1 Introduction

Mango processing generates a variety of residues including the seed, which consists of a husk and a kernel. The mango seed kernel contains carbohydrates (58–80%), protein (6–13%), and essential amino acids and lipids (6–16%) [1], and it is a good source of phenolics, carotenoids, vitamin C, and dietary fiber that can improve human health and nutrition [1, 2]. As a result, there is a potential for biorefining of the mango seed kernel into such low-volume but high value-added products. The mango seed husk, on the other hand, has less commercial value compared with the mango seed kernel, and it is therefore disposed of into the environment, which causes pollution, or used as compost [1]. One of the ways to add economic value to the mango seed husk is through the production of activated carbon. Therefore, together with biorefining of the mango seed kernel, it can potentially increase the economic value of the mango seed as a feedstock in a biorefinery. Activated carbon is one of the commonly used adsorbents [3] in several applications including removal of dyes, odors, and contaminants, in water purification processes as well as in adsorption cooling processes [4,5,6]. The functional properties of activated carbon are among the other factors influenced by the production method. Activated carbon can be produced through physical and chemical means. The physical activation method, which is generally applied to non-renewable feedstocks such as coke, pitch, and coal [7], involves carbonization in an inert atmosphere [8,9,10,11] followed by activation using steam (800–1000 °C) or carbon dioxide [12]. Such physical activation method is costly due to its high carbonization temperature, high processing time, and low carbon yield [7] and has high safety risks for small-scale applications. Unlike the physical activation method, the chemical activation method which is applied to biomass materials is economically feasible due to its shorter processing time, higher activated carbon yield [7], and lower activation temperature (400–500 °C) [6, 13,14,15].

Activated carbon from biomass materials such as the mango seed husk is produced through chemical means where carbonization and activation occur simultaneously [8, 11, 12, 16,17,18]. The activation chemicals are added before reaching the carbonization temperature of 400–500 °C [6, 13,14,15]. Many chemicals such as CaCl2, ZnCl2, H3PO4, K2CO3, and KOH have been used for the production of activated carbon [6, 12, 19, 20]. Each of these chemicals affects pore formation differently and has different safety concerns that affect the application of the activated carbon produced. For example, H3PO4 restricts the escape of the tar from the carbon during carbonization, thus affecting the development of pore structures [21]. The KOH, on the other hand, is normally applied to already carbonized materials because of their lower activated carbon yield when virgin biomass is used than other activating chemicals such as H3PO4 and ZnCl2 [7]. Similar to ZnCl2, the NaCl is a strong dehydrating agent, which prevents the formation of tars and enhances the release of volatile matters from the carbon and the formation of well-developed pore structure in the carbon [7]. Furthermore, NaCl is not as toxic as other chemicals. The NaCl has a boiling point of 1465 °C, which is higher than the carbonization temperature, and, thus, does not decompose during carbonization to produce hazardous fumes [22, 23]. On the other hand, the H3PO4 and ZnCl2 with boiling points of 213 °C and 732 °C, respectively, may decompose at the carbonization temperature to produce toxic fumes [22, 23]. Arguably, KOH has a boiling point of 1327 °C, which is equally higher than the typical carbonization temperatures [7]. However, the presence of KOH in the effluent during the washing stage may be hazardous to the environment and humans [7]. Consequently, activated carbon produced using NaCl activation may be suitable for application in pharmaceutical and food industries unlike that from KOH and ZnCl2 activation due to safety and contamination issues [7]. The time-weighted average recommended an airborne exposure limit for NaCl of 5–10 mg m−3 [24] while those of ZnCl2 and H3PO4 are 1–2 mg m−3 [22, 23].

The NaCl is considered to be an effective catalyst for activated carbon production from wood sources [25]. Besides, NaCl has a high thermal conductivity than most of the chemicals used [26], which can increase the rate of heating the mango seed husk during carbonization, and, thus, becomes superior to other activation chemicals. Furthermore, most of these activation chemicals are not readily available for small-scale production. NaCl can be easily accessed at a small-scale level.

The use of NaCl as an activating agent in the production of activated carbon from non-wood sources has been reported for the production of powdered activated carbon from mango seed husk [27] but not for granular activated carbon. Powdered activated (< 0.045 mm) carbon is normally suited as an adsorbent in wastewater treatment whereas granular activated carbon is suited for applications such as adsorption cooling because of its high diffusion rate through the adsorbent bed and for not being easily sucked out of the adsorber during vacuum creation [28].

The physical and functional properties of activated carbon apart from the type of activation chemical are affected by soaking time, carbonization temperature, and particle size [9, 15]. Therefore, it is important to understand the effects of the impregnation ratio, soaking time, and carbonization temperature on the properties of the activated carbon. This study aimed to investigate the production of granular activated carbon from mango seed, targeting specifically the husk, through NaCl activation, as a potential adsorbent in the adsorption cooling system. The husk is targeted instead of the whole mango seed for the production of the activated carbon in order to allow potential integration with mango seed biorefinery, which would produce low-volume but high value-added bioproducts. Specifically, the study assessed the optimum production conditions for making the granular activated carbon from the mango seed husk by simultaneously considering the effects of NaCl impregnation ratio, soaking time, carbonization temperature, and the interaction of these parameters on the physicochemical and functional properties, including surface area, ash content, and bulk density using response surface methodology (RSM) [29].

2 Materials and methodology

2.1 Materials

The NaCl (99.5% in purity) was purchased from Kimix Chemical and Laboratory Supplies CC. Tommy Atkins mango kernels were kindly donated by Hoedspruit Fruit Processors (South Africa). Nitrogen gas (technical grade; 99.5% purity) 5.0 (Afrox Ltd.) was used for the pyrolysis of the mango husk, while carbon dioxide and liquid nitrogen baseline 5.0 (Afrox Ltd.) were used in the characterization of the activated carbon.

2.2 Mango husk preparation and characterization

Pre-dried Tommy Atkins mango kernels were obtained from Hoedspruit Fruit Processors (South Africa). The mango kernels were opened to separate the seeds from the husk. The mango husks were reduced into sizes ranging from 1 cm to 2 cm using a pair of scissors. The resulting mango husks were dried in an oven at 105 °C for 24 h until a constant weight was reached. A sample of the mango husks was milled in a Condux-Werkbei (Hanau) mill. The resulting particles were then sieved using a Vibratory Shaker Retsch AS200. The fractions that were retained on 425 μm and 625 μm were used for chemical analysis. The samples were subjected to a proximate analysis based on ASTM standards as follows: moisture [30], ash content [31], volatile matter [32], and fixed carbon. The lignocellulosic composition of the material was determined by the NREL method [33].

2.3 Proximate analysis

The fractions (425 μm and 625 μm) were subjected to proximate analysis to determine the moisture, ash, fixed carbon, and volatile contents, based on the ASTM-E-1131 [30] method using a Metler Toledo TGA/DSC 1 thermogravimetric analyzer. About 24 ± 3 mg of each lignin sample was loaded onto a 600-μl alumina crucible and placed on the TGA pan. Heat was supplied in a sequential manner to remove moisture removal at 110 °C, volatile content removal at 900 °C under nitrogen, and fixed carbon combustion with oxygen at 900 °C.

2.4 Activated carbon production

The procedure for production of the activated carbon from mango husk is shown at Fig. 1. NaCl solutions were prepared to the required impregnation ratios of 0.25, 0.50, and 0.75 defined as the ratio of dry weight of NaCl to the weight of the mango husk based on similar study [27]. Dried mango seed husk sample (100 g) containing sizes ranging from 1 to 2 cm was mixed in a beaker with 250 mL of NaCl solution of specified concentration (10% w/v, 20% w/v, and 30% w/v). The size of the mango seed husk was chosen by taking into account the requirement for producing granular activated carbon, which is suited for application in adsorption cooling systems that involve vacuum creation. Notably, activated carbon produced from smaller sized mango seed would likely be close to powdered form, which could easily be sucked out during vacuum creation. Mango husks were soaked in NaCl solution for a period of 2, 4 and 6 h to obtain impregnation ratios of 0.25, 0.5 and 0.75 at room temperature (25 °C). The impregnated husks samples were dried in an oven at 50 °C for 72 h. The moisture content at the end of the oven drying was about 25%. Approximately 50 g of the impregnated sample was placed in a stainless steel container, which was inserted into the reactor tube of the pyrolysis furnace for carbonization under nitrogen atmosphere, at a flow rate of 1 l min−1 and a heating rate of 10 °C min−1 (Fig. 2). The carbonization temperatures were 400 °C, 450 °C, and 500 °C, which were chosen based on previous studies. The optimum carbonization temperature for most biomass materials generally falls between 400 °C and 500 °C [4, 9, 15, 34]. Once the carbonization temperature was reached, the samples were kept constant at the carbonization temperature for 1 h. At the end of this process, the flow of the nitrogen gas continued to avoid carbon reacting with oxygen at high temperature to produce CO2 which could lead to a loss of carbon. After carbonization, the sample was cooled down outside the furnace at room temperature.

Fig. 1
figure 1

Process flow for the production of mango seed husk activated carbon using NaCl. The activated carbons produced were rinsed with demineralized water several times to remove any excess NaCl that did not react. The activated carbon was dried in an oven at 105 °C for 24 h to evaporate the water until constant weight was reached. The dried activated carbon was placed in a sealed container and stored in a desiccator

Full size image
Fig. 2
figure 2

Setup for pyrolysis of the treated mango husk for production of activated carbon

Full size image

2.5 Experimental design and statistical analysis

The experiments for the production of the activated carbon were carried in a 33 Box-Behnken fractional factorial [29] designed experiment with three center runs for impregnation ratio, soaking time, and carbonization temperature to give a total of 15 runs. The Box-Behnken fractional factorial was used because the optimum values were expected to be within the limits set for each variable [15]. The dependent variables (responses) analyzed are yield, bulk density, ash content, and surface area. Commercial activated carbon was used as the benchmark in this study.

2.6 Activated carbon characterization

The yield of activated carbon produced was determined using the relation shown in Eq. (1):

$$ \mathrm{Yield}\ \left(\%\right)=\frac{W_1}{W_0}\times 100 $$
(1)

where W1 is the dry weight of the final activated carbon and W0 is the dry weight of the precursor material. The bulk density was determined by a standard method [34,35,36]. Samples of activated carbon were weighed into a known volume graduated cylinder and tapped gently until the volume of activated carbon remained constant in the cylinder. The bulk density was calculated as the ratio of the weight of the activated carbon to the known volume of the closely packed sample. Ash content was determined by a standard method [35]. Since chemical activation normally produces narrow pores [11, 37, 38] and nitrogen has a problem of diffusion of the molecules inside the narrow pores (< 0.7 nm) [39], it was suggested to use CO2 for characterization of micropores to complement to N2 characterization [38]. Therefore, CO2 was used for the characterization of the activated carbon to be used in the optimization process. After that, N2 characterization was performed on the optimized product. The CO2 characterization of the activated carbon was done using a 3Flex surface characterization instrument from Micromeritics Instrument Corporation (Micromeritics ASAP 2020). About 0.3–0.6 g of each activated carbon sample was first freed of moisture and atmospheric vapor by application of electrically induced heat and evacuation in a VacPrep 061 degasser (Micromeritics Instrument Corporation). The sample was heated to 90 °C with a vacuum pump run and held for 1 h and thereafter heated to 250 °C and kept at this temperature for 20 h while the vacuum was still running. Thereafter, the sample was transferred to the 3Flex instrument (Accelerated Surface Area and Porosimetry System, Micromeritics Instrument Corporation) where in situ degassing was done at a temperature of 175 °C with a vacuum pump run for 1 h. The sample temperature was then reduced to that of CO2 (273 K) at a relative pressure range (P/P0) of 0.00005–0.025. The CO2 gas was then admitted in an incremental dosage of 3.0 cm3 g−1. Surface area and pore volume were determined on CO2 adsorption isotherms measured at 273 K, and the accumulated gas quantity adsorbed and CO2 gas pressure data at that temperature (273 K) were then graphed to generate an adsorption isotherm. The isotherm data were then treated in accordance with BET theory [40] to arrive at a specific surface area. These results were used for the optimization process through desirability analysis. The characterization of the activated carbon was done using N2 by following similar steps. However, in this case, the sample temperature was then reduced to that of liquid nitrogen (77 K) after degassing. Nitrogen gas was then admitted in an incremental dosage of 3.0 cm3 g−1. Surface area and pore volume were determined on N2 adsorption isotherms measured at 77 K at a relative pressure range (P/P0) of 0.0000002–0.1026. The accumulated gas quantity adsorbed and N2 gas pressure data at that temperature (77 K) were then graphed to generate an adsorption isotherm. The isotherm data were then treated in accordance with BET theory [40] to arrive at a specific surface area.

2.7 Regression analysis and optimization

Regression analysis was performed on data in order to derive appropriate equation for each response. All variable parameters and their interactions were considered for a model for each response. Statistical analysis software (Statistica 13) was used to solve the coefficients of the second-order model with three variables for each response as shown below:

$$ Y={\beta}_0+{\beta}_1{X}_1+{\beta}_2{X}_2+{\beta}_3{X}_3+{\beta}_{11}{X_1}^2+{\beta}_{22}{X_2}^2+{\beta}_{33}{X_3}^2+{\beta}_{12}{X}_1{X}_2+{\beta}_{13}{X}_1{X}_3+{\upbeta}_{23}{X}_2{X}_3 $$
(2)

where β0, β1, β2, β3, β11, β22, β33, β12, β13, and β23 are the regression coefficients; X1, X2, and X3 are the coded independent variables/regressors for soaking time, impregnation ratio, and carbonization temperature; and Y is the particular response evaluated. Predicted values were solved from the derived equations for each of the response. The relationship of each response variable to the input variable was evaluated for its significance at a probability value of lower than 0.05 (p < 0.05), and the strength of the relationship was evaluated using regression coefficient (R2). Subsequently, a lack-of-fit test was performed to show the adequacy of the model at p > 0.05. In addition, normal distributions of the residuals were checked to validate the assumptions made in the ANOVA [29]. Identification of the optimal production conditions involved surface plots (contour plots) of the effect on two variables while holding one at a set target. In addition, the desirability function approach has been used to obtain the production conditions at which the responses exhibit the ideal optimal value (maximum).

2.8 Fourier transform infrared spectroscopy analysis of surface functional groups on the activated carbon

The changes in surface functional groups on the activated carbon were surface analyzed using Fourier transform infrared spectroscopy (FTIR) in a Thermo Scientific Nicolet iS10 apparatus equipped with smart ITR diamond attenuated total reflectance (ATR). The experiments were carried out in the wavelength range of 250 cm−1 to 4500 cm−1 with the resolution of 4 cm−1 and total scans of 64 for each sample. This analysis was done to analyze the changes in the functional groups of the raw mango seed husk at different carbonization temperatures.

3 Results and discussion

3.1 Mango seed husk characterization

Characterizations of feedstock are important for determining the functional properties and quality of the activated carbon that can be formed [13, 41]. The chemical composition analysis (Table 1) indicates that the mango husk contains about 4.24% moisture, 19.56% fixed carbon, 74.43% volatile matter, and 2.73% ash (Table 1).

Table 1 Proximate analysis and lignocellulosic composition of mango seed husk
Full size table

The percentage of lignin, hemicellulose, and cellulose is comparable to what is reported in the literature of 5–20.71%, 15.6–16.63%, and 34.68–39.4%, respectively [42, 43]. Furthermore, the higher the content of volatile matter, the greater the porosity and, subsequently, the higher the surface area of the activated carbon produced [11]. According to Suhas and Ribeiro Carrott [13], materials with a greater content of lignin produce activated carbon with a predominantly macroporous structure, while raw materials with a higher content of cellulose produce activated carbon with a predominantly microporous structure.

3.2 Effect of production conditions on characteristics of the activated carbon

Varying the impregnation ratio, soaking time, and carbonization temperature resulted in the production of activated carbon with different characteristics in terms of surface area, ash content, and bulk density (Table 2). High carbonization temperature provided activated carbon with high surface area, low ash content, and bulk density (Table 2). The contour plots (Fig. 3a–c) show the trends for the effects of each parameter and interactions of the parameters on the specified responses for the activated carbon. The surface area of the activated carbon increased with increased soaking time and carbonization temperature (Fig. 3a (i and iii)). The surface area is a measure of the porosity of activated carbons on which adsorption can take place. The relationship between the surface area and the impregnation ratio was different from that of carbonization temperature. The surface area decreased with an impregnation ratio up to a value of 0.45, and thereafter, the surface area increased (Fig. 3a (ii and iii)).

Table 2 Characteristics of activated carbon produced using the pyrolysis method at different process conditions
Full size table
Fig. 3
figure 3

a Effects on the surface area of (i) temperature vs soaking time at 0.50 impregnation ratio, (ii) impregnation ratio vs soaking time at 450 °C, and (iii) temperature vs impregnation ratio at a 4-h soaking time. b Effects on the ash content of (i) temperature vs soaking time at 0.50 impregnation ratio, (ii) impregnation ratio vs soaking time at 450 °C, and (iii) temperature vs impregnation ratio at a 4-h soaking time. c Effects on the bulk density of (i) temperature vs soaking time at 0.50 impregnation ratio, (ii) impregnation ratio vs soaking time at 450 °C, and (iii) temperature vs impregnation ratio at a 4-h soaking time

Full size image

Thus, the greater the impregnation ratio (beyond 0.45), coupled with either longer soaking times or high carbonization temperature, the higher the surface area for adsorption (Fig. 3a (i–iii)). The impregnation ratio and carbonization temperature were the two parameters that showed significant effects (p < 0.05) on the surface area (Fig. 4a). However, the carbonization temperature had the greatest positive impact on the surface area of the activated carbon, whereas the impregnation ratio had the most negative effect (Fig. 4a). The increase in carbonization temperature increased the surface area (Fig. 3a (i and ii), which can be attributed to the degradation of the hemicellulose, cellulose, and lignin that produces volatile compounds, bio-oil, and carbon [13]. The degradation of these compounds, similar to dissolution by NaCl, leads to further opening of pores in the seed husk structure. However, the degradation of these components does not occur at the same temperature [44]. For example, thermal degradation of hemicellulose, cellulose, and lignin would occur between 220 °C and 315 °C, between 315 °C and 400 °C, and between 160 °C and 900 °C, respectively [44]. Therefore, the higher carbonization temperature of 400–500 °C applied (Table 2) most likely degraded hemicelluloses as well as the cellulose and lignin. The results are similar to the study of Adinata et al. [14] where activated carbon produced with both KOH and K2CO3 as activation agents showed similar trends for such temperature ranges.

Fig. 4
figure 4

a–c Pareto charts showing the size and significance of the effects of activation, temperature, soaking time, and activation temperature on the properties of mango seed husk activated carbon produced using NaCl as an activation agent

Full size image

The increase in soaking time increased the surface area of the activated carbon alone and also in interaction with impregnation ratio and carbonization temperature. The effect suggests that during manufacturing of activated carbon, these parameters may have to be monitored and controlled simultaneously for optimal surface area. However, since only the temperature (linear) and the impregnation ratio (quadratic) showed significant effects (p < 0.05) on the surface area (Fig. 4a), the desired surface area for the activated carbon can be tailored made by controlling the two parameters independently.

The presence of ash in the activated carbon indicates the presence of inorganic content, which reduces its quality [45]. Similar to the surface area, the carbonization temperature and impregnation ratio were the most important factors to consider when high-quality activated carbon is desirable from the mango seed husks. The results show that the effects of these two parameters were significant at p < 0.05 on the ash content (Fig. 3b (i–iii)). On the other hand, a combination of high impregnation ratio and longer soaking time is not desirable for high-quality activated carbon because of increased ash content (Fig. 3b (ii)) unless it is accompanied by high carbonization temperature above 450 °C. At the stated temperature, the ash content is less than 8% for soaking times and impregnation ratio of less than 6 h and 0.6, respectively (Fig. 3b (i and iii)). The effect of carbonization temperature on ash content could be because the sodium absorbed in the husk is reduced to metallic sodium at elevated temperatures and is subsequently lost during the washing stage [11]. Similar to the surface area, the impregnation ratio (linear) has the most negative effect on the quality of the activated carbon because of its positive relationship with the ash content (Fig. 4b) whereas the temperature has a negative relationship with the ash content.

Bulk density is an important parameter mainly for handling of the activated carbon in the adsorption cooling system, in particular the adsorbent unit. The bulk density affects the volume that can be handled in the adsorbent unit per unit time.

The bulk density was largely affected by all the three parameters: carbonization temperature, soaking time, and impregnation ratio (Fig. 3c (i–iii)). The effects of carbonization temperature (linear), impregnation ratio (quadratic), and soaking time (quadratic) on the bulk density were significant (p < 0.05) (Fig. 4c). Activated carbon of low bulk density from the seed husks can be obtained if the production is done using impregnation ratios between 0.5 and 0.6 at temperatures above 500 °C and soaking times between 3.5 and 5.5 h (Fig. 3c (i and iii)). The increase in bulk density at reduced carbonization temperature may be related to the limited breakdown of the structural components (hemicellulose, cellulose, and lignin) [44], which leads to reduced porosity.

3.3 Optimal conditions for the production of activated carbon from mango seed husk

The non-linear regression coefficients for the model (Eq. (2)) are shown in Table 3 (a). The positive relationship between carbonization temperature and surface area is reflected by the positive coefficient whereas the negative relationship with bulk density is reflected by the negative regression coefficient (Table 3 (a)). The experimental data fitted the model for all the three outputs (surface area, ash content, and bulk density) with an adjusted R2 value of at least 0.8 (Table 3 (b)). The p value from the lack-of-fit test showed that the data fitted the model at p > 0.05 (Table 3b). The optimum production conditions were identified to be 5.8 h for the soaking time, 0.25 for the impregnation ratio, and 500 °C for the carbonization temperature (Table 4). The optimum carbonization temperature is comparable to the optimum temperature of 456 °C obtained when ZnCl2 was used as an activating chemical with the optimum impregnation ratio of 1.08, to produce activated carbon from agave bagasse [9].

Table 3 Final regression coefficients, after discarding the insignificant terms and values of the statistical test parameters that validate the model
Full size table
Table 4 Optimized conditions and predicted values of responses
Full size table

At the optimized conditions, the mango seed husk activated carbon had a BET surface area of 415 m2 g−1, an ash content of 6.92%, and a bulk density of 243 kg m−3 (Table 4). The BET surface area falls within the range of 3 m2 g−1 to 1718 m2 g−1 reported for activated carbon produced from agricultural residues [6]. However, it is lower than the surface area of 618–661 m2 g−1 for mango seed husk, which used NaCl as an activation agent mainly because the activated carbon produced was in powder form whereas, in this study, it was in granular form [27].

Similarly, activated carbon powder from coal, wood, and coconut was reported to have a higher surface area of 750–850 m2 g−1, 900–1200 m2 g−1, and 590–1500 m2 g−1, respectively [46] than the activated carbon from the mango seed husk produced in this study. The activated carbon for the commercial activated carbon had a particle size of 3 mm compared to 10 mm for this study.

The ash content and bulk density of the mango seed husk activated carbon produced in this study were 6.92% and 243 kg m−3, respectively, which shows improved quality when compared with the ash content of 13.55–14.75% for the activated carbon produced by Mise and Patil [27], but it is close to about 6% reported for commercial activated carbon [46]. The bulk density of the mango seed husk activated carbon produced is higher than the 204–232 kg m−3 bulk density reported by Mise and Patil [27]. The relatively high bulk density of the activated carbon produced in this study makes it a better choice for use in vapor-phase applications such as adsorption cooling, since it would not be easily sucked out during the vacuum creation process.

3.4 Validation of model production conditions for mango husk activated carbon

Using the predicted optimum production conditions in Table 4 to produce activated carbon, the surface area, ash content, and bulk density of the activated carbon produced were measured to validate the model (Table 5).

Table 5 Validated model production conditions for mango husk activated carbon
Full size table

The results (Table 5) showed that the values for the surface area (\( {\mathrm{BET}}_{{\mathrm{CO}}_2} \)), ash content, and bulk density of the activated carbon produced at the optimal conditions were very close to the predicted values deviating by 2.69%, 1.01%, and 2.06%, respectively. Furthermore, the surface area (\( {\mathrm{BET}}_{{\mathrm{N}}_2} \)) obtained when nitrogen gas was used to estimate the BET surface area of the activated carbon produced was found to be 415 m2 g−1 (Table 4), which falls within the surface area of 300–2500 m2 g−1 reported by Campbell et al. [47] for activated carbon.

3.5 Changes in surface functional groups on the activated carbon as determined by FTIR analysis

The infrared spectral bands (Fig. 5) depicted the changing structure of the various samples of mango seed husk activated carbon. The infrared spectrum of the raw mango seed husk shows the presence of several functional groups such as alkene, aromatic, ketone, hydroxyl, and carboxyl functional groups (Fig. 5). These kinds of functional groups are typical of many lignocellulose raw materials. The absorbance for the hydroxyl stretching vibrations of the water in the raw mango seed husk is correlated to the band at 3281 cm−1 [11, 48, 49]. The other bands detected in the raw mango seed husk are C=C (alkene functional group) at 1617 cm−1, C=O (carbonyl functional group) with a spectral band at 1732 cm−1, organic acid (carboxyl functional group) at 2934 cm−1, aromatic group (benzene) at 1443 cm−1, and C–O (ether functional group) at 1038 cm−1 [11, 50]. Several of the spectral bands in the raw mango seed husk disappeared with increasing carbonization temperature due to thermal degradation of bonds of the lignocellulosic compounds present. The activated carbon produced at 400 °C and 450 °C showed a small peak at 1617 cm−1, suggesting that there could be cellulose and lignin present. However, the activated carbon produced at 500 °C showed less remnants of lignin because the lignin is totally broken down at such high temperatures [44]. The result supports the optimum carbonization temperature obtained (500 °C) as evident from the absent or reduced absorbance from volatiles initially present in the sample.

Fig. 5
figure 5

Comparison of the FTIR spectra of raw mango seed husk (A), mango seed husk activated carbon produced at 500 °C (B), mango seed husk activated carbon produced at 450 °C (C), and mango seed husk activated carbon produced at 400 °C (D)

Full size image

Therefore, activated carbon was produced from mango seed husk through simultaneous carbonization and activation with NaCl as the activation agent, which falls within the quality range (in terms of BET surface area and ash content) of commercial activated carbon. There are some precautions to be considered when using NaCl. Although it is a relatively safer activation agent than most activating chemicals, NaCl has some challenges because of the corrosive effect to most metals; therefore, anti-corrosive materials should be used to protect the equipment used for soaking the biomass material and for the adsorber construction. Secondly, the discharged NaCl water needs special disposal procedures to avoid damage to the environment.

4 Conclusion

The optimum mango husk activated carbon production conditions determined using the response surface methodology for impregnation ratio, soaking time, and carbonization temperature were 0.25, 5.8 h, and 500 °C, respectively, which resulted in activated carbon with an ash content of 6.92%, a bulk density of 243 kg m−3, and a surface area (\( {\mathrm{BET}}_{{\mathrm{CO}}_2} \)) of 223 m2 g−1, which was 415 m2 g−1 (\( {\mathrm{BET}}_{{\mathrm{N}}_2} \)). These optimal activated carbon production conditions were within the surface area ranges of 300–2500 m2 g−1 found in the literature for similar materials [47]. At the validated production conditions, the ash content (6.69%), the bulk density (248 kg m−3), and the surface area (229 m2 g−1 (\( {\mathrm{BET}}_{{\mathrm{CO}}_2} \))) were within 2% error (1.01%, 2.06%, and 2.6%, respectively) of the predicted values, which suggests the possibility of using the model to predict and control the activated carbon production process. Furthermore, the lack of interaction among the production conditions suggests that the production of the activated carbon from mango seed husks can be controlled by varying each parameter independently to tailor make the quality of the activated carbon in terms of surface area, bulk density, and ash content for different applications. Notably, the NaCl is a relatively safe and, potentially, an environment-friendly alternative activation agent for the production of granular activated carbon from mango seed husk of comparable quality to commercial activated carbon as well as to activated carbon from biomass sources. Therefore, the study has provided an optimized process that is not only effective in producing functional activated carbon but has also a potential to reduce the environmental impact of activated carbon production and, at the same time, provide an opportunity for the development of an integrated biorefinery.