Valorization of Solid Wastes from Textile Industry as an Adsorbent Through Activated Carbon Production

Materials

All wastes were supplied by a textile finishing company (Batı Basma A.Ş., Turkey) where dyeing and printing processes were carried out. Since they are common fiber types in the textile industry, CO and CO/PES (1:1)-based fabric wastes (white fabrics) were chosen as textile fabric waste. They were cut into 10 mm sizes. The waste sludge was dried in an oven at 105°C and then ground below 200 μm. Some properties of wastes are given in Table 1.

OPEN IN VIEWER

Table 1.

Elemental analysis and mass yield of resultant chars and activated carbons.

		Elemental analysis (%) a						Mass yield (%) c	Mass yield (%) d
		C	H	N	S	O b	Ash
CO	Raw	44.8	6.4	0.0	0.0	48.8	0.0	–
	Char	57.3	5.3	0.0	0.0	37.4	0.0	60	60.0
	Activated carbon	83.3	1.8	0.0	0.0	13.9	1.0	28 (46)	31.7
CO/PES	Raw	52.4	4.7	0.0	0.0	42.9	0.0	–
	Char	64.4	4.1	0.0	0.0	31.5	0.0	72	72.0
	Activated carbon	84.0	2.1	0.0	0.0	11.7	2.2	31 (43)	31.7
TWS	Raw	55.5	6.9	4.8	1.1	21.4	10.3	–
	Char	59.0	1.5	4.9	0.7	20.0	13.9	79	91.6
	Activated carbon	61.3	1.8	5.1	0.7	15.1	16.0	34 (43)	40.5

CO: cotton; CO/PES: cotton/polyester; TWS: textile wastewater treatment sludge.

a

On dry basis.

b

Calculated by difference,O = 100 − (C + H + N + S + ash).

c

Based on initial (raw) feedstock.

d

On ash free basis.

C.I. Reactive Red 141 (RR141) dye which is extensively used in the textile industry (chemical formula= C₅₂H₂₆O₂₆S₈Cl₂N₁₄Na₈, and M_W = 1774.159 g/mol) ²⁵ was used as an adsorbate and supplied from DyStar (Singapore). Aqueous solutions of RR141 have a maximum absorbance (λ_max) at 544 nm, independent of solution pH between 2 and 12. ²⁶

Activated Carbon Production

Textile wastes were first converted into biochar, which is carbon-rich and homogeneous structure by low temperature pyrolysis. Then, biochar was activated with ZnCl₂.

A stainless steel vertical bed reactor (L = 210 mm, Ø = 60 mm) was used for both biochar production and activation of biochar. In biochar production, the reactor was filled with wastes and heated up to 300°C at a rate of 10°C/min, and held at this temperature for 15 min. The reactor was continuously purged with nitrogen (25 mL/min flow rate) to remove volatile degradation products. The resulting solid products (biochar) were weighed, and then stored for activated carbon production.

In activation, biochars were impregnated with ZnCl₂ at a ratio of 1:1 (w:w) and kept in an oven at 105°C for 24 h. Then, the samples were loaded into the reactor, and the reactor was heated up to 550°C at a rate of 10°C/min, under nitrogen gas stream (25 mL/min), and held at this temperature for 1 h. After cooling, activated carbons were boiled with 10% HCl solution for 1 h, and then washed with hot distilled water for the removal of chlorides and other inorganics. After post-treatments, activated carbon samples were milled in a blade grinder (Grinding mill, A11 basic, IKA, Germany) and then sieved through 200 mesh screen, weighed and stored in sealed containers until further analysis.

The mass yields (MY, %) of solid products (biochar and activated carbon) were calculated by equation (1), where M_p (g) is the mass of solid product (biochar or activated carbon) and M_r (g) is the mass of the precursor raw material:

MY = ( M p M r ) × 100

(1)

Characterization

Elemental analysis of the solid products (biochar and activated carbon) was carried out using LECO CHNS 932 elemental analyzer (St. Joseph, MI, USA) according to ASTM D5291-96. Also, X-ray fluorescent (XRF) spectrometry analysis was applied to TWS, TWS-based biochar, and activated carbon considering the high inorganic content of TWS in contrast to the negligible inorganic content of textile fibers.

Surface functional groups were analyzed using a Fourier-transform infrared spectroscopy (FT-IR) spectrometer coupled with attenuated total reflectance (ATR) (PerkinElmer Spectrum Two, Waltham, MA, USA). Moreover, acidic and basic surface functional groups were determined according to the Boehm titration method. ²⁷

Multi-point BET (Brunauer, Emmett, and Teller) analysis by N₂ adsorption at 77 K was employed by Quantachrome, Autosorb iQ Station 2 (Boynton Beach, FL, USA) to determine the surface area and pore characteristics of the resultant activated carbons. The pore size distributions of activated carbons were analyzed by the DFT (Density-Functional Theory) method.

Reactive Dye Adsorption Experiments

Since the reactive dyes are found in the wastewater mainly in their hydrolyzed form, in order to simulate a reactive dye liquor effluent, hydrolysate solution was prepared by the treatment of 2 g/L of RR141 with 50 g/L salt (Na₂SO₄) and 20 g/L Na₂CO₃ at 70°C for 1 h.

Adsorption experiments were carried out using the batch adsorption technique in conical flasks stirred on an orbital shaker (Nüve, ST 402, İstanbul, Turkey) at 25°C. A total of 50 mL of dye solutions of different initial concentrations (10, 50, 100, 150, and 200 mg/L) were prepared, and three different adsorbent dosages (0.5, 1, and 1.5 g/L) were used for each type of activated carbon at contact times of 0–24 h. Considering the pH of a typical textile wastewater, the adsorption experiments were conducted at pH 9. After adsorption, the samples were filtered and the supernatant liquids were analyzed by a UV-Vis spectrophotometer (PerkinElmer lambda 25) to calculate the residual dye concentration.

The decolorization efficiency was determined by the following equation:

% Removal = C 0 − C e C 0 × 100

The amount of reactive dye adsorbed onto activated carbons (q_e, mg/g) was calculated by equation (3), where C₀ (mg/L) is the initial dye concentration, C_e (mg/L) is the equilibrium dye concentration, V (L) is the total volume of dye solution, and m (g) is the mass of the adsorbent.

q e = ( C 0 − C e ) V m

Isotherm studies were used to describe the equilibrium states of the produced activated carbons at different adsorbent concentrations after adsorption. For this purpose, the adsorption test data made with the adsorption time of 24 h and the adsorbent dosage of 1.5 g/L were used. In this context, experimental results were analyzed with the Langmuir and Freundlich isotherms.

Equations (4) and (5) express the Langmuir equation in a linear form and the dimensionless separation factor (R_L), respectively, where q_max is the maximum adsorption capacity (mg/g), and K_L is the Langmuir constant (L/mg): ²⁸

C e q e = 1 q max K L + C e q max

R L = 1 1 + K L C 0

The Freundlich isotherm is expressed by equation (6) in a linear form, where K_F is the Freundlich constant ((mg/g)(L/mg)^1/n) and 1/n is the heterogeneity factor. ²⁸

ln q e = ln K F + 1 n ln C e

Mass Yield and Elemental Analysis

The mass yields of biochars and activated carbons obtained from CO, CO/PES, and TWS wastes are shown in Table 1. In the case of fibrous textile wastes, the mass yield of CO/PES-based biochars was found to be higher compared with CO-based samples. This was attributed to the higher thermal decomposition temperature of polyester fibers, which possibly did not breakdown under the operating temperature of biochar production. This result was in general agreement with the studies of Hanoğlu et al. ²⁹ and Zhong et al. ²¹ TWS-based biochar had the higher mass yield compared with fibrous wastes, which may be due to the fact that it has different chemical composition from both natural and synthetic fibers. The major components of the organic fraction of activated sludge are extracellular polymeric substances, which are complex aggregates consisting of various microorganisms (mainly bacteria) embedded in a polymeric network, besides protein, polysaccharides, and lipids as well as organic micropollutants. ³⁰ The yield of TWS char was similar with the other types of sludge reported in the literature. ³¹

As in biochar, the mass yield of wastewater treatment sludge-based activated carbon (TWS-AC) was higher than that of fibrous textile wastes. The yield of cotton textile waste-based activated carbon (CO-AC) was comparable with the yield of activated carbon obtained from jute waste ¹⁵ and waste CO fabrics ³² by ZnCl₂ activation in one step. However, Carraro et al. ⁶ reported the activated carbon yields of 46.0% and 48.0% from CO and CO/PES, respectively, by H₃PO₄ activation in one step. Similarly, activated carbon from denim fabric waste with H₃PO₄ produced with a yield of 53% which was higher compared with activated with ZnCl₂. ⁷ Compared with ZnCl₂, H₃PO₄ strongly retards the evolution of volatiles, which lead to the promoting of the carbon yield. ³³ In this study, ZnCl₂ was chosen as the activation agent, because, at the activation temperature, phosphoric acid can combine with organic species in the carbon precursor and cause the phosphorus content by forming polyphosphate esters. ³⁴ In addition, in case of TWS, the use of H₃PO₄ may lead to a mixture of P-metal complexes including Fe–P, Al–P, and Ca–P, with the various metal in the TWS.

Although the chars were free of inorganics, the fiber-based activated carbons contained little ash, probably due to the incorporation of zinc into the carbon matrix via C–O–Zn bonds. As expected, sludge-derived activated carbon contained high ash content, even about half of the ash in sludge was removed by demineralization with HCl. The reason probably is the existence of water- and HCl-insoluble compounds in raw sludge. Supplemental Table S1 shows the XRF analysis results of TWS-based samples. Silicon, sodium, and calcium were found to have the highest content in TWS. By acid washing, 60–95% of the inorganics (except Cl and Zn) in TWS were removed from the activated carbon, while Mg and P were completely removed. It should be noted that TWS-based activated carbon contains more Cl and Zn than predicted, due to the use of ZnCl₂ for activation, despite repeated washings after activation.

Elemental analyses of the resultant biochar and activated carbons are also given in Table 1. Pyrolysis and activation led to an increase in the carbon content and a decrease in the oxygen content of CO and CO/PES samples as expected. Regarding elemental analysis composition, TWS-AC had much lower carbon and higher nitrogen and ash contents whereas its contents of ash, hydrogen, nitrogen, and oxygen were higher compared with the fiber-based activated carbons. However, only a slight difference in the carbon, oxygen, and nitrogen ratio was observed for the TWS-AC compared with the raw TWS. We can assume that the removal of carbon and non-carbon species occurs at the same extent during carbonization.

Surface Chemistry of Activated Carbons

FT-IR analyses were applied to precursor wastes and resultant biochars and activated carbons to detect the structural changes and surface functional groups. Supplemental Figure S1 illustrates the FT-IR spectra of CO waste-based samples. It was observed that the O–H stretching peak at 3100–3600 cm⁻¹ weakened after biochar production to a great extent and completely disappeared by the activation process. In addition, the peak at 2900 cm⁻¹ corresponding to the aliphatic C–H stretching disappeared by activation, which indicates the aromatization of the structure. In the spectrum of CO fiber, characteristic peaks of carbohydrates were observed in the range of 900–1200 cm⁻¹, such as the C–O–C stretching in pyranose ring at 1160 and 1109 cm⁻¹ and C–O stretching at 1030 cm⁻¹. ³⁵ The peak intensity of carbohydrate bands was significantly reduced by pyrolysis and activation, which denotes the dehydration and depolymerization of cellulose. Furthermore, new peaks were appeared at 700–900 cm⁻¹ and 1556 cm⁻¹ assigned to aromatic C–H deformation and C=C vibration of aromatic rings, respectively.^36,37 A peak characterizing carboxylic acid groups emerged at 1697 cm⁻¹ in activated carbon samples. As a result, it can be reported that along with aromatization, acidic functional groups emerged by an activation process.

Supplemental Figure S2 shows the FT-IR spectra of CO/PES-based samples. As in CO-based samples, the O–H stretching band in the range of 3100–3600 cm⁻¹ for CO/PES wastes disappeared as a result of the pyrolysis process. Carbonyl (C=O) stretching at 1713 cm^–1, aromatic ester asymmetric stretching at 1241 cm^–1, and aromatic C–H out-of-plane vibration at 722 cm⁻¹ can be seen as the characteristic peaks of polyester fiber. ¹⁶ It was observed that there was no change in the bands characterizing polyester in the spectrum of the CO/PES biochar. This indicated that the polyester part of the blend did not decompose during the biochar production process. The high mass yield of CO/PES-based biochar (Table 1) also supports this result. However, the peaks that characterize the polyester disappeared after the activation process, indicating that the polyester part was also carbonized. As in CO-based samples, the intensity of the peaks in the range of 900–1200 cm⁻¹ that characterizes CO fibers decreased by pyrolysis and new aromatic bands appeared (i.e. aromatic C=C stretching at 1555 cm⁻¹). It was investigated that the spectra of cotton/polyester-based activated carbon (CO/PES-AC) have the same peaks as CO-AC.

The FT-IR spectra of TWS-based samples came up with many functional groups as shown in Supplemental Figure S3. The wide peak in the range of 3100–3600 cm⁻¹ corresponds to the O–H stretching associated with water, alcohols, and carboxylic acids ³⁸ and the N–H stretching associated with organic compounds. ³⁹ It was observed that this peak disappeared after pyrolysis and activation. The peaks found at 2919 and 2851 cm⁻¹ indicated asymmetric and symmetrical C–H stretching in the alkyl groups. After pyrolysis and activation, these peaks were largely eliminated, which indicates that aliphatic C–H chains decomposed into carbon dioxide, methane, or other gases or turned into aromatic structures. ³⁸ Peaks in the range of 1400–1750 cm⁻¹ were assigned to COO, C=O, and C–N stretching, while peaks in the range of about 1000–1050 cm⁻¹ corresponded to C–O–C and C–O stretching, probably due to carbohydrates in the structure. ⁴⁰ The peak at 874 cm⁻¹ attributed to aromatic rings. After pyrolysis and activation, these peaks were observed to be largely eliminated and new aromatic structures emerged (i.e. aromatic C=C stretching at 1576 cm⁻¹ and aromatic C–H stretching at 880 cm⁻¹).

Besides textural properties, the oxygen functional groups are also important characteristics of the carbons, which determine their quality as adsorbents. The amount of acidic oxygen groups (carboxyl, lactonic hydroxyl, carbonyl groups) and basic oxygen groups (pyrones, ethers, carbonyls) on the activated carbon surfaces were determined by the Boehm titration. The amount of acidic surface oxides in TWS-AC was higher than those of other activated carbons, and no basic group was encountered (Table 2). CO-AC involved both acidic and basic groups. In the case of CO/PES-AC, acidic surface oxides were found to be predominant and traces of basic groups were identified. These results for CO and CO/PES-based wastes are in parallel with the studies of Silva et al. ⁷ and Carraro et al., ⁶ in which activated carbons were produced from CO and CO/PES denim fabrics, respectively.

OPEN IN VIEWER

Table 2.

Surface oxygenated groups on activated carbons and BET analysis results.

		CO-AC	CO/PES-AC	TWS-AC
Surface oxygenated groups	Acidity, mmol/g	1.302	1.516	3.228
	Basicity, mmol/g	0.932	0.086	–
BET analysis	Surface area, m2/g	1378.937	809.177	587.910
	Micropore area, m2/g	906.515	594.315	246.136
	Microporosity, %	65.7	73.4	41.9
	Total pore volume, m3/g	0.8044	0.5251	0.4946
	Micropore volume, m3/g	0.436	0.287	0.119

BET: Brunauer, Emmett, and Teller; CO-AC: cotton textile waste–based activated carbon; CO/PES-AC: cotton/polyester-based activated carbon; TWS-AC: wastewater treatment sludge-based activated carbon.

BET Surface Area and Pore Size Distribution

BET surface area and pore size distribution of the produced activated carbons are given in Table 2. The use of CO wastes as precursor led to the highest surface area, followed by CO/PES and TWS, respectively. The pore size distribution according to the DFT method is illustrated in Supplemental Figure S4. The results indicate that all of the activated carbons produced have pores distributed over a broad range, including micropores and mesopores. And the DFT distributions showed that the pore structure in CO-AC and CO/PES-AC was mainly micropores, while it was mainly mesopores in TWS-AC.

To explain the observed surface area and porosity, we can speculate that the molten ZnCl₂ acted as templates inside the carbon to form pore structures. The obtained surface areas of CO-AC and CO/PES-AC were comparable with the fabric-based activated carbons reported in the recent literature obtained by one-stage activation as given in Table 3. Literature includes a few references regarding the activated carbon production from textile sewage sludge. In those references, reported BET surface areas were 135 m²/g ²² and 336 m²/g ²³ by KOH activation and 221.52 m²/g² by H₂SO₄ activation. However, in the literature, the use of sludge-derived char to produce activated carbon has been less studied. Li et al. ⁴¹ produced activated carbon from sludge-derived char by ZnCl₂ activation. They obtained activated carbon having BET surface areas of 765.88 m²/g. In another study by Alvarez et al., ⁴² the activated carbon with a BET surface area of 440 m²/g was produced by CO₂ activation of the sewage sludge char. As can be seen, TWS-AC had higher or comparable BET surface area compared with textile sewage sludge and/or sludge derived char-based activated carbons. It is worth noting that the adsorption performance of activated carbons depends on not only the surface area but also the pore size, target adsorbate, and surface functional groups, as discussed in the following section.

OPEN IN VIEWER

Table 3.

Recent reported properties of activated carbons produced from solid textile wastes in the literature.

Precursor material	Activating agent	Activation temperature (°C)	BET surface area, (m2/g)	Microporosity (%) and micropore volume (m3/g)	Yield (%)	Reference
Denim fabric waste	H3PO4	500	1582	0.17	53	Silva et al. 7
Cotton–polyester mixed waste	ZnCl2	500	1605.481	0.4373	–	Zhong et al. 21
Cotton textile wastes	ZnCl2	500	1854.70	0.20	34.61	Xia et al. 8
	H3PO4	400	1261.68	0.23	36.02
	FeCl3	400	510.20	0.13	37.00
Cotton textile waste	FeCl3	700	837.39	50.58	32.66	Xu et al. 9
Cellulose denim residues	H3PO4	500	1714	0.12	46.0	Carraro et al. 6
Cellulose/polyester denim residues			1743	0.19	48.0
Waste polyester fabric	ZnCl2	500	467.60	0.1636	–	Yu et al. 17
		900	1101.50	0.7470	–
Cotton cloth waste	H3PO4	600	1175	0.48	–	Boudrahem et al. 5
Cotton fabric	H3PO4	500	690.79	–	26.53	Salehi et al. 43
Waste cotton fibers	H3PO4	450	694	0.2886	–	Ekrami et al. 44

BET: Brunauer, Emmett, and Teller.

Reactive Dye Adsorption

The decolorization efficiency of TWS-AC, CO-AC, and CO/PES-AC for the adsorbent dosage of 1.5 g/L is illustrated in Figure 1 according to the adsorption time (0.5–24 h) and dye concentration (10–200 mg/L). The results for all adsorbent dosages (including 0.5 and 1 g/L) are also given in Supplemental Figures S5–S7 for TWS-AC, CO-AC, and CO/PES-AC, respectively. The decolorization efficiency of the activated carbons was ranked in the order of TWS-AC > CO-AC > CO/PES-AC. In all cases, by increasing the initial dye concentration, the percentage removal decreased, while it increased with increase in adsorbent dosage, as expected. For instance, in case of TWS-AC, depending on the contact time, 65–80% of dye was adsorbed for the 200-mg/L initial dye concentration and adsorbent dosage of 0.5 g/L. When the adsorbent dosage increased up to 1.5 g/L, 81–95% of adsorption was observed even for the 200 mg/L initial dye concentration, and 83–99% for 10 mg/L.

OPEN IN VIEWER

Figure 1.

RR141 adsorption % on (a) TWS-AC, (b) CO-AC, and (c) CO/PES-AC for the adsorbent dosage of 1.5 g/L.

The dye adsorption was rapid in the first stage of the adsorption process, followed by a gradual adsorption. As known, the adsorption rate varies depending on several parameters such as stirring rate, structural properties of adsorbent, adsorbent dosage, and adsorbate properties. The high adsorption rate in the first stage of the adsorption process might be that the dye adsorption takes place first on the exterior surface of the adsorbent. When saturation is reached on the outer surface, the dye molecules begin to diffuse from the pores to the inner surfaces of the adsorbent, which takes a relatively longer time. ⁴⁵ Although, most of the dye was adsorbed within the first 2 h of contact between the adsorbate and the adsorbent, an equilibrium time of 24 h could be proposed to make sure that fully equilibrium was attained.

Among the activated carbons, the highest dye removal was achieved by TWS-AC. The three main parameters affecting the adsorption behavior of an adsorbent are surface area, surface groups, and porosity. Reactive dyes are the large molecules (e.g. the dimensions of RR141 is 1.72 nm × 2.68 nm × 1.66 nm); ⁴⁶ therefore, mesopores play a large part in adsorption. Although TWS-AC has the lowest surface area in this study, the reason for its high adsorption performance might be related to its higher mesoporous structure. In addition, it was thought that the higher inorganic content of TWS-AC might contribute to the higher adsorption capacity. However, CO/PES-AC showed similar adsorption capacity compared with CO-AC, although it had lower surface area.

Table 4 shows the Langmuir and Freundlich isotherm parameters for each type of activated carbon. Adsorption isotherms were also illustrated in Supplemental Figure S8. The Langmuir isotherm assumes that the adsorption occurs on a homogeneous surface containing sites with identical sorption energies. ⁴⁷ The Freundlich isotherm, on the contrary, is used for heterogeneous systems. The results obtained from adsorption of RR141 on sludge-based activated carbons conformed well with both the Langmuir and Freundlich equations; however, higher R²- values were obtained from the Freundlich model for the description of the adsorption on the fabric-based activated carbons. This might be an indication that surface heterogeneity may have a significant role in dye adsorption on fibrous waste-based activated carbons and a multi-layer adsorption might be involved. ⁴⁵ The value of n indicates the type of isotherm as linear (n =1), a chemical process (n < 1), or a physical process (n > 1). ⁴⁸ The n-values of the textile solid waste-based activated carbons showed that the adsorption of RR141 dye is a physical process and favorable. However, if we take into account that the values of 1/n < 1 indicate a normal Langmuir isotherm and 1/n < 1 indicate the cooperative adsorption, respectively, ⁴⁸ we can assume that dye adsorption on all activated carbons is a physical process and the normal Langmuir isotherm is favorable. Also, the calculated R_L values (0 < R_L- < 1) showed that the textile waste-based activated carbons are favorable for RR141 dye adsorption.

OPEN IN VIEWER

Table 4.

The Langmuir and Freundlich isotherm parameters.

	Langmuir isotherm parameters				Freundlich isotherm parameters
	qmax (mg/g)	KL (L/mg)	RL	R 2	1/n	KF (mg/g) (L/mg)1/n	R 2
CO-AC	114.94	0.26	0.0191	0.93	0.39	27.72	0.99
CO/PES-AC	111.11	0.12	0.1391	0.93	0.43	18.30	0.99
TWS-AC	161.29	0.31	0.0161	0.98	0.68	30.98	0.97

CO-AC: cotton textile waste–based activated carbon; CO/PES-AC: cotton/polyester–based activated carbon; TWS-AC: wastewater treatment sludge-based activated carbon.

As is well known, both the surface charge of adsorbent and the ionic forms of the dye in solution play an important role in adsorption process. One can think that surface charge of adsorbent becomes negative at higher pH values, and therefore, the adsorption of anions is not favored. In this study, the adsorption experiments were conducted at pH 9 by hydrolysate RR141 solution to simulate a typical textile wastewater. The fact that sorption capacity of activated carbons produced in this study is comparable with that of other activated carbons from different type of sources,^{26,49,50–53} even though their surfaces are negatively charged could be explained by the adsorption of dye by non-electrostatic. As discussed in the study of Ip et al. ⁵⁴ who investigated the Reactive Black dye adsorption onto different adsorbents, the effect of the van der Waals attraction and π–π interactions seems to be dominant and controls the adsorption. Similarly, although CO/PES-AC had lower meso- and macro-porosity but higher surface area than CO-AC, its dye adsorption capacity was found almost similar with that of CO-AC. This showed that the RR141 adsorption on these adsorbents was controlled mainly by the van der Waals attraction and π–π interactions, as in the case of TWS-AC.