Poly(benzoxazine-co-sulfur): An efficient sorbent for

Poly(benzoxazine-co-sulfur): An efficient sorbent for mercury removal from aqueous solution Sema Akay,1 Berkant Kayan,1 Dimitrios Kalderis,2 Mustafa Arslan,3 Yusuf Yagci,3 Baris Kiskan

3

1

Faculty of Arts and Science, Department of Chemistry, Aksaray University, Aksaray, Turkey Department of Environmental and Natural Resources Engineering, School of Applied Sciences, Technological and Educational

2

Institute of Crete, Chania, Crete, Greece 3

Faculty of Science and Literature, Department of Chemistry, Istanbul Technical University, Maslak Istanbul 34469, Turkey

Correspondence to: B. Kiskan (E - mail: [email protected]) and B. Kayan (E - mail: [email protected])

ABSTRACT: A novel sulfur-rich adsorbent, poly(BA-ala-co-sulfur), was synthesized by reacting allyl functional benzoxazine (BA-ala) and elemental sulfur. Simultaneous inverse vulcanization and ring-opening reactions of benzoxazine generated copolymers in several feed ratios. The adsorption behavior of these copolymers was investigated in aqueous solutions containing Hg21. A three level Box– Behnken design with four factors was applied in order to examine the interactive effect of Hg21 concentration (ppm), S % in adsorbent, temperature, and pH. The optimum adsorption conditions were determined as: 10.33 ppm Hg21, 68% S content, 329 K, and pH 6.3. Common isotherm and kinetic models were applied to the experimental data, where the Langmuir isotherm provided the C 2017 Wiley better fit (qmax 5 79.36 mg g21) and the pseudo-second order fit indicated chemisorption as the process-controlling step. V

Periodicals, Inc. J. Appl. Polym. Sci. 2017, 134, 45306.

KEYWORDS: benzoxazine; box-behnken design; mercury removal; polybenzoxazine; response surface methodology

Received 24 January 2017; accepted 16 May 2017 DOI: 10.1002/app.45306 INTRODUCTION

Water contamination by heavy metal ions has been a major environmental issue throughout the world thus limiting the available water resources.1,2 A range of heavy metals are hazardous to living organisms and among these, mercury has received particular attention since the toxicity of mercury compounds is high and exposure possibility is greater compared to other heavy metals.3–6 For example, mercury salts can cause serious health problems namely mercury poisoning7 (also known as hydrargyria or mercurialism) that result in acrodynia (pink disease),8 Hunter–Russell syndrome, and Minamata diseases.9 Generally, kidneys, liver, brain, and central nervous system are severally effected by mercury exposure, hence mercury can be classified as neurotoxic, nephrotoxic, hepatotoxic, and immunotoxic.10 On the other hand, mercury compounds are indispensable in several industries and these compounds are used or produced as by-product in fungicides, bactericides, batteries, gold mining, coal burning power stations, some electronics, dyes and pigments, catalysts, auto brakes, thermometers, and pharmaceuticals.11–13 Due to the wide spread usage, mercury circulation and its levels in air, water, soil, and food chain should be controlled and suitable removal procedures be implemented. Coagulation, precipitation, reduction, adsorption,

chemisorption, ion exchange, and membrane separation methods are generally used to remove mercury compounds from water, air, and sediments.14–16 Among these approaches, several simple adsorption processes have been successfully applied in laboratory and bench-scale. Thus, adsorbents with sulfur-, nitrogen-, and oxygen-containing functional groups as binding sites for mercury ions can effectively be used in this manner. For example, thiourea,17 xanthate,18 thiol,15,19 and dithiozone20 were successfully employed as ligands on adsorbents for binding Hg(II) ions. Alternatively, non-modified adsorbents such as activated carbons, specially designed polymers, silicates, clays, chitosan, and specific bacteria have shown promising behavior toward mercury removal.21–25 Specifically, the complexing capability of N and O atoms against Hg(II) can lead to the development of alternative mercury removal systems. From this point of view, polybenzoxazines, which contain both tertiary amino and free phenolic groups, can be used as a cost-effective material for chemisorption-reinforced adsorbents for mercury withdrawal, especially from water. Generally, polybenzoxazines are used as thermosetting resins, alternative to conventional phenolic resins (resol or novolac) or replacement to epoxy resins. This is due to the favorable features such as low water up-take, high Tg, high char yield,

C 2017 Wiley Periodicals, Inc. V

WWW.MATERIALSVIEWS.COM

45306 (1 of 11)

J. APPL. POLYM. SCI. 2017, DOI: 10.1002/APP.45306

ARTICLE

WILEYONLINELIBRARY.COM/APP

St. Louis/ Missouri, 99.9%) mercury (II) chloride (Aldrich, St. Louis/ Missouri, 99.9%), hydrogen chloride (Fluka, St. Louis/ Missouri, 37%) were used as received. All the chemicals were of analytical reagent grade. Figure 1. Synthesis and polymerization of a mono-functional benzoxazine monomer.

near-zero shrinkage during curing, limited release of byproducts during the polymerization thus avoiding formation of voids in final resins.26,27 At the same time, no strong acid catalysts are required for the polymerization in contrast to conventional phenolic resin synthesis.28–31 Therefore, polybenzoxazines have become one of the few new polymers developed and commercialized in the past 20 years. Another reason for that is benzoxazine monomers have enormous molecular design flexibility.32–38 Basically, the synthesis of a monomer is carried out by the reaction of any suitable phenol (mono or multifunctional), formaldehyde, and primary amines (aliphatic or aromatic) (Figure 1). Moreover, the synthesis of polybenzoxazines is performed by thermally activated ring opening reaction without using any catalyst or curative.39,40 Depending on the monomer structure, polymerization temperature of benzoxazine monomer vary approximately in a range of 180–250 8C.41–43 Most of polybenzoxazines advantageous properties are mainly due to the ACH2ANACH2A linkages and phenolic AOH groups that generate strong intra- and inter-molecular hydrogen bonds. Moreover, this network can attract and bind mercury salts through oxygen and nitrogen atoms. Previously, this issue has already been shown by our group and polybenzoxazines were used as effective adsorbents for the removal of Hg21 ions from sea water.44 However, even though N and O atoms are good ligands for mercury salts, sulfur-based systems are known to have much higher binding affinity for mercury compounds. Hence, by using flexible benzoxazine chemistry, polybenzoxazine adsorbents can also be designed as sulfur atom containing systems to cover all possible strong mercury binding atoms except selenium.45 As part of our continuing quest in using polybenzoxazines in novel applications, we propose a simple method to synthesize sulfur-containing polybenzoxazine adsorbents for the removal of mercury salts from water. The influence of important variables on adsorption efficiency, such as initial Hg(II) concentration (ppm), S % in adsorbent, temperature, and pH were investigated and optimized by Box–Behnken design (BBD) combined with response surface methodology (RSM) to maximize the removal efficiency of Hg21. Additionally, the kinetics and isotherms of adsorption process was studied to gain inside into the mechanism. EXPERIMENTAL

Materials Allylamine (Aldrich, St. Louis/ Missouri, 98%), 4,40 -isopropylidenediphenol (Aldrich, St. Louis/ Missouri, 97%), paraformaldehyde (Acros, Geel/Belgium, 96%), 1,4-dioxane (Aldrich, St. Louis/ Missouri, 99%), sodium hydroxide (Honeywell, Morris Plains/New Jersey, 98.0%), magnesium sulfate (Acros, Geel/Belgium, 99%), methanol (Aldrich, St. Louis/ Missouri, 99%), diethyl ether (Carlo Erba, Milano/Italy, 99.8%), sulfur (Aldrich,


Characterization Gel permeation chromatography (GPC) measurements were performed on a Viscotek GPC max auto sampler system consisting of a pump, light-scattering detector (k 5 670 nm, Model 270 dual detector, Viscotek Corp.) and Viscotek a differential refractive index (RI) detector. The GPC was equipped with three ViscoGEL GPC columns (G2000H HR, G3000H HR, and G4000H HR, 7.8 mm internal diameter, 300 mm length) in series. Tetrahydrofuran (THF) was used as an eluent at flow rate of 1.0 mL min21 at 30 8C. Both detectors were calibrated with polystyrene (PS) standards having narrow-molecular-weight distribution. Data were analyzed using ViscotekOmniSEC Omni-01 software. Inductively coupled plasma emission spectroscopic (ICP-OES) measurements were performed on a Perkin Elmer Optima 2100 DV-Waltham. Scanning electron microscopic (SEM) analysis was performed using FEI Quanta FEG 250. The pH of solutions was measured by pH-meter-Thermo ORION 3 STAR. Synthesis and Characterization of Allylbenzoxazine (BA-ala) Monomer A modified procedure was used46: Allyamine (15.4 g, 0.27 mol) was dissolved in 200 mL of 1,4-dioxane at room temperature in a 250-mL flask. This solution was cooled in an ice bath below 5 8C and then paraformaldehyde (16.3 g, 0.55 mol) was added slowly for 15 min with continuous stirring to prevent smoke formation. Then, bisphenol A (30.8 g, 0.13 mol) was added to the solution. The mixture was refluxed for 24 h. After removal of 1,4-dioxane with a rotary evaporator, the crude product was dissolved in 200 mL of diethyl ether and washed several times with 1 N NaOH to remove phenolic impurities. Then, washed with distilled water for neutralization. After drying with MgSO4, filtering, and solvent evaporation, a sticky oil was obtained. The product was further purified by dissolving it in 25 mL of methanol, followed by addition of water that increased the turbidity of the solution. The solution was cooled in a refrigerator and sticky solid was precipitated. The product was obtained after decanting the solvent. The solid was dried under vacuum at 60 8C for 24 h (yield: 59%). Preparation of Poly(BA-ala-co-sulfur) A modified procedure was used45: To a 20-mL glass tube, equipped with a magnetic stir bar, elemental sulfur and BA-ala were placed (sulfur/BA-ala wt/wt ratios were as follows: 50/50, 60/40, 70/30). The tube was heated up to 185 8C with vigorous stirring in an oil bath. In the first 2 min a clear orange solution is formed and became transparent yellow subsequently. Then, the color of the solution changed to brown-black after 10 min. The overall reaction time was 30 min in total. After cooling the tube to room temperature the solid product was collected. Molecular weights of the poly(BA-ala-co-sulfur)s for 50, 60, 70% S8 amounts were determined by GPC according to polystyrene standards and by using light scattering detector. Mn: 38,650 (Mn/Mw: 2.5), 28,850 (Mn/Mw: 2.9), and 31,090 (Mn/Mw: 2.6), respectively.

45306 (2 of 11)


ARTICLE


mol L21 HCl or NaOH solutions. Then, 0.15 g of the solid sample was added and the final pH was measured after 24 h under agitation at temperature of 25 6 1 8C. The pHpzc was taken as the point at which the curve crossed the line where pH final equals to pH initial. RESULTS AND DISCUSSION Figure 2. Synthesis of BA-ala and sulfur copolymer poly(BA-ala-cosulfur).

Adsorption Process The adsorption studies were performed at different temperatures in the batch mode. Mercury stock solution was prepared by dissolving mercury chloride (HgCl2) in 18.2 MX cm Milli-Q ultra-pure water. A batch solution of 100 mg L21 was prepared from commercial HgCl2. Each working solution was formulated at the experimental concentrations by diluting the batch solution with ultra-pure water. All adsorption experiments were performed in Erlenmeyer flasks and a heated water bath was used where necessary. About 0.1 g poly(BA-ala-co-sulfur) and 20 mL of Hg21 solution was used for each of the adsorption experiments. The pH adjustments were made by adding either 0.1M HCl or NaOH and monitored by a conventional laboratory digital pH meter. The mixture was stirred for 120 min at 200 rpm. At the end of treatment, the adsorbents were recovered and the remaining Hg21 concentration in solution was determined in an inductively coupled plasma optical emission spectrometry (ICP-OES). The effect of solution pH, initial mercury concentration, temperature, and S % ratio in polybenzoxazine were investigated. The adsorption percentage was calculated from the relationship given in eq. (1). Co 2Ce Adsorption ð%Þ5 3100 (1) Co where Co and Ce correspond to the concentration of Hg21 before treatment and after equilibrium, respectively. Point of Zero Charge The zeta-potential measurements were carried out as a function of pHpzc determination. pH of the solution in each flask was adjusted to pH values of 2, 4, 6, 8, 10, and 12 by adding 0.1

Initially, allyl groups bearing bisbenzoxazine monomer,46–49 6,60 (propane-2,2-diyl)bis(3-allyl-3,4-dihydro-2H-benzo[e][1,3]oxazine) (BA-ala), was produced from bisphenol A, allylamine, and paraformaldehyde. In the subsequent step, sulfur-based copolymers were prepared by reacting BA-ala and elemental sulfur in various amounts at 185 8C (Figure 2). At this temperature, S8 forms polysulfide structure which is thermodynamically unstable at room temperature and slowly reverts to cyclooctasulfur (S8). Thus, the direct use of pristine polysulfide is uncommon and limited. However, polysulfides can be stabilized by combination of elemental sulfur with vinylic structures through a radical process, namely inverse vulcanization.50–52 This process can be applied to various vinylic monomers. Therefore, it is appropriate to replace vinylic systems with allyl-containing benzoxazines to form polybenzoxazine-co-sulfides. The allyl groups of BA-ala react with the sulfur radical and generate a CAS covalent bond between the benzoxazine units and polysulfide moiety.37,53 In the meantime, ring-opening of the oxazine takes place by the help of S8 producing both polybenzoxazine and additional CAS bonds from oxazines (ASACH2AN(R) ACH2A)54 (Figure 2). In this way, a stabilized polysulfide as polybenzoxazineco-sulfide appears [abbreviated as poly(BA-ala-co-sulfur)] and mercury (II) sorption capacity of the polybenzoxazine copolymer was determined. The structure and thermal behavior of BA-ala, poly(BA-ala-co-sulfur) was analyzed by spectral and thermal methods, previously.45 Apart from these characterizations, poly(BA-ala-co-sulfur) was morphologically investigated using a high resolution scanning electron microscope. Accordingly, the surface of the copolymer is nonporous and completely smooth indicating the adequate mixing of sulfur and polybenzoxazine without any phase separated zone. Non-porous structure was also evidenced by the nitrogen gas adsorption experiments. These measurements showed only macroporosity with a low surface area. Thus, the sorption process of Hg(II) salts proceeds over these macro surfaces and may limit the overall sorption capacity of the material. However, sorption

Figure 3. SEM images of 50 wt % S containing poly(BA-ala-co-sulfur) before treatment (a), and after treatment (b,c).


45306 (3 of 11)


ARTICLE


Table I. The Independent Variables Examined and Their Experimental Range Range and level Variables 21

Hg

concentration (ppm)

Factor

21

0

11

X1

10

15

20

S % in adsorbent

X2

50

60

70

Temperature (K)

X3

298

315.5

333

pH

X4

4

6

8

experiments revealed that even with the surface area limitation, high sorption capacity was obtained for poly(BA-ala-co-sulfur). The morphological changes of the adsorbents’ surface were examined by using high resolution scanning electron microscope, after the adsorbent had been recovered by centrifugation. As can be from seen in Figure 3, the surface has lost the smoothness and a

rougher surface has developed. This change may be due to the reaction between polysulfides and Hg(II) ions, which break the polysulfide chains and result in a phase separation on the surface. This observation points out that the sorption process occurs through both adsorption and chemisorption pathways. Box–Behnken Design and Adsorption Optimization RSM is a useful technique for the development and optimization of the adsorption process. The main advantages of RSM are the reduced numbers of experimental runs needed to evaluate multiple parameters and their interactions, and its usefulness for developing, improving, and optimizing processes.55,56 A three level BBD with four factors was carried out to define optimal levels of X1 (initial Hg21 concentration), X2 (sulfur % in adsorbent), X3 (temperature), and X4 (pH). Actual and coded values of the independent variables are depicted in Table I. Combinations of initial Hg21 concentration (10,15, 20 ppm), sulfur % in adsorbent

Table II. Experimental and Predicted Adsorption % at Each Experimental Run Experimental design X1

X2

X3

X4

Run number

CHg

S%

T (K)

pH

Experimental Y adsorption %

Predicted Y adsorption % (% error)

1

15

70

333

6

92.04

90.22 (1.97)

2

15

60

333

4

61.10

63.87 (4.53)

3

15

50

315.5

8

73.20

72.58 (0.84)

4

15

60

315.5

6

82.61

81.59 (1.23)

5

15

60

333

8

78.66

80.85 (2.78)

6

15

70

298

6

69.26

70.29 (1.48)

7

10

60

315.5

8

82.74

82.55 (0.22)

8

15

60

315.5

6

80.46

81.59 (1.40)

9

15

60

315.5

6

81.99

81.59 (0.48)

10

20

70

315.5

6

77.92

77.91 (0.01)

11

10

60

315.5

4

62.31

61.89 (0.67)

12

15

50

333

6

87.75

85.57 (2.48)

13

15

60

315.5

6

81.62

81.59 (0.03)

14

20

60

298

6

61.06

61.42 (0.58)

15

15

50

298

6

62.37

63.03 (1.06)

16

20

60

315.5

8

69.29

68.55 (1.06)

17

10

60

333

6

95.44

94.61 (0.86)

18

15

70

315.5

8

75.29

75.78 (0.65)

19

20

60

333

6

84.04

83.92 (0.14)

20

20

50

315.5

6

72.21

73.68 (2.03)

21

15

60

298

4

41.57

41.01 (1.34)

22

10

50

315.5

6

82.27

83.91 (1.99)

23

15

70

315.5

4

59.77

59.92 (0.25)

24

15

50

315.5

4

52.18

51.21 (1.85)

25

10

70

315.5

6

91.43

91.59 (0.17)

26

15

60

315.5

6

81.26

81.59 (0.40)

27

10

60

298

6

74.99

74.64(0.46)

28

15

60

298

8

62.39

61.25 (1.82)

29

20

60

315.5

4

52.94

51.98 (1.81)


45306 (4 of 11)


ARTICLE


Table III. Analysis of Variance Terms for the Proposed Quadratic Model Source

Degrees of freedom

Sum of squares

Mean square

F-value

Model

14

4772.63

340.90

134.91