
removed all dissolved iron, most of the initial phenols in the
solution and about 30% of the initial pollutant load. The LAB
and yeast population slightly decreased during EC. The
chemical depuration efciency increased signicantly as
the current density increased.
(2) The bubbling of ozone produces a decrease in total phenols, a
discoloration of the solution but only an initial limited reduction of COD and without removing any iron ions. The ozone
bubbling also produced a decrease in the microbiological population. The energy cost of ozonation is much higher than EC.
(3) Combined EC and ozonation, regardless of the order in which
treatments are performed, produces a colorless solution,
with lower phenol concentration and pollutant load than
when only EC or ozonation are applied. Ozonation reduces
the population of LAB and yeast faster than EC. When it is
applied after EC, it is possible to eliminate all
microorganisms.
(4) The best working conditions consist in applying 40 min EC
with aluminum in anode and iron in the cathode at
25 mA cm2 current density, followed by storage of the partially puried solution after bubbling ozone to reduce the
microorganism population and even eliminate it completely.
Thus the formation of possible unusual odors and tastes is
eliminated. Total cost is due to EC (6.5 kW h m3) plus ozone
(approximately 13.7 kW h g1 O3 m3); therefore the aluminum consumed at the anode is 0.89 g m3.

Acknowledgments
This work was supported by the Spanish Government (project
AGL2010-15494 partially nanced from European regional development funds, ERDF) and Junta de Andaluca (through nancial
support to group AGR-125).
References
[1] IOOC, Trade Standard Applying to Table Olives, International Olive Oil Council,
Madrid, Spain, 2004.
[2] IOOC, On line reference included in World table olives gures: production,
2012. <http://www.internationaloliveoil.org/estaticos/view/132-world-tableolive-gures> (accessed 2.13).
[3] A.H. Snchez, P. Garca, L. Rejano, Elaboration of table olives, Grasas Aceites 57
(2006) 8694.
[4] E. Medina, P. Garca, C. Romero, M. Brenes, Recycling preservation solutions in
black ripe olive processing, Int. J. Food Sci. Technol. 46 (2011) 16851690.
[5] P. Garca, M. Brenes, C. Romero, A. Garrido, Colour xation in ripe olives. Effects
of type of iron salt and other processing factors, J. Sci. Food Agric. 81 (2001)
13641370.
[6] N. Daneshvar, A. Oladegaragoze, N. Djafarzadeh, Decolorization of basic dye
solutions by electrocoagulation: an investigation of the effect of operational
parameters, J. Hazard. Mater. 129 (2006) 116122.
[7] Y. Yavuz, EC and EF processes for the treatment of alcohol distillery
wastewater, Sep. Purif. Technol. 53 (2007) 135140.
[8] I.A. Sengil, M. zacar, Treatment of dairy wastewaters by electrocoagulation
using mild steel electrodes, J. Hazard. Mater. 137 (2006) 11971205.
[9] S. Khou, F. Feki, S. Sayadi, Detoxication of olive mill wastewater by
electrocoagulation and sedimentation processes, J. Hazard. Mater. 142 (2007)
5867.
[10] U. Tezcan, A. Savas, U. Bakir, Electrocoagulation of vegetable oil renery
wastewater using aluminium electrodes, J. Environ. Manage. 90 (2009) 428
433.
[11] A. Kyriacou, K.E. Lasaridi, M. Kotsou, C. Balis, G. Pilidis, Combined
bioremediation and advanced oxidation of green table olive processing
wastewater, Process Biochem. 40 (2005) 14011408.
[12] P. Garca-Garca, A. Lpez-Lpez, J.M. Moreno-Baquero, A. Garrido-Fernndez,
Treatment of wastewaters from the green table olive packaging industry using
electro-coagulation, Chem. Eng. J. 170 (2011) 5966.
[13] A. Deligiorgis, N.P. Xekoukoulotakis, E. Diamodopoulos, D. Mantzavinos,
Electrochemical oxidation of table olive processing wastewater over borondoped diamond electrodes: treatment optimization by factorial design, Water
Res. 42 (2008) 12291237.
[14] United States Enviromental Protection Agency Ofce of Water, Alternative
disinfectants and oxidants guidance manual EPA 815-R-99-014., <http://
www.epa.gov/OGWDW/mdbp/alternative_disinfectants_guidance.pdf>.
[15] M.A. Khadre, A.E. Yuoseft, J. Kim, Microbiological aspects of ozone applications
in food: a review, J. Food Sci. 6 (2001) 12421252.
[16] P. Garca, M. Brenes, J. de Vicente, A. Garrido, Depuracin de las aguas
residuales de las plantas envasadoras de aceitunas verdes mediante
tratamientos fsico-qumicos, Grasas Aceites 41 (1990) (1990) 263269.
[17] N. Zouari, Decolorization of olive oil mill efuent by physical and chemical
treatment prior to anaerobic digestion, J. Chem. Technol. Biotechnol. 73 (1998)
297303.
[18] K.A. Segovia, F.N. Arroyo, P. Garca P, M.-C. Durn, A. Garrido, Treatment of
green table olive solutions with ozone. Effect in their polyphenol content and
on Lactobacillus pentosus and Saccharomyces cerevisiae growth, Int. J. Food
Microbiol. 114 (2006) 6068.
[19] K.A. Segovia-Bravo, F.N. Arroyo-Lpez, P. Garca-Garca, M.C. Durn-Quintana,
A. Garrido-Fernndez, Reuse of ozonated alkaline solutions as fermentation
brines in Spanish green table olives, J. Food Sci. 72 (2007) 126133.
[20] K.A. Segovia-Bravo, P. Garcia-Garcia, F.N. Arroyo-Lpez, A. Lpez-Lpez, A.
Garrido-Fernndez, Ozonation process for the regeneration and recycling of
Spanish green table olive fermentation brines, Eur. Food Res. Technol. 227
(2008) 463472.
[21] APHA-AWWA-WPCF, Standards methods for the examination of water and
wastewater American Public Health Association. Washington, DC, 1985.
[22] M. Brenes, P. Garca, A. Garrido, Regeneration of Spanish style green table olive
brine by ultraltration, J. Food Sci. 53 (1988) 17331736.
[23] P. Garca, C. Romero, M. Brenes, A. Garrido, Validation of a method for the
analysis of iron and manganese in table olives by ame atomic absorption
spectrometry, J. Agric. Food Chem. 50 (2002) 36543659.
[24] A. Montao, A.H. Snchez, L. Rejano, Mtodo para la determinacin del color en
salmueras de aceitunas verdes de mesa, Alimentaria 193 (1988) 137139.
[25] N.K. Khosla, S. Venkachalan, P. Sonrasundaram, Pulsed electrogeneration ob
bubbles for electrootation, J. Appl. Electrochem. 21 (1991) 986990.
[26] T. Picart, G. Cathalifaud-Feillade, M.Mazet.C. Vandensteendam, Cathodic
dissolution in the electrocoagulation process using aluminium electrodes, J
Environ. Monit. 2 (2000) 7780.
[27] C.A. Martinez-Huitle, S. Ferro, Electrochemical oxidation of organic pollutants
for the wastewater treatment: direct and indirect processes, Chem. Soc. Rev.
35 (2006) 13241340.

P. Garca-Garca et al. / Separation and Purication Technology 133 (2014) 227235

[28] Ch. Comnillelis, A. Nerini, Anodic oxidation of phenol in the presence of NaCl
for wastewater treatment, J. Appl. Electrochem. 25 (1995) 2328.
[29] P. Garca, M. Brenes, T. Vattan, A. Garrido, Kinetic study at different pH values
of the oxidation process to produce ripe olives, J. Sci. Food Agric. 60 (1992)
327331.
[30] T. Grahl, H. Markl, Killing of microorganisms by pulsed electric elds, Appl.
Microbiol. Biotechnol. 45 (1996) 148157.
[31] H.F. Diao, X.Y. Li, J.D. Guc, H.C. Shi, Z.M. Xie, Electron microscopic investigation
of the bactericidal action of electrochemical disinfection in comparison with
chlorination, ozonation and Fenton reaction, Process Biochem. 39 (2004)
14211426.

235

[32] S. Cotillas, J. Llanos, P. Caizares, S. Mateo, M.A. Rodrigo, Optimization of an

integrated electrodisinfection/electrocoagulation process with Al bipolar
electrodes for urban wastewater reclamation, Water Res. 47 (2013) 1741
1750.
[33] K.P. Dreesa, M. Abbaszadeganb, R.M. Maiera, Comparative electrochemical
inactivation of bacteria and bacteriophage, Water Res. 37 (2003) 22912300.
[34] J. Beltran, Ozone reaction kinetics for water and wastewaters systems, Lewis
Publisher, Washington, DC, 2004.