The Use of Exoskeletons and Molts of Farmed Mealworm ( Tenebrio molitor ) for the Removal of Reactive Dyes from Aqueous Solutions

.


Introduction
Dyes are heavily used in industry to color fibers, textiles, leather, paper, and other products.Reactive dyes are the most popular choice due to their long shelf-life and considerable versatility [1].However, up to 50% of the dyes can be left in the effluent due to processing limitations [2].Colored wastewater can exert a number of adverse environmental effects if discharged untreated into natural water bodies [3].Dyes can significantly reduce the transparency of water, even at concentrations as low as a few mg/L.The resultant blockage of sunlight affects aquatic plants and can inhibit the primary production process [4].Colorants also limit the diffusion of atmospheric oxygen in water, which, combined with the inhibition of photosynthesis, may lead to the deoxygenation of water bodies.Moreover, dyes tend to be toxic to aquatic organisms [5].Further exacerbating the problem, colorants are hard to biodegrade because of their often complicated chemical structure, which means that water contamination with colorants can persist for extended periods of time [6].These potential environmental effects mean that the decolorization of wastewater requires the most effective methods available.
The sorption process is deemed to be one of the most attractive options for treating colored effluent [7].During sorption, pollutants are adsorbed onto the surface of a sorbent.The exact cost and efficiency of the process depend mainly on the type of sorption materials used [8].Activated carbon is currently the most popular choice of sorbent for removing dyes [9].Its large surface area, which can exceed 3000 m 2 /g, bestows activated carbon with a very high sorption capacity [10,11].The drawback is that the substance is expensive to produce and regenerate, which is why cheaper and/or more efficient alternatives are being sought.
Chitin has been gaining popularity as a sorption material in recent years.This polysaccharide is the main building block of arthropod exoskeletons, as well as a major component of the cell walls of fungi, some algae, and bacteria [12].Industrial volumes can be harvested from shells of shrimps and crabs, i.e., waste products of seafood processing [13].Chitin-based sorption materials can prove very useful in the treatment of colored wastewater [14].Compared with commercial activated carbons, chitin can boast higher sorption capacity towards anionic reactive dyes, which are the major component of colored wastewater [15].These considerable sorptive properties of chitin are due to the high concentrations of acetamide and amine functional groups, which act as key sorption centers [16].Unfortunately, raw materials for chitin production are scarce in countries where marine crustaceans are not fished.Furthermore, chitin availability is limited by pharmaceutical companies, which purchase high volumes of biopolymer to produce drugs and dietary supplements.
As an alternative, chitin may be harvested from exoskeletons and molts of insect larvae farmed for food and feed [17].The mealworm (Tenebrio molitor) is one example of an insect farmed on a large scale for human consumption in many countries [18] and classified by the European Food Safety Authority as a "novel food" in June 2021.The chitin from mealworm molts-a waste product of mealworm farming-has very similar sorptive properties to commercial chitin from snow crab shells [15].
However, just as with crab and shrimp shells, the production of chitin from mealworm exoskeletons and molts is time-consuming and expensive.Other constituents of the exoskeletons/molts, such as protein and minerals (up to 90% of the raw material by weight), are irrecoverably wasted in the process [17,[19][20].Theoretically, these components could be capable of adsorbing dyes.It can therefore be hypothesized that "the total sorption capacity of mealworm exoskeletons and molts against dyes is greater than the sorption capacity of pure chitin extracted from the same quantity of raw material".If confirmed, it would imply that harvesting pure chitin for sorbents from insect exoskeletons and molts is not commercially reasonable.The present work aims to verify the aforementioned hypothesis.To the best of the authors' knowledge, fragments of insect shells have not been investigated as sorbents so far.
Our study investigated the performance of exoskeletons and molts of mealworms as sorption materials for the removal of popular industrial dyes: Reactive Black 5 and Reactive Yellow 84, from aqueous solutions.

Sorbents
The molts and exoskeletons used in the study came from mealworm (Tenebrio molitor) grown for animal feed by the Department of Veterinary Prevention and Feed Hygiene, University of Warmia and Mazury in Olsztyn, Poland.The insects were fed oatmeal and carrots.No instances of cannibalism were observed, indicating that the mealworm was provided sufficient protein.Chitin fractions in dry matter were 8.4-9.6% for the exoskeletons and 17.0-18.0%for the molts [17,19].

Sorbates (Dyes)
The Reactive Black 5 and Reactive Yellow 84 were purchased from the "BORUTA S.A." dye and pigment production plant in Zgierz (Łódzkie Province, Poland).
Reactive Black 5 (RB5) is an anionic, reactive azo dye popular in the textile industry.Its anionic (acidic) nature is attributable to the sulfone groups present in its structure (Figure 1).It has two characteristic vinyl-sulfone groups, which provide it with the capacity to undergo chemisorption and permanently bind to the material being dyed.Its molar mass is 991.8 g/mol, and the maximum absorbance peak is at λmax = 600 nm.The material is used to dye cellulose, viscose, cotton, and polyamide fiber and is sold under the popular trade names Begazol Black B and Remazol Black B. Similar to Reactive Black 5, Reactive Yellow 84 (RY84) is an anionic, reactive azo dye popular in the textile industry.Just like RB5, RY84 is acidic due to its sulfone functional groups (Figure 1).Its reactivity is attributable to chlorotriazine functional groups, which allow it to permanently bind to the material via a condensation reaction.Its molar mass is relatively high at 1628.2 g/mol, and its maximum absorbance peak is at λmax = 356 nm.It is generally used to dye cotton, viscose, and polyester fibers.Alternative trade names include Active Yellow HE-4R and Lamafix Yellow HER.
All chemical reagents used were purchased from POCH S.A., Gliwice, Poland, and were of p.a. (analytical purity) grade or higher.

Laboratory Equipment
The following laboratory equipment was used for the study:  Gemini VI (Micromeritics, Norcross, GA, USA)-for the measurements of porosity and surface area of the sorbent.

Sorbent Preparation
The mealworm exoskeletons and molts were subjected to the same pre-sorption procedure.The material was ground in a laboratory mill and passed through laboratory sieves with mesh diameters of 3 mm and 2 mm.The 2-3 mm biomass fraction was washed with deionized water, then poured into a beaker with petroleum ether, and left for 24 h (25 °C) to dewax.The biomass was then filtered and rinsed with deionized water to remove the petroleum ether.Once dried in a laboratory desiccator (105 °C), the sorbent composed of crushed and cleaned mealworm exoskeletons (MES) and mealworm molts (MMS) was ready for analysis.

Determination of pH Effect on Dye Sorption Efficiency
A total of 0.20 g portions of either the exoskeleton-based (MES) or molt-based (MMS) sorbent was weighed into a series of Erlenmeyer flasks (500 mL).Then, 200 mL of either the RB5 or RY84 solution (50 mg/L, pH 2-11) was added to each flask.The flasks were placed on a laboratory shaker (150 RPM, vibration amplitude 30 mm) for 120 min.Afterward, 10 mL samples were collected into a polypropylene test tube to quantify the dye remaining in the solution.The solutions were also pH-tested after the sampling.

Determination of Dye Sorption Kinetics
A total of 1.00 g portions of the tested sorption material (either MES or MMS) were distributed into a series of narrow laboratory beakers (1000 mL).Then, 1000 mL of a dye solution (either RB5 or BR46, 50 mg/L or 250 mg/L) was added into the beakers while ensuring optimal sorption pH (determined based on the tests in Section 2.6).The beakers were placed on a multi-station magnetic stirrer (150 RPM).Standard 40 × 8 mm Tefloncoated cylindrical magnetic stirrers were used to mix the solution with the sorbent.At specific time intervals (0, 10, 20, 30, 45, 60, 90, 120, 150, 180, 240, and 300 min), an automatic pipette collected 2 mL solution samples into the prepared test tubes.

Determination of pH Effect on Dye Sorption Efficiency
A total of 0.20 g portions of a sorbent (either MES or MMS) were transferred into 500 mL Erlenmeyer flasks, then diluted in the same flasks in 200 mL of a dye solution (RB5 or RY84) with a concentration of 10-500 mg/L and optimal pH (determined based on the tests in Section 2.6).The flasks were placed on a laboratory shaker (150 RPM, vibration amplitude 30 mm).Once the established time to sorption equilibrium had passed (individually set for each dye and sorbent according to the tests in Section 2.7), samples were collected from the flasks into test tubes.
All experimental procedures described in Sections 2.6-2.8 were performed in triplicate at a constant 25 °C.The stirring rate was set to ensure thorough mixing of the sorbent with the entire volume of the solution.The concentration of dye in the samples was spectrophotometrically determined using 10 mm quartz cuvettes.Dye curves were plotted for concentrations of 0.0-50.0mg/L at wavelengths of 600 nm (RB5) and 356 nm (RY84).Solutions of >50 mg/L were diluted in deionized water.

Computation Methods
The amount of dye adsorbed onto MES/MMS was computed from Formula (1).T-absolute temperature (K).

Characteristics of Tested Sorbents (FTIR, Surface)
The FTIR spectra of the sorbents from mealworm exoskeletons and mealworm molts are very similar to each other, suggesting their similar chemical composition.Both materials show absorption bands characteristic of chitin, proteins, and lipids (Figure 2).The saccharide-specific peaks at 1068 cm −1 and 1025 cm −1 indicate the presence of chitin and suggest C-O stretching for the C3 and C6 of the pyranose ring [21].The peaks at 1155 cm −1 (symmetrical C-O-C stretching) and 1114 cm −1 (asymmetrical C-O stretching) are also typical of saccharides [21].The small peak of 2961 cm −1 suggests a methylene group in the saccharide ring of chitin [22].
The distinct peak at 1626 cm −1 (amide I region) is indicative of a hydrogen bond between a carbonyl and hydroxyl group of the same polysaccharide chitin chain.The peak at 1540 cm −1 (amide II region) is attributable to N-H bending-a feature of the chitin acetamide group [22,23] (Figure 2).The presence of acetamide groups in the MES and MMS structure is also indicated by the typical peaks at 3290 cm −1 and 2870 cm −1 , corresponding to N-H (Amide A) and C=O bonds [24].The peak at 1233 cm −1 , visible in the amide III region, is also indicative of a C=O bond [25].
The peaks at 1450 cm −1 and 1374 cm −1 correspond to the CH2 and CH3 stretching vibrations typical of lipids.Finally, 2917 cm −1 and 2850 cm −1 peaks are due to symmetrical and asymmetrical stretching of CH2/CH3, lipid chains, and terminal protein groups [26].
Using a Gemini VI apparatus, the porous structure and pore distribution were measured for the tested sorbents using the low-temperature nitrogen sorption method.The BET area determined for both MES and MMS is 0.13 ± 0.02 m 2 /g and 0.24 ± 0.03 m 2 /g, respectively, whereas the total pore volume was 0.00115 cm 3 /g and 0.00166 cm 3 /g, respectively.

Influence of pH on the Dye Sorption Efficiency on MES and MMS
RB5 and RY84 sorption efficiency on MES and MMS peaks at pH 2 and deteriorates as the pH increases, with the worst performance noted at pH 11.The highest drop in dye adsorption is observed between pH 2 and 4. In the RB5 groups, a slight increase in sorption intensity is noted at pH 9. Dye sorption of RB5 and RY84 to MES and MMS is the lowest at pH 11 (Figure 3a,b).The high efficiency of anionic dye sorption at low pH resulted from a strong positive charge acquired by the sorbent.The hydronium ions of the acidic solution caused some functional groups (e.g., amine acetamide, hydroxyl) to protonate.
-NHCOCH3 + H3O + → -NH2COCH3 + + H2O -NH2 + H3O + → -NH3 + + H2O -OH + H3O + → -OH2 + + H2O The positively charged functional groups of the sorbents (-NH3 + , -OH2 + ) electrostatically interact with the negatively charged groups of the dyes (-SO3 − ), greatly promoting sorption.As the pH increases, there are fewer and fewer hydronium ions in the system, limiting the protonation of functional groups.The lower number of positively charged functional groups translates to lower sorption efficiency of the anionic dyes.At pH > 4, the dye binding rate became relatively low, as only the primary amine groups (fairly sparse on the MES and MMS surface) are still being protonated.This explains the significant decrease in RB5 and RY84 sorption on the tested sorbents at pH 2-4.At high pH (pH > 9), dye sorption may have been further impeded by the negative charge attained by the sorbent, potentially caused by the deprotonation of certain (e.g., hydroxyl or carboxylic) functional groups upon interaction with the dissolved hydroxyl ions.
The negatively charged functional groups electrostatically repulse the anionic dyes, suppressing sorption.RB5 and RY84 removal at high pH may have also been inhibited by competition with OH-ions for sorption centers.
The positive effect of low pH on anionic dye sorption efficiency has been corroborated by studies on RB5 sorption onto purified chitin [27], feathers [28], and activated carbon [29], as well as RY84 sorption onto cotton fiber [30] and sunflower hulls [31].
The slight increase in RB5 sorption efficiency observed at pH 9 (Figure 3a) is caused by the primary amine groups in the dye structure.At pH 9, MES and MMS already possessed a total negative charge, while a significant amount of the amine groups of RB5 have been ionized (-NH3 + ).Despite the generally anionic nature of the RB5, its protonated amine functional group generates a local positive electric charge, which may have electrostatically interacted with the negatively charged sorbent's surface, driving RB5 sorption.The positive effect of pH 9 on RB5 sorption has also been reported in studies investigating its sorption on eggshell membranes [32] and cotton fiber [30].However, the same positive effect is not noted for the RY84 group due to the lack of primary amine groups in the dye.
Sorbents based on mealworm exoskeletons and molts affect solution pH during sorption (Figure 3c,d).With starting pH of 4-10, the solution pH after sorption is 6.88-7.49for MES and 6.77-7.40 for MMS.The type of dye has little effect on pH trends in the system.The pH-modifying capacity of the sorbents results from their high content of readily-ionized functional groups, such as amine or carboxyl groups.At low pH, protons from hydronium cations are bound to basic functional groups (-NH2 + H3O + → -NH3 + + H2O).The solution pH is inversely proportional to the levels of hydronium ions.In contrast, at high pH, acidic functional groups release their protons, resulting in the neutralization of the hydroxide anions (-COOH + OH − → -COO − + H2O) and lower pH.The sorbents always tend towards a system pH close to the pHPZC for the given sorbent (PZC-point of zero charges), i.e., the pH at which the number of positively charged sites on the tested sorbent is the same as the number of negatively charged ones.The point of zero charges determined for MES and MMS using Boehm titration is pHPZC = 7.26 and pHPZC = 7.02, respectively (Figure 3e,f).The higher pHPZC determined for MES compared with MMS may be due to a higher number of basic functional groups (e.g., -NH2 or -NHCOCH3).
Subsequent stages of the study, as described in Sections 3.3 and 3.4, were carried out at optimal sorption pH for RB5 and RY84-pH 2.

Kinetics of Dye Sorption on MES and MMS
The time needed to reach the sorption equilibrium is within the 120-150 min range for both dyes (RB5 and RY84) and sorbents (MES and MMS) (Figure 4).Similar sorption times have also been reported in studies investigating RB5 sorption on carbonized carob tree biomass (120 min) [33] and loquat seed shells (150 min) [34], as well as RY84 sorption on wool [35] (180 min).The rate of dye sorption on MES and MMS was the fastest at the start of the process.After just 10 min, the fraction of dyes sorbed is 54% to 61% of qe for MES and 46% to 51% for MMS qe (qe-the amount of dye sorbed at equilibrium state).
The faster sorption at higher initial concentrations of RB5/RY84 probably resulted from the higher likelihood of dye molecules colliding with sorbent active sites, resulting in faster saturation of the sorption centers and termination of the process.
In each experimental group, the pseudo-second-order model shows the greatest fit to the data (higher values of R 2 and smaller values of Chi-square-χ 2 ) (Table 1, Figure 4), which is typical of organic dye sorption on biosorbents [36][37][38].
Table 1.Kinetic parameters of sorption of anionic dyes onto MES and MMS, determined from the pseudo-first order and pseudo-second order models (based on the average of three measurements).Initial dye levels have a significant effect on the k2 and qe values derived from the pseudo-second-order model (Table 1).This may suggest that RB5 and RY84 have relatively low affinity to the sorbent active sites and that the sorbent efficiency is high only at high dye concentrations.The similar k2 values for experimental groups with the same initial dye levels (Table 1) suggest that RB5 and RY84 follow similar mechanisms of sorption on MES and MMS.
The data on dye sorption kinetics are also modeled using the intraparticle diffusion model (Table 2, Figure 5).The analysis of the resultant constants shows that the process unfolded in two primary sorption phases, characterized by different sorption rates and durations.In the first phase, the dye molecules most likely diffused from the solution onto the surface of the sorbent and bound to the most accessible active sites.This phase is characterized by high sorption rates in a short amount of time.Once most of the sorbent active sites are saturated, the second phase commences, in which the dyes mainly occupy hardto-reach sorption sites in the deeper layers of the sorbent.Due to the lower accessibility to remaining sorption sites and the increased competition between dyes, phase 2 is longer than the first, with lower sorption intensity (Table 2, Figure 5).
As already mentioned, the shorter length of phase 1 in the experimental groups with higher initial dye concentrations is due to the faster saturation of the active sites and, thus, the faster transition of the process to phase 2 (Table 2).
The MMS groups feature generally lower values of kd1 and a longer phase 1, suggesting that the sorbent surface had fewer sorption centers than MES (Table 2).The higher kd1 values determined for MES may also point to a higher sorption efficiency compared with MMS.

Maximum Sorption Capacity of MES and MMS
Maximum sorption capacity data for MES and MMS are modeled using three isotherms: Langmuir 1, Langmuir 2, Freundlich, and Dubinin-Radushkevich (Table 3, Figure 6).Judging by the R 2 coefficient of determination and the Chi-squared test (χ 2 ), the Langmuir 2 model shows the greatest fit to data in all experimental groups.This indicates that at least two types of sorption centers play an important role in RB5/RY84 sorption on MES/MMS.Presumably, for both of the tested sorbents, the active sites in question are protonated amine (-NH3 + ), acetamide (-NH2COCH3 + ), and hydroxyl groups (-OH2 + ) of chitin and protein.The K1/K2 constants derived from Langmuir 2, which measure the affinity of sorbate functional groups to those of the sorbent, have similar ranges for both RB5 and RY84 sorption (0.091-0.104 for K1 and 0.002-0.06for K2).This suggests that RB5 and RY84 follow similar mechanisms of sorption on MES and MMS, presumably based on ionic interaction between the ionized groups of the dye (e.g., -SO3 − ) and the protonated function groups of the sorbent (-NH3 + , -NH2COCH3 + -OH2 + ).
The maximum RB5 and RY84 sorption capacity of the tested sorbents is 78.70 mg/g and 60.49 mg/g, respectively, for MES, as well as 55.72 mg/g and 44.25 mg/g, respectively, for MMS.The inferior sorption capacity of MMS compared with MES may be due to the lower density of sorption centers on the sorbent's surface.
The slightly lower sorption capacity of RY84 compared with RB5 may stem from the higher molar mass of the former (RY84-1628 g/mol; RB5-991 g/mol).The heavier dye may, to some extent, restrict access to the active centers in the deeper layers of the sorbent.
Table 4 shows the sorption capacity of the sorbents towards RB5 and RY84.The sorption capacities of the tested mealworm exoskeletons and molts towards RB5 and RY84 are several times higher than those of plant-based sorbents (e.g., seed hulls and shells, fruit peel, and crop stems).MES and MMS have higher sorption capacity than some types of activated carbons (e.g., those derived from walnut wood, palm tree husks, or bamboo) and are exceeded only by commercial activated carbon and pure chitin (Table 4).The sorption capacity of non-processed mealworm exoskeletons and molts towards RB5 is 35.0%and 54.0% lower than that of pure chitin.Nevertheless, it seems that extracting chitin from mealworm exoskeletons/molts to be used as a sorbent for RB5 and RY84 is not commercially reasonable.Bearing in mind that exoskeletons and molts contain only up to 10% and 18% chitin, respectively, it follows that 1 g of mealworm exoskeletons and 1 g of mealworm molts have at least 549.3% and 155.4% higher sorption capacity, respectively, than chitin extracted from the equivalent amount of biomass (Table 5).This seems to confirm the starting hypothesis assuming that "the total sorption capacity of mealworm exoskeletons and molts against dyes is greater than the sorption capacity of pure chitin extracted from the same quantity of raw material".

Conclusions
Mealworm exoskeletons and molts can be successfully used as sorbents to remove dyes from aqueous solutions, providing greater performance than some types of activated carbon.The results confirmed the starting hypothesis assuming that "the total sorption capacity of mealworm exoskeletons and molts against dyes is greater than the sorption capacity of pure chitin extracted from the same quantity of raw material".This suggests that extracting chitin for dye removal from mealworm exoskeletons/molts is not commercially reasonable.
The sorption pH is the primary factor determining dye sorption onto MES and MMS.RB5 and RY84 sorption efficiency on the tested sorbents peaked at pH 2. Sorbents based on mealworm exoskeletons and molts had a significant effect on solution pH changes during sorption.This is due to the system's tendency toward a pH close to the pHPZC for the sorbent.The pHPZC values for MES and MMS were pHPZC = 7.26 and pHPZC = 7.02, respectively.The higher pHPZC for MES compared with MMS may point to a higher number of basic functional groups (e.g., -NH2 or -NHCOCH3) on the sorbent's surface.
The time needed to reach the sorption equilibrium fell within the 120-150 min range for both dyes (RB5 and RY84) and sorbents (MES and MMS).The constants derived from the intraparticle diffusion model indicate that the dyes were adsorbed onto the tested sorbents in two distinct phases.The first, short-lasting phase (10-20 min) had high removal rates.In contrast, the second phase tended to last longer (100-130 min) and showed relatively lower intensity.
At least two types of sorption centers were shown to play an important role in RB5/RY84 sorption on MES/MMS.Presumably, for both of the sorbents, the active sites in question were protonated amine (-NH3 + ), acetamide (-NH2COCH3 + ), and hydroxyl groups (-OH2 + ) of chitin and protein.

Figure 1 .
Figure 1.Structural formulas of Reactive Black 5 and Reactive Yellow 84.

Figure 3 .
Figure 3.The effect of pH on the sorption effectiveness of (a) RB5 and (b) RY84 onto MES and MMS (mean + range).The effect of MES and MMS on solution pH during dye sorption: (c) RB5 and (d) RY84.(e,f) Determination of pHPZC for MES and MMS with the drift method.

Figure 4 .
Figure 4. Trends in dye levels throughout the sorption process: (a) RB5 on MES, (b) RY84 on MES, (c) RB5 on MMS, and (d) RY84 on MMS.Pseudo-first-order model and pseudo-second-order model.

Figure 6 .
Figure 6.Isotherms of sorption of (a) RB5 on MES and MMS and (b) RY84 on MES and MMS.
CS-concentration of the dye after sorption [mg/L]; V-the volume of the solution [L]; m-sorbent mass [g].The data on dye sorption onto the tested sorbents were modeled using the pseudofirst-order model (2), pseudo-second-order model (3), and the intraparticle diffusion model (4).

Table 4 .
Comparison of the sorption capacity of biosorbents and activated carbons against BV10 and BR46.