Dithiocarbamate-modified cellulose-based sorbents with high storage stability for selective removal of arsenite and hazardous heavy metals

A series of cellulose derivatives bearing dialkyl dithiocarbamate (DTC) groups were synthesized. Their ability of sorption of arsenite (As(iii)) and heavy metals and their storage stability in the solid state were investigated. Among them, DTC-modified cellulose derived from l-proline showed the highest sorption capacity for As(iii) and heavy metals to selectively remove them from aqueous media. It also showed exellent storage stability in air at 40 °C.


Introduction
Compounds having a dithiocarbamate (DTC) group work as good chelating agents to capture heavy metals because the DTC group is a so Lewis base that has strong affinity toward so Lewis acids such as heavy metals to form stable complexes according to the HSAB rule. 1,2 Several small organic molecules having a DTC group, such as sodium diethyldithiocarbamate, have been industrially used as sorbents for the removal of hazardous heavy metals from aqueous or organic media. 3,4 Such small molecule-based sorbents are readily available, but further treatment is oen required for the efficient removal of the resultant complexes from aqueous media due to the difficulty in precipitation of the heavy metal complexes. 5 Polymer-based sorbents carrying a DTC group are potential materials for the removal of heavy metals from aqueous media because they can work as heterogeneous sorbents for solid-liquid extraction and can be easily recovered from water. 6 However, typical polymerbased sorbents are synthesized from petroleum-based chemical materials, and the production of a large amount of waste acid solutions poses serious problem for the environment. 7 Cellulose is the world's most abundant natural polymeric raw material with a fascinating structure and properties. 8 This polysaccharide is capable of being chemically modied through the hydroxyl groups in order to develop eco-friendly and costeffective biosorbents for wastewater treatment. Recently, our group synthesized a DTC-modied cellulose material 1 with excellent ability as a selective sorbent for highly toxic arsenite, which is an inorganic As(III) compound, from aqueous media (Fig. 1). 9,10 It is generally believed that DTC compounds are stable in the solid state. During the course of the above study, however, we found that the sorption capacity of compound 1 for As(III) signicantly decreased with time when it was stored even in the solid state under ambient conditions, probably due to decomposition of the DTC groups by moisture or oxygen. Although degradation behaviors of polymer-based sorbents having DTC groups in the solid state have not been investigated in detail, degradation might be a general problem for DTC-modied polymer-based sorbents. This drawback would greatly limit the applicability of the material because manufacturing and transport of materials with poor storage stability are problematic.
We therefore decided to develop DTC-modied cellulose materials with good storage stability that are capable of efficiently removing toxic As(III) and heavy metals from aqueous media. Since monoalkyl DTC compounds (R-NH-C(]S)S À ; R ¼ alkyl group) have several competitive decomposition pathways based on the N-H group, 11 they do not have sufficient stability compared with the stability of dialkyl DTC compounds. In this study, we synthesized a series of dialkyl DTC-modied (-R 1 -N (R 2 )-C(]S)S À ; R 1 and R 2 ¼ alkyl group) cellulose materials and evaluated their ability for sorption of As(III) and other heavy metals as well as their storage stability. As a result, we identied a novel biopolymer material derived from cellulose and Lproline as a potential sorbent with excellent storage stability for selective sorption of As(III) and other heavy metals (Fig. 1).

Results and discussion
We began our study with the design and synthesis of four new DTC-modied cellulose compounds 5a-d (Fig. 2). First, commercially available microcrystalline cellulose was converted to the corresponding cellulose esters 3a-d by condensation between acyclic and cyclic N-protected amino acid derivatives 2a-d in the presence of 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC-HCl) and 4-(N,N-dimethylamino)pyridine (DMAP). Compounds 3a-d were readily soluble in organic solvents, and 1 H NMR and elemental analyses of 3a-d indicated that the degree of substitution (DS) was almost 3. Next, the tert-butoxycarbonyl group of 3a-d was removed by treatment with triuoroacetic acid (TFA) to give secondary ammonium salts 4a-d. Finally, treatment of 4a-d with CS 2 and tetramethylammonium hydroxide provided the corresponding  dialkyl DTC-modied cellulose compounds 5a-d as white powders in good yields. § Since compounds 5a-d were slightly soluble in water or acetic acid, 1 H NMR analysis of these compounds could be performed unlike in our previous study. 9 Although 1 H NMR spectra of 5a-d showed rather broadened signals, they clearly indicated the existence of each side chain and tetramethylammonium cation (see ESI †). IR spectra of 5ad showed a typical and strong band at around 1730-1740 cm À1 based on C]O stretching of ester groups. The spectra also displayed characteristic bands at around 1375-1485 cm À1 and 1155-1185 cm À1 , which correspond to N-C(S) and C]S stretching of DTC groups, respectively. The N-C(S) vibration characteristically shied to a lower value in the order of acyclic 5a (1485 cm À1 ) > 6-membered cyclic 5b (1410 cm À1 ) > 5membered cyclic 5c (1384 cm À1 ) and 5d (1375 cm À1 ) (see ESI †). Next, we evaluated the sorption ability of the obtained dialkyl DTC-modied cellulose compounds 5a-d for As(III) ( Table 1). Our previous study showed that the monoalkyl DTC-modied cellulose 1 had high capacity for sorption of As(III) in acidic and neutral conditions (#pH 7) because As(III) exists as a neutral form over a wide pH range (pK a1 ¼ 9.2), 9 and Ascontaining wastewater such as mining and smelting wastewater is usually acidic. 12 Therefore, we tentatively compared the sorption capacities of compounds 5a-d for As(III) at pH 3. Compound 5a, which is an N-methylated analogue of 1, showed a lower sorption capacity than 1 ( Table 1, entries 1 and 2). The sorption capacity of compound 5b having a piperidine ring was also not good (Table 1, entry 3). Compound 5c having a pyrrolidine ring showed better sorption capacity than 5a or 5b, but its sorption capacity was still lower than that of 1 (Table 1, entry 4). Finally, we found that the sorption capacity of compound 5d having a L-proline side chain is superior to that of 1 (Table 1, entry 5). The impact of reaction time in the introduction of DTC groups to 4d does not seem to be signicant because 5d synthesized from 4d in a prolonged reaction time (7 h / 24 h) did not show improved sorption capacity for As(III) (503.6 AE 65.8 mmol g À1 ).
We investigated the change in sorption capacity of compounds 1 and 5a-d aer they had been stored in air at 40 C for 2 weeks ( Table 1). The sorption capacity of compound 1 having an N-H group was signicantly decreased by about 89% aer 2 weeks (Table 1, entry 1), and this was consistent with the gradual degradation observed in the course of storage under ambient conditions as mentioned in the introduction section. On the other hand, the sorption capacity of N-substituted DTC derivatives 5a-d did not decrease as much as that of compound 1 even aer 2 weeks (Table 1, entries 2-5). Notably, the sorption capacity of 5c and 5d having a pyrrolidine moiety was hardly changed aer 2 weeks (Table 1, entries 4 and 5), indicating that these sorbents are sufficiently stable to be stored for a long time under ambient conditions.{ Consequently, 5d having a Lproline side chain was identied as a practical DTC-modied cellulose-based sorbent with both high sorption capacity and excellent storage stability. These good properties of 5d might be due to the high nucleophilicity of a pyrrolidine moiety, which could strongly ligate CS 2 . 13,14 However, it is unclear why the sorption capacity of 5d is different.
Finally, we preliminary tested the efficiency of compound 5d for the removal of other metal ions using aqueous solutions Fig. 3 Removal percentages of metals from 5 mg L À1 multi metal solution at pH 3 using 5d. § An appropriate caution (e.g., the use of safety glass and glove) should be paid for the use of potentially toxic reagents such as DMAP and Me 4 NOH. In addition, the use of hazardous CS 2 as a reagent is unavoidable for the synthesis of DTC compounds, but reacted CS 2 is essentially incorporated to the material as a stable DTC group.
{ Aer exposing the materials to air at 40 C for 2 weeks, the IR spectrum of 1 greatly changed whereas those of 5d hardly changed ( Fig. S1 and S2 †).
containing 21 representative metal ions together with As(III) as shown in Fig. 3. Compound 5d adsorbed heavy metals, including V(IV), Fe(III), Co(II), Ni(II), Cu(II), Zn(II), Ga(III), Cd(II), In(III), Pb(II) and Bi(III), with high efficiency. 1,9 In contrast, compound 5d hardly adsorbed alkaline earth metal ions, which are hard metals. This trend is very similar to that of 1 and is consistent with the HSAB rule. 2 Exceptionally, Ti(IV), which is a hard acid, was efficiently adsorbed by 5d, and this might be because Ti(IV) could form a stable multidentate complex with DTC groups like other heavy metals. 15 Thus, 5d is a potential sorbent for the selective and efficient removal of As(III) and other hazardous heavy metals from natural water or wastewater with high concentrations of hard metal ions such as alkaline earth metal ions.

Conclusions
We have developed a new dialkyl DTC-modied biomass-based sorbent 5d derived from cellulose and L-proline. This sorbent is a potential material for the selective removal of toxic As(III) and other heavy metals from aqueous media because it has high capacity for As(III) and other hazardous heavy metals but hardly adsorbs alkaline earth metal ions. The sorption capacity and stability was maintained even aer exposure to air at 40 C for 2 weeks, indicating its excellent storage stability for practical use. In contrast, a signicant decrease in the sorption capacity for As(III) was observed for monoalkyl DTC-modied compound 1, suggesting that monoalkyl DTC-modied polymer-based sorbents might have poor storage ability, and caution is therefore required for the practical use of such materials. The present study demonstrated development of an improved sorbent for the selective removal of As(III) and other heavy metals based on a solid molecular design with biopolymer. Studies for further improvement and practical applications of the material are ongoing in our laboratory. 1.2. For batch sorption experiments. All laboratory wares were soaked in an alkaline detergent (Scat 20X-PF; Nacalai Tesque) overnight, and then rinsed with deionized water. Subsequently, they were soaked in 3 mol L À1 HCl overnight, and then washed again with deionized water. As(III) standard solution (1000 mg L À1 ), sodium hydroxide (NaOH), nitric acid (HNO 3 , 60%) and acetic acid (AcOH, 99%) were purchased from Kanto Chemical. Sodium acetate (AcONa) was purchased from Nacalai Tesque. ICP multi-element standard solution IV containing 21 elements (Al, Ba, Be, Bi, Ca, Cd, Co, Cu, Fe, Ga, In, K, Li, Mg, Mn, Na, Ni, Pb, Sr, Y, Zn) was purchased from GL science.
The metal concentrations were quantied with inductively coupled plasma optical emission spectrometry (ICP-OES; iCAP 6300; Thermo Fisher Scientic). For pH measurements, a pH meter (Navi F-52; Horiba Instruments) was used. In order to prepare deionized water with a resistivity of > 18.2 MU cm, an Arium Pro water purication system (Sartorius Stedium Biotech GmbH) was used. A natural incubator (NIB-82; Iwaki Asahi Techno Glass) was used for heating.
4a. TFA (7.8 mL) and 3a (1.17 g, 1.63 mmol) were used for the reaction to give 4a (1.08 g, 92% yield) as a white solid. 1  4c. TFA (32 mL) and 3c (4.80 g, 6.37 mmol) were used for the reaction to give 4c (4.73 g, 93% yield) as a white solid. 1  4d. TFA (32 mL) and 3d (4.82 g, 6.40 mmol) were used for the reaction to give 4d (4.09 g, 80% yield) as a white solid. 1  2.2. Batch sorption experiments. The durability of the sorbents was investigated by comparing the sorption capacities of As(III) before and aer keeping them at 40 C for fortnight. Sorption tests were conducted in 50 mL centrifuge tubes containing 0.02 g of sorbent and 10 mL of 2 mmol L À1 As(III) solutions by agitating the tubes for 20 minutes at 25 C and 200 rpm. Then, the solutions were collected by ltrating suspensions through a 0.45 mm membrane lter. Subsequently, the metal concentrations in the solutions were determined with ICP-OES. The sorption capacities of As(III) (q e ) were calculated according to the equation shown below: where q e (mmol g À1 ) represents the sorption capacity of As(III), C i and C e (mmol L À1 ) refer to initial and equilibrium concentrations of As(III), m (g) is the weight of the sorbent, and V (L) is the solution volume. The removal efficiency of various elements was examined using polymer 5d. The solution containing 21 elements was prepared by diluting the desired amount of the ICP multielement standard solution with deionized water so that the concentrations became 5 mg L À1 . Then, the pH was adjusted to pH 3 using 0.1 mol L À1 HNO 3 or NaOH solution, and sorption tests were carried out according to the procedure mentioned above. The removed metal percentages (%) were calculated from the following equation:

Conflicts of interest
The authors declare the following conict of interest(s): Kanazawa University and Daicel Corporation hold or have a led patent related to this work (Patent Application No. PCT/JP2020/ 21903).