1536

Regeneration of Spent Earth by Wet Oxidation Abul Kalam and J.B. Joshi* Department of Chemical Technology, University of Bombay, Matunga, Bombay - 400 O19, India

Regenerat ion of deoiled spent earth by wet oxidation has been invest igated. Molecular oxygen was used as an oxidizing agent. Ef fects of operating parameters such as speed of agitation, temperature, oxygen partial pressure, slurry concentration, and cycle of regenera- t ion have been studied. Kinetics of regeneration was found to be first order with substrate concentrat ion and also with dissolved oxygen concentration.

Decolorizat ion capacity of regenerated earth was the same as that of its virgin bleaching earth. Adsorpt ion i sotherm data was f itted in Freundlich equation. The values of Freundlich cons tants of regenerated earth were comparable with the values of virgin bleaching earth.

Natural/active Fuller's earth and activated carbon are widely used adsorbents in the vegetable oil industry. In view of reducing operating costs, various attempts have been made to desorb the adsorbed material from spent earth, to make it suitable for reuse (1-3}. The methods generally used in the past include thermal, solvent and chemical regeneration. Wet oxidation, which falls into the category of chemical regeneration, is attractive because of its cleanliness, compactness and cost effective- ness over other processes.

Wet oxidation, often known as the "Zimmerman process," refers to the chemical reaction between molec- ular oxygen and suspended/dissolved organics in the presence of water at elevated temperature and pressure t4). The operating temperature varies over the range of

100

80

60

40

20

TEMPERATURE = 2 0 0 ~ OXYGEN PARTIAL PRESSURE = 0-5 M Pa SLURRY CONCENTRATION = 5 % ( w / w )

,'2 ,Is ,I8 AGITATION SPEED ( r p $ )

FIG. 1. Effect of speed of agitation on regeneration efficiency.

*To whom correspondence should be addressed.

125 - 350~ and 1-20 MPa pressure. At high temper- ature, the oxidation reaction is fast enough to exploit it in practice. The reaction proceeds via fragmentation of large molecules into smaller molecules and finally oxida- tion to carbon dioxide and water. The process is being used for a variety of environmental control and energy related applications (5).

Charest and Chornet (6) reported wet oxidation of activated carbon. Effects of agitation speed, tempera- ture and oxygen pressure have been studied. The oxida- tion proceeds readily with oxygen pressure as low as 4 MPa and at temperatures higher than 200~ The re- action was found to be first order with respect to oxygen partial pressure. It was observed that the reaction takes place throughout the particle.

Gitchel et al. (7, 8) reported regeneration of spent carbon used in a variety of waste water treatment applications by wet oxidation. Chemical and physical tests of the regenerated carbon showed that wet oxida- tion retains all the adsorbate relative efficiencies and the pore structure of the virgin carbon with the exception of the smallest pore surface (~< 10 A in diameter). The effect of successive regeneration also was studied.

Natural coloring pigments in oils and fats are higher carbon chain organics like p-carotene, xyenthophyll, chlorophyll and others. As all these compounds are

SYMBOL TEMPERATURE r ~

�9 -- 'O' . , - 200

180

160

140

100 OXYGE-N PARTIAL PRES"SURE = O.5MPa - - SLURRY CONCENTRATION = 5 o: (w /w )

8 0 -

':60

"

s

0 t I I I 1800 3600 5400 7200 9000

T IME ( s e c l

FIG. 2. Effect of substrate concentration on regeneration effi- ciency.

JAOCS, Vol. 65, no, 9 (September 1988)

1537

REGENERATION OF SPENT EARTH USING WET OXIDATION

sensitive to oxidative degradation, it was hypothesized tha t the wet oxidation method would be useful for regeneration of spent bleaching earth.

MATERIALS AND METHODS

Spent earth (86% bleaching earth, 14% activated carbon) was taken from a commercial vegetable oil refinery which had been refining peanut oil. The bleaching clay consisted mainly of aluminosilicate minerals.

Bleachability was tested on refined peanut and cotton- seed oils to check regeneration efficiency on a moder- ately colored and highly colored oil. Bleachability test- ing was done in a similar experimental set-up as reported by Kheok (9). Additional arrangements were made to introduce bleaching ear th in the bleaching pot under vacuum. Test ing was carried out at 85~ + I~ with 1% bleaching earth for peanut oil and at 92 _ 1 ~ with 2% bleaching ear th for cot tonseed oil. Bleaching was done under pressure of 5-10 mmHg. The time of 35-40 min was found to be sufficient to get the extent of bleaching near its equilibrium value. Oil color was mea- sured by a Lovibond tintometer, Model E. The regenera- tion efficiency was calculated on the basis of reduction in red color. Results were compared with virgin bleach- ing ear th containing the same ratio of bleaching ear th

2-4

2.0

i 1.5

t~

c 1. 0 t

0-5

SYMBOL TEMPERATURE (~

200 180

- ~ 160 , ~ 11,0

A

O ~ ~ 1 . . . .L . ._ 0 1800 3600 5400 7200 9000

TIME { s e c }

F I G . 3. Kinetics of regeneration with substrate concentration.

and act ivated carbon as in spent earth. The results of regeneration efficiency of bleaching earth on cottonseed oil were used for finding the kinetics. Regeneration efficiency (RE) of regenerated earth was calculated by

RE -= BR X 100 [1]

B V

where B R is the bleachability of regenerated earth (per- cent) and Bv is the bleachability of virgin earth (percent).

Adsorpt ion isotherm data was obtained by t rea t ing the oil with different weight ratios of oil/bleaching earth. The adsorption isotherm data was f i t ted in the mathematical expression developed by Freundlich:

n C ---- K C e [2]

where C is the concentration of adsorbates/g of adsorbent at equilibrium (Lovibond red unit/g); K is the decolorizing power constant of the adsorbent; C~ is the oil phase equilibrium concentrat ion of adsorbate (Lovibond red unit), and n is the decolorization sensit ivi ty constant toward the range of concentrat ion of the adsorbate.

EXPERIMENTAL

Deoiling of spent ear th was done by the conventional solvent extract ion method using a soxhlet apparatus. n-Hexane t60-80 ~ was used as solvent. Deoiled ear th was dried at l l 0 ~ for four hr to remove traces of

SYMBOL OXYGEN PARTIAL PRESSURE (MPa)

-,-O-,-- 0.5

0.4

0.2

C 0.1

TEMPERATURE = 200~ SLURRY CONCENTRATION = 5% (w/w)

D

0 1800 3600 5400 7200 9000 TIME (sec)

FIG. 4. Effect of oxygen partial pressure on regeneration effi- ciency.

JAOCS, Vol, 65, no. 9 (September 1988)

1538

A. KALAM AND J.B. JOSHI

solvent. Dried earth was used for regeneration. Regenera- tion was carried out in a two-1 SS 316 autoclave equipped with an electrical heat ing jacket, gas inlet, gas outlet, pressure releasing valve and rupture disc. Mechanical agitation was provided by six-blade disc turbine impellers.

The autoclave was charged with 1000 cm3 of the desired concentrat ion of slurry, us ing distilled water as an aqueous phase. The st i rr ing speed was adjusted as desired. A vacuum p u m p was used to evacuate air f rom the autoclave. The charge was prehea ted to the desired tempera ture . Oxygen f rom the cylinder was sparged into the autoclave directly beneath the impeller at the desired oxygen par t ia l pressure. This t ime was consid- ered as 'Zero ' t ime for the reaction. Purging of gas was continued to mainta in the desired oxygen part ia l pres- sure during the course of reaction. The t empera tu re was mainta ined within __ 1~ of the desired tempera- ture. In te rmedia te samples were wi thdrawn during the course of the reaction. The regenerated earth was pul- verized to its original particle size (~< 75 microns) and tested for bleachability. Periodic samples at higher slurry concentrat ions could not be wi thdrawn due to choking of the sample in the outlet line.

Effect of speed of agitation. The effect of agi ta t ion speed was studied to ensure t ha t all the mass t ransfer res is tance was overcome and the t rue intrinsic kinetics was being invest igated. Exper imen t s were conducted at 200~ and oxygen part ia l pressure of 0.5 MPa. The slurry concentration was 5% (w/w) in all the experiments. The agi ta t ion speed varied over the range of 6-18.3 r/s. Results are p lo t ted in Figure 1. The external mass t ransfer was el iminated at a speed higher than 15 r/s. Hence, all the addit ional exper iments were conducted above 15 rps. Because the mass t ransfer resis tance was eliminated at the highest t empera tu re (200~ it did not contr ibute any resis tance at lower tempera tures .

Effect of substrate concentration. Exper imen t s were carried out in the t empera tu re range of 140-200~ and at an oxygen part ia l pressure of 0.5 MPa. The s lurry concentrat ion was 5% (w/w) in all the experiments . The resul ts are plot ted in Figure 2.

TABLE 2

RESULTS AND DISCUSSION

The adsorba tes on spent ear th are carotenoids, soaps and other impurities. Carotenoids are unsaturated hydro- carbons, highly sensit ive to oxidat ive degradat ion (10). Al though these compounds are insoluble in water at a tmospher ic tempera ture , their desorpt ion in aqueous phase at high t empera tu re was observed. The reaction between organics and oxygen in the aqueous phase can be represented by

Organic subs t ra te -* Lower molecular --- CO2 + H20 weight compounds

Complete des t ruct ion of organics and desorpt ion of non-oxidizable and par t ia l ly oxidizable compounds into the aqueous phase revives the adsorbing efficiency of the adsorbent. Thermal regeneration also occurs, together with oxidative regeneration. Therefore, the studies on thermal regenerat ion were carried out separa te ly (11).

The color of peanut oil in the 5.25" cell was 27 yellow and 2 red, and of cot tonseed oil in the one-inch cell was 42 yellow and 7.8 red. The bleachabil i ty of virgin bleaching earth was 80% on peanut oil and 57.7% on cot tonseed oil.

TABLE 1

Reaction Rate Constants at Various Temperatures

Temperature (~

Rate constant (k) m 3

(k-E~ol~ sec-1)

140 160 180 200

1.66 • 10 -2 3.42 • 10 -2

5.0 • 10 -2 7.98 X 10 -2

Reaction Rate Constants at Various Oxygen Partial Pressures

Oxygen partial pressure Rate constant (k') (MPa) (sec -1)

0.1 9.17 • 10 -5 0.2 1.64 X 10 -4 0.4 3.03 • 10 _4 0.5 5.19 • 10 -4

2.~,

2.0~-

1.5

j 1,o

0,5

SYMBOL OXYGEN PARTIAL PRESSURE ( M P o )

0 .50

0- / ,0

0 .20

0 .10

0 1800 3600 5~00 7200 9000

TIME ( sec )

FIG. 5. Kinetics of regeneration with oxygen partial pressure.

JAOCS, Vol. 65, no. 9 (September 1988)

4-5

REGENERATION OF SPENT EARTH USING WET OXIDATION

The rate expression can be wri t ten as

dCA = kiCA] m[o2]m' dt [3]

where CA is the concentration of adsorbates on adsorbent in terms of percent deactivation; t is the time in seconds;

_~3 sec_l); k is the second order reaction rate constant (kr~o~e m is the order with respect to substrate concentration; [02] is the concentrat ion of dissolved oxygen in liquid phase (kmoles/m3), and m' is the order with respect to dissolved oxygen concentration.

At constant oxygen part ial pressure equation (3) can be wri t ten as

d C A = k,[CA]m [4] dt

where CA is the concentration of adsorbates on adsorbent in terms of percent deactivation; k' is the specific first order reaction rate constant (sec-1), and m is the order with respect to subst ra te concentration.

where k' = k[02] m [5]

where k' is the specific first order reaction rate con- s tant (sec-1); k is the second order reaction rate con-

3 , s tant (km--~'ole sec-1); [02] IS the concentrat ion of dissolved oxygen in liquid phase (kmoles/m3), and m is the order with respect to subs t ra te concentration.

From Figure 3 it can be seen tha t the order with respect to substra te concentrat ion {percent deactiva- tion) is first order. The specific rate constants were calculated by plot t ing - In CA/CAo (concentration of adsorbates on adsorbent in terms of percent deactiva- tion/initial concentrat ion of adsorbates on adsorbent in terms of percent deactivation) against time. The values of first-order ra te constants are given in Table 1.

Effect of oxygen partial pressure. The effect of oxygen partial pressure was investigated over the range of 0.1

4.C

3 . 0

ta

3.5 o O

/ I ,. 1

2.0 2.5 3.0 LO0 [0 z] (kmo les / rn 3)

FIG. 6. Plot of log k versus log-dissolve oxygen concentration.

- 0.5 MPa and at 200~ The slurry concentration in all the exper iments was 5% (w/w). The ex ten t of re- generation increases with an increase in the oxygen part ial pressure. Results are shown in Figure 4. The rate constants were calculated by plot t ing - In CA/CAo against time, as shown in Figure 5. The values of reaction ra te constants are given in Table 2. The reactio ra te was found to be first order with respect to oxygen partial pressure. A plot of log k against log-dissolve oxygen concentrat ion is shown in Figure 6. The order with respect to dissolved oxygen concentrat ion was found to be first order.

4.0

3-5

2.Sh

2.0[

3.0 Tu

r

i

O

I I I 2.1 2.2 2.3

I/T x 103{~

FIG. 7. Arrhenius plot of regeneration rate constants.

2.4

11r

~I00

z

z ~ eo

7c

w

50

TEMPERATURE = 200~

OXYGEN PARTIAL PRESSURE = 0"SM Pa

REACTION TIHE = 6hr

1539

- ' ~ PEANUT OIL -.{:3- COTTONSEED OIL

l i i I S 10 ZO 35

SLURRY CONCENTRATION ( % )

FIG. 8. Effect of slurry concentration on regeneration efficiency.

JAOCS, Vol. 65, no. 9 (September 1988)

1540

A. KALAM AND J.B. JOSHI

110

1 0 0

z 9O w

m 8o

z O

~ 70

z

TEMPERATURE ffi 2 0 0 ~

OXYGEN PARTIAL PRESSURE = 0 .5 MPa

REACTION T I M E = 6 hr

- SLURRY CONCENTRATION = 5 % ( w / w }

qD [] O 0

- -O-- P E A N U T OIL

@ C O T T O N S E E D OIL

I I I I 1 2 3

CYCLE OF REGENERATION

FIG. 9. Effect of decolorization efficiency on the cycle of regen- eration.

Effect of temperature. To find the dependence of reaction rate constant on temperature, the Arrhenius equation was used to find activation energy (Fig. 7). The value of apparent activation energy was found to be 10.0 kcal/gm mole.

Effect of slurry concentration. Exper iments were con- ducted in a slurry concentration range of 5-35% (w/w). Regeneration was done at 200~ and 0.5 MPa of oxygen partial pressure for six hr. Results are shown in Figure 8. The extent of regeneration decreases with an increase in the slurry concentration. Only 80-82% regeneration was obtained at 35% slurry concentration. With an increase in solid loading the amount of adsorbates to be dest royed increases. Therefore, for a given reaction time, the extent of regeneration decreases. Complete regeneration was found even for 35% loading when the reaction t ime was 14 hr.

Effect of cycle of regeneration. Successive regenera- tion of spent ear th was studied in order to use bleach- ing ear th repeatedly. Results are shown in Figure 9. The extent of regeneration was found to be indepen- dent of the cycle of regeneration. The extent of regenera- tion was found to be 100% even after the fourth cycle. Further, the regenerated earth works well on both types of oil. In the oxidative regeneration process, probably no degraded products accumulate on the surface of the adsorbent, unlike aqueous phase thermal regeneration.

Adsorption equilibricL Adsorption isotherms of virgin and regenerated earth (after the first cycle of regenera- tion) on peanut and cot tonseed oils are shown in Figure 10. The values of K and n for virgin and regenerated ear th were found to be similar. This indicates that , apar t from the revival of adsorption capacity, the basic nature of bleaching ear th remains unchanged during the regeneration process. The Freundlich equations for

,,=,

o

z

=: c~ o,

- - O,

2~ o.6[ w 0 - 5

0 0.4P

0-3 O O

u 0"2 O.Z

6

4

'r q

PEANUT OIL /

0 VIRGIN BLEACHING EARTH �9 REGENERATED BLEACHING EARTH

COTTONSEED OIL

Q VIRGIN BLEACHING EARTH �9 REGENERATED BLEACHING EARTH

! I I I I I ~ [ J I

0!, 0., 0.s 0 ! , . 2 ~ ,' s C e ~ EQUILIBRIUM COLOR IN OILiLOVIBONO RED UNITS)

FIG. 10. Adsorption isotherm equilibrium plot for virgin and regenerated bleaching earth.

virgin and regenerated ear th for peanut and cot tonseed otis, respectively, are represented by the following equa- tions:

C -- 3.12 Ce ~176 [6]

C : 0.32 Ce 1'5 [7]

where C is the concentration of adsorbates/g of adsorbent at equilibrium (Lovibond red unit/g) and Ce is the oil phase equilibrium concentration of adsorbate (Lovebond red unit).

REFERENCES

1. Feuge, R.O., and H.J. Janssen, J. Am. Oil Chem. Soc. 28:429 (1951).

2. Thomopoulos, C., and C. Liapis, Chem. Abstr. 58:6468 (1963). 3. Kuck, J.C., W.A. Pons Jr. and u Frampton, J. Am. Oil

Chem. Soc. 39:48 (1962). 4. Zimmerman, F.J., Chem. Eng. 65:115 (1958). 5. Schaefer, P.T., Hydrocarbon Processing 60:100 (1981). 6. Charest, F., and E. Chornet, Can. J. Chem. Eng. 54:190 (1976}. 7. Gitchel, W.B., J.A. Meidl and W. Burant Jr., Chem. Eng.

Prog. 71:90 (1975). 8. Gitchel, W.B., J.A. Meidl and W. Burant Jr., AIChE. Syrup.

Series 71:414 (1975). 9. Kheok, S.C., J. Am. Oil Chem. Soc. 59:129 (1982).

10. Christopher B.J., Carotenoids as Colorants and Vitamin A Precursors, Academic Press, New York, 1981, p. 81.

[Received October 22, 1987; accepted April 6, 1988]

JAOCS, Vol. 65, no. 9 (September 1988)