Results and Discussion

3.1. ACTIVATED CARBON FROM CO-MINGLED WASTES

In order to gain some insight into the reactions which may occur between the components of the co-mingled wastes, the thermal decomposition of the individual components, as well as of their blends, was studied by thermal analysis (DTG/TG). The emphasis is given here to the correlation between the experimental data and the theoretically predicted curves under the assumption of no interaction between components, in order to evaluate possible synergetic effects.

The thermal decomposition of the blend gives rise to simultaneous cracking reactions generating volatile products, and to polycondensation reactions favoring the formation of solid-state products.

The TG/DTG data reveal that the blend components interact during pyrolysis, changing the temperatures of initial weight loss, maximum rate of weight loss, yield of the volatile products, and char (Table 1).

According to the obtained data, there are three main zones of temperature decomposition of the single components and of the blends (Figure 1, Table 2): water decomposition in the temperature range 383-443 K, the first thermal decomposition at 463-623 K, and a second one over the temperature interval 643-773 K.

The first thermal decomposition step shows two overlapping peaks with maximum at ca. 523-548 K and 568-583 K; the process starts at 453-463 K, then gradually increases and ends at 608-623 K (Figure 1). Therefore, this initial thermal decomposition step occurs over the temperature range for the decomposition of hemicellulose, cellulose, aromatic tar stock, and maltenes fraction of the aliphatic petroleum stock.

However, in contrast to the theoretical predicton, the maximum rate of weigh loss for the blends starts at a lower temperature, and the overall process for the first thermal decomposition step is found to be stronger. The yield of volatiles (12-24%) is higher than the calculated ones based on the additive behavior of the blend components. Thus, there is evidence for some interaction between the blend components (Figure 1).

The second thermal decomposition step is observed in the temperature range of Tmit = 633 K - Tmax = 668-688 K - Tfinal = 773 K, which corresponds to the temperature range for decomposition of lignin, coal, and aliphatic (asphaltenes) petroleum stock. Synergetic effects are observed for the decomposition of the blend component: a negative trend (19% wt. loss) for aliphatic petroleum stock, and a positive trend (12% wt. loss) for aromatic tar stock.

The interaction between the blend components is more apparent when the total product yield (char, volatiles) is compared to the theoretical calculated one (Table 1). The deviation of 7-10% in the char yield is more pronounced for the blends with aliphatic petroleum stock and is related to repolymerisation (cross-linking) reactions.

Despite of the different decomposition character of individual petroleum and tar stocks, the degradation of their co-mingled blends with coal and biomass was found to be nearly similar (Figure 1). Differences between the aromatic tar and aliphatic petroleum wastes were found only for volatiles/ char yield (Table 1), and char properties.

From the DTG/TG analysis, the temperature ranges for the synergetic effects were evaluated. The data obtained were used for the optimization of the carbonisation/activation processes of the co-mingled liquid and solid organic wastes. Emphasis is done to the polycondensation and repolymerisation reactions via secondary vapour-solid blend components interaction.

According to the laboratory tests, the best conditions for the synthesis were found to be direct high-temperature (1,123 K for 1.5 h) activation with steam, or two-step activation via low temperature (at 623 K for 2 h) carbonisation under inert atmosphere. The optimum blend composition was found to be 25% of biomass, 25% of low-grade coal, 50% of petroleum wastes (either aliphatic petroleum stock or aromatic tar stock), and 5% wt of blends of binary eutectic Na/K carbonates.

The activated carbons produced under optimal conditions exhibited a high surface area and a well-balanced pore size distribution (Table 2). It can be concluded that direct activation seems to be more efficient for the surface and porous structure development (Table 2).

The surface roughness and homogeneity of the materials were studied by Scanning Electron Microscopy (SEM). It was observed that the surface topography drastically changes depending on the conditions of the activation process. Activation via carbonization stage completely eliminates the carbon heterogeneity, and the porous carbon obtained exhibits uniform particle and

TABLE 1. Thermogravimetric analysis of the decomposition of the single components and of their blends.

Samples

Decomposition

Average rate of

Total

temperature range

weight loss

weight loss at

C^init ■Tmax-■^final)

(min-1)

1,223 K (%)

Bituminous coal

398-673-968

0.27

38.7

543-578-628

2.04

Sunflower husks

83.4

628-783

0.35

Aliphatic petroleum

488-608-623

0.33

85.4

stock

623-688-748

0.83

438-553-668

0.63

Aromatic tar stock

44.7

668-1223

0.13

Coal/biomass/aliphati

313-408-453

0.36

47.3

c petroleum stock

453-548-608

0.74

(Experimental)

1:1:1

608-688-888

0.46

53.9 (Theoretical )

448-543

0.51

Coal/biomass/aromati

41.9

c tar stock

543-583-623

0.70

(Experimental)

1:1:1

623-763

0.26

38.1 (Theoretical )

TABLE 2. Textural parameters of the activated carbons from co-mingled wastes. Influence

of blends composition and conditions of the activation process.

Activated carbon

SBET

Vtotal Vmicro

Vmeso/Vmicro

Precursors

(m2/g)

(cm3/g) (cm3/g)

Direct activation

Coal/biomass/aliphatic

petroleum stock

1,038

0.47 0.39

0.20

Coal/biomass/aromatic tar

stock

638

0.32 0.23

0.39

Activation via carbonisation

Coal/biomass/aliphatic

petroleum stock

979

0.46 0.38

0.21

Coal/biomass/aromatic tar

stock

111

0.11 0.05

1.2

pore size distribution. On the other hand, samples obtained by direct activation, despite exhibiting a high surface area, contain both charcoal and some husk residue with wide pores of few microns.

T, K

300 350 400 450 500 550 600 650 700 750 800 T, K

Figure 1. Differential thermogravimetry (DTG) profiles of the co-mingled wastes decomposition: (A) blend on the base of Aliphatic petroleum stock; (B) blend on the base of Aromatic tar stock. ▲ - Theoretical calculation; A - DTG experimental data.

300 350 400 450 500 550 600 650 700 750 800 T, K

Figure 1. Differential thermogravimetry (DTG) profiles of the co-mingled wastes decomposition: (A) blend on the base of Aliphatic petroleum stock; (B) blend on the base of Aromatic tar stock. ▲ - Theoretical calculation; A - DTG experimental data.

The topography study and the elemental analysis of the surface were also studied by the Auger method. Auger Depth Profiling provides quantitative information on composition as a function of depth below the surface (near-surface characteristics). The presence of some elements, such as Si, Ba, and Fe, was detected due to initial mineral content of the components in the blend. The distribution of these elements was found to be non-uniform. Under the optimal conditions used, the porous materials synthesized consisted mainly of carbon (up to 86%). Furthermore, measurements of the depth profile concentration by Auger technique did not reveal the presence of any other elements besides of carbon.

3.2. KINETIC DATA FOR SIMULTANEOUS METALS ADSORPTION FROM MULTI-COMPONENT TECHNOLOGICAL SOLUTION

The kinetic data for the simultaneous adsorption of Fe (II), Cu (II), Ni (II), Co (II) and Mn (II) on the parent and post-oxidized carbons from comingled wastes (25% sunflower husk, 25% bituminous coal, 50% aromatic petroleum stock; direct activation at 1,123 K with steam) are presented in Figure 2.

The Lagergren equation, Eq. (1), was applied to the kinetic data in order to evaluate the rate constants for simultaneous metal adsorption on activated carbons with different surface oxygen functionalities.

The linear plots of log (qeqi - qt) vs t [an example is provided in Figure 3 for the case of Fe (II)] for all of the studied metals corroborates the applicability of Lagergren equation, thus providing evidence for first-order adsorption kinetics. The specific rate constants, £ads, were calculated from the slopes of the linear plots and are listed in Table 3.

Simultaneous adsorption of the 3-d transition metals on the parent activated carbons from co-mingled waste shows that Cu (II) and Fe (II) ions adsorb faster (specific rate constants in the range 0.41-0.48 x 10 -4 min-1) and more efficiently (Merem of 21-26%) than Mn (II) and Co (II) (kads of 0.12-0.24 x 10-4 min-1 and Merem of 15-20%). There was almost no removal of Ni (II).

However, the situation is drastically changed when the post-oxidized activated carbon from co-mingled wastes is used as adsorbent. There is an increase of the specific adsorption-rate constants for Fe (II) and Cu (II) ions and also an increase of the percentage of removal (89% and 54%). Also, for Mn (II), Ni (II) and Co (II) a slightly increase is observed for the specific rate constant and percentage of removal (Figure 2).

From the TPD experiments it is seen that the oxidized carbon is enriched with carboxylic (hydrous and anhydrous) and quinones groups, which are, probably, responsible for the adsorption of the metals. Furthermore, postoxidation by nitric acid essentially modifies the structure of the parent activated carbon. The specific surface area decreases due to partial destruction of the initial micropores, along with the formation of a mesoporous structure. As a result, the post-oxidized carbon adsorbs heavy solvated metal ions from liquid phase better than the parent carbon (Figure 2).

Based on the obtained kinetics results, on previous work on adsorption equilibrium of Cu (II) and Cr (III) on synthetic and natural activated carbons [21, 24], and on data reported in the literature [25-31], we propose a mechanism for the adsorption of 3-d transition metals in real aqueous solutions.

TABLE 3. Adsorption capacity and adsorption efficiency of parent and post-oxidized activated carbons from wastes for 3-d transition metals.a The specific rate constants of the simultaneous adsorption of 3-d transition metals from multi-component technological solutions.

Metal

Me uptake

Me removal

Specific rate constants

(mg/g)

(%)

X10-4 (min-1)

Adsorption on parentb activated carbon from wastes

Cu (II)

0.41

20.6

0.48

Fe (III)

0.57

25.9

0.41

Co (II)

0.28

14.9

0.12

Ni (II)

0.08

4.2

0.04

Mn (II)

0.38

20.0

0.16

Adsorption

on post-oxidized activated carbon from wastes

Cu (II)

1.17

54.0

1.05

Fe (II)

1.94

89.1

7.41

Co (II)

0.44

23.4

0.23

Ni (II)

0.56

28.5

0.28

Mn (II)

0.43

22.5

0.37

aData are given for 72 h of contact time bActivated carbon from co-mingled 25% sunflower husk, 25% bituminous coal, 50% aromatic petroleum stock; direct activation at 850°C with steam aData are given for 72 h of contact time bActivated carbon from co-mingled 25% sunflower husk, 25% bituminous coal, 50% aromatic petroleum stock; direct activation at 850°C with steam

In previous work on Cu (II) adsorption over activated carbon in the absence of oxygen surface groups we observed the formation of a metallic film on the carbon surface (similar to the galvanic covering process) [24]. However, copper adsorption on carbons enriched with oxygen surface groups originated a decolorized solution with no metallic film formation [24]. The obtained results suggest that there are different mechanisms for copper adsorption on oxidized and non-oxidized carbons.

According to the theory of reductive adsorption [31], the spontaneous process of reduction of electrochemical ions is thermodynamically favored when the equilibrium potentials of metal ions reduction are more positive than the carbon surface potential:

The standard potential of Cu2+/Cu° pair is 340 mV (for the diluted metal solution this value is slightly lower), whereas the potential of gaseous oxygen electrode on carbon surface is 770 mV, which is significantly higher than the potential of reduction of copper cations. According to thermodynamics, the reductive adsorption of metals on the carbon surface in the presence of oxygen is not allowed. Thus, carbons enriched with oxygen groups exhibit an oxidation-reduction potential higher than the parent ones and the oxidized carbon surface can not reduce the metal cations to the metallic state.

40 Time, h

40 Time, h

Figure 2. Adsorption kinetics on parent (A) and post-oxidized (B) activated carbons from multi-components technological solution containing Z[Ni> Co, Cu Fe, Mn] = 0.05 g/L.

In the case of the activated carbon from co-mingled wastes, the carbon/ metal ions interaction via electron transfer or redox-reactions accompanied by metal ions reduction to a lower oxidation state is not favorable.

The pHPZC of the parent carbon is 5.8, and if electrostatic interactions take place then the net carbon surface will be positive charged at pH = 3.6, and the repulsion forces will inhibit the adsorption of cations, which seems to explain the low carbon adsorption capacity.

According to the pHPZC of the oxidized carbon from co-mingled waste (pHPZC = 3.4), and taking in account that the initial pH of the real solution is 3.6, the carbon surface charge is almost neutral. Thus, the electrostatic

Time, min

Figure 3. Typical Lagergren plot for heavy metals adsorption on parent (a) and post-oxidized (b) activated carbons from co-mingled wastes [example is given for Fe (II) adsorption].

Time, min

Figure 3. Typical Lagergren plot for heavy metals adsorption on parent (a) and post-oxidized (b) activated carbons from co-mingled wastes [example is given for Fe (II) adsorption].

interactions (attraction or repulsion) do not seem to have an important role on the adsorption mechanism of the metal.

It is plausible that the adsorption mechanism be a fast ion-exchange, Eqs. (3) and (4), of the aqueous metals ions, followed by their surface hydrolysis and slower chemisorption, Eq. (5), or/and an outer-sphere complexation (most probably with the first hydrolyzed hydrolytic species) that converts to inner-sphere complexation with time, Eq. (6):

RxOOH + Me(OH)2+ = RxOOH... Me(OH)2+_^ RxOOMeOH+ + H+ (6)

It was observed that the oxidation with nitric acid modified the structure of the parent carbon increasing the pore size. Considering purely a physisorp-tion process an increase of the pore size may explain the increase of the uptake.

0 0

Post a comment