Reactions between Methanethiol and Biologically Produced Sulfur

Jan 6, 2011 - Paques BV, P.O. Box 52, 8560 AB Balk, The Netherlands. ) UNESCO-IHE, P.O. Box 3015, 2601 DA Delft The Netherlands. ^. Shell Exploration ...
0 downloads 0 Views 867KB Size
ARTICLE pubs.acs.org/est

Reactions between Methanethiol and Biologically Produced Sulfur Particles R. C. van Leerdam,†,‡,* P. L. F. van den Bosch,§ P. N. L. Lens,†,|| and A. J. H. Janssen†,^ †

Sub-department of Environmental Technology, P.O. Box 8129, 6700 EV Wageningen, The Netherlands KWR Watercycle Research Institute, P.O. Box 1072, 3430 BB Nieuwegein, The Netherlands § Paques BV, P.O. Box 52, 8560 AB Balk, The Netherlands UNESCO-IHE, P.O. Box 3015, 2601 DA Delft The Netherlands ^ Shell Exploration and Production International, Kesslerpark 1, 2288 GS Rijswijk, The Netherlands

)



ABSTRACT: Recently, new biotechnological processes have been developed to enable the sustainable removal of organic and inorganic sulfur compounds from liquid and gaseous hydrocarbon streams. In comparison to existing technologies (e.g., caustic scrubbing or iron based redox technologies) far less chemicals are consumed, while reusable elemental sulfur is formed as the main end-product. This research shows that in these processes a number of consecutive reactions occur between methanethiol (MT) from the hydrocarbon stream and the formed biosulfur particles, leading to the formation of (dimethyl) polysulfides. This is an important feature of this family of new bioprocesses as it improves the MT removal efficiency. The reaction kinetics depend on the MT and biosulfur concentration, temperature, and the nature of the biosulfur particles. The first reaction step involves a S8 ringopening by nucleophilic attack of MT molecules to form CH3S9-. This work shows that CH3S9- reacts to polysulfides (S32-, S42-, S52-), dimethyl polysulfides [(CH3)2S2, (CH3)2S3], and dissociated H2S, while also some longer-chain dimethyl polysulfides [(CH3)2S4-7] are formed at μM levels. Control experiments using orthorhombic sulfur flower (S8) did not reveal these reactions.

’ INTRODUCTION Removal of volatile sulfur compounds like H2S and shortchain thiols (methanethiol, ethanethiol and propanethiol) from hydrocarbon streams such as natural gas and liquefied petroleum gas (LPG) is needed for reasons of toxicity, environmental protection, and corrosivity. As an alternative for the currently applied physical-chemical treatment methods, Van Leerdam et al.1 described a biological process for LPG desulfurization consisting of four integrated process steps. After extraction under mildly alkaline conditions, MT is degraded in an anaerobic bioreactor to form H2S, CH4, and CO2 by methanogenic archaea. Hereafter, sulfide-oxidizing bacteria oxidize the formed H2S to elemental biosulfur particles that are removed in a gravity settler. From previous research it is known that these particles exhibit hydrophilic properties.2,3 Van Leerdam et al.4 found that part of the MT, ethanethiol, and propanethiol may not be degraded in the anaerobic reactor. Hence, thiols also enter the aerobic stage, where a spontaneous reaction with biosulfur particles will occur. In our experimental work, MT was used as a model compound. Although the oxidation of thiols by orthorhombic sulfur (S8) crystals to (organic) polysulfides has been studied in the presence of a base-catalyst,5 no literature could be found on the reactions between biosulfur particles and MT. Thiolates (RS-) are strong nucleophiles, which are able to open any sulfur-sulfur single bond in a polysulfane.6 Methylthiolate, the dissociated form of MT, reacts with elemental sulfur to form methyl polysulfide (eq 1) CH3 S - þ S8 / CH3 S9 r 2011 American Chemical Society

ð1Þ

The formed methyl polysulfides are however unstable, and evi- 6,7 dence was only found for the existence of RS and RS2 . The ring-opening (eq 1) is the rate determining step. In subsequent reactions, shorter dimethyl polysulfides are formed along with the formation of polysulfides (eq 2) CH3 S - þ CH3 S9 - / CH3 Sx CH3 þ Sy 2 -

ð2Þ

with x þ y = 10. The molar ratio of thiols to sulfur partly determines the range of reaction products. For instance, Jocelyn 8 found that dimethyl disulfide (DMDS) and H2S are formed if sulfur is present in limited concentrations (eq 3), while at excess amounts longer chain polysulfides and dimethyl polysulfides are the main endproducts (eq 4) 2RS - þ Hþ þ 1=8S8 / HS - þ RSSR 2RS - þ S8 / Sy 2 - þ RSx Rðsum of reactions1and2Þ

ð3Þ ð4Þ

Reactions 1-4 proceed conveniently at temperatures up to 60 C. Dimethyl polysulfide with ten or more S atoms are however very unstable9 and tend to decompose by equilibrium reactions with other chain lengths or by the formation of elemental sulfur, even after pure organic polysulfides have been Received: June 15, 2008 Accepted: December 16, 2010 Revised: December 5, 2010 Published: January 6, 2011 1320

dx.doi.org/10.1021/es102987p | Environ. Sci. Technol. 2011, 45, 1320–1326

Environmental Science & Technology

ARTICLE

Figure 1. Particle size distribution of the dialyzed biosulfur particles (dashed line: sample 1, continuous line: sample 2).

obtained. Hence, species with a sulfur chain length between 2 and 8 can be expected6 (eqs 5 and 6) 2RSx R / RSx þ n R þ RSx - n R

ð5Þ

RSx R / RSx - n R þ n=8S8

ð6Þ

The establishment of the equilibria 5 and 6 can be accelerated by light, heat, and numerous catalysts, of which strong nucleophiles (e.g., RS-, HS-, Sx2-) are the most effective.10 Obviously, after the reaction between MT and sulfur has proceeded for some time, a mixture of elemental sulfur, (di)methyl polysulfides, polysulfides, and other nucleophilic species is formed. The objective of this paper is to gain more insight in the selectivity and kinetics of the reactions between MT and biosulfur particles that are formed in biological desulfurization processes.

’ MATERIAL AND METHODS Elemental Sulfur. Freshly prepared biosulfur samples were obtained from a full-scale H2S-oxidizing bioreactor located at the wastewater treatment plant of Industriewater Eerbeek (Eerbeek, The Netherlands). Samples were collected at two different moments in time. It is known that the surface of these particles is negatively charged at neutral to alkaline pH,11 and these particles exhibit hydrophilic properties, resulting from the presence of amphiphilic compounds that are attached to the surface.2,3 Immediately after sampling, the biosulfur suspensions were repeatedly dialyzed to remove dissolved salts (mainly sodium bicarbonate, sodium sulfate, and sodium thiosulfate) until the specific conductivity of the suspension was below 40 μS/cm. The biosulfur concentration in the stock solution was 30 g/L and 16 g/L, respectively. The stock solutions were used for experiments without further filtration or drying. Particle sizes in sample 1 ranged from 0.01 to 100 μm, while in sample 2 all particles were smaller than 33 μm (Figure 1). The differences in particle size distribution can, most likely, be attributed to small fluctuations in the operating conditions of the wastewater treatment plant. The overall specific surface area (As) of the biosulfur particles was 6.9 m2/g (sample 1) and 10.2 m2/g (sample 2). In the experiments, biosulfur concentrations varied between 1 and 16 mM of sulfur atoms (32 g/mol). Hence, the initial concentration of the

reactive surface (Ac) varied between 221 and 3533 m2/m3 (sample 1) and 327 and 5234 m2/m3 (sample 2). Experimental Setup. Batch tests were performed in continuously shaken (120 rpm, 30 C) 120- and 250-mL flasks. The dialyzed biosulfur suspension was added to a deaerated buffer solution (20 g/L NaHCO3) where after the flasks were flushed with N2 gas and sealed with Viton stoppers. In thermostatted (30-60 C) 250- and 500-mL round-bottom flasks, the biosulfur suspension was mixed with a 200 mL (bi)carbonate buffer solution and then nitrogen flushed. Deaerated solutions of 20 g/L NaHCO3 (ionic strength I = 0.24 M) were used as buffer solutions (final pH = 8.7). At the start of the experiment, a concentrated MT solution was supplied to the flask through a septum to prevent oxygen intrusion. During the course of the experiment (5-25 min), the liquid content was continuously circulated over a flow-through quartz cuvette (path length 1 cm) while the UV absorbance (λ = 285 nm) was recorded once every 4 s. Some experimental runs lasted up to 18 h to assess any long-term changes. Analysis. Undiluted samples taken from the batch flasks were analyzed for the presence of MT and dimethyl polysulfides by HPLC equipped with a 20 cm Chrompack C-18 column. The oven temperature was controlled at 30 C. The eluent flow rate was 0.6 mL/min. The injection volume of the samples was 20 μL, and an ultraviolet detector was used to monitor MT and dimethyl polysulfides simultaneously at wavelengths of 220, 230, 265, and 285 nm. For MT analyses the eluent consisted of 40% methanol and 60% water (method A). To avoid very long retention times, higher dimethyl polysulfides were measured with an eluent consisting of 70% methanol and 30% water (method B). The peaks in the HPLC chromatogram were identified by using solutions of pure reference compounds, i.e. MT, DMDS, and dimethyl trisulfide (DMTS). Peaks of other compounds were identified by relying on the dependency of the retention time of dimethyl polysulfides on the number of sulfur atoms in the molecule. A strong correlation exists between the logarithm of the chromatographic retention time, ln(Rt), and the number n of sulfur atoms in the molecule (R-Sn-R, where R represents an alkyl group).12,13 Moreover, the dimethyl polysulfides could also be identified in the HPLC chromatogram by their characteristic UV spectra between 200 and 400 nm. Concentrations of MT, DMDS, and DMTS were calculated after calibration with standard solutions. More complex dimethyl polysulfides with four or more sulfur atoms are less stable and not commercially available. Therefore, concentrations of dimethyl polysulfides from CH3S4CH3 up to CH3S8CH3 were calculated from calibration curves of a K2S5 solution, synthesized according to the method described by Pearson and Robinson.14 The K2S5 solution disproportionates to an equilibrium mixture with a known distribution of S32- up to S82- polysulfide ions. After methylation of these polysulfides with methyl trifluoromethanesulfonate (methyl triflate) to dimethyl polysulfides, the individual concentrations were related to the peak areas on a HPLC chromatogram as shown by Kamyshny et al. and Rizkov et al.12,13 This rapid chemical methylation of polysulfides is normally used to determine polysulfide concentrations. The K2S5 solution and reaction samples (0.1 mL) were dissolved in 0.8 mL of methanol together with 0.1 mL of methyl triflate. The polysulfides are rapidly methylated by the methyl triflate and measured by HPLC as dimethyl polysulfides. Unfortunately, this method is not suitable to distinguish between Sx2- and CH3Sx- as both are methylated to CH3SxCH3. 1321

dx.doi.org/10.1021/es102987p |Environ. Sci. Technol. 2011, 45, 1320–1326

Environmental Science & Technology A disadvantage of HPLC analyses is that rapid changes in reaction medium may not be observed. Hence, spectrophotometric measurements were performed to record semionline changes in the medium’s composition. In batch experiments the UV absorbance was determined at a wavelength of 285 nm, representing a sum parameter for the produced polysulfides and (di)methyl polysulfides.11 MT and biosulfur solutions exhibit a low absorption rate, i.e. A285 is less than 0.1. It has been assumed that an initial increase in absorbance rate is related to changes in the presence of CH3S9- ions as these are the first reaction products. The total sulfide concentration includes free sulfide and polysulfides.15 Immediately after sampling, the sample (1 mL) was added to 1 mL of zinc acetate (20 g/L) to prevent oxidation of sulfide. The formed precipitate was diluted with demineralized water. The total sulfide concentration in the diluted zinc sulfide precipitate was measured with the Dr. Lange cuvette test LCK653 (Hach Lange, Germany). Residual biosulfur concentrations were determined by measuring the dry weight after filtration over a 0.45 μm filter and drying at 50 C for 24 h. The particle size distribution was determined by laser scattering image analysis (Coulter laser LS 230). Calculation of Reaction Rate. The initial slope of the absorbance curve (ΔA285/Δt) was used to calculate the rate of the reaction between biosulfur and MT (eq 1). The initial slope was calculated in the linear part of the curve, i.e. after passing a 30 s “dead-time” slot. The concentration of CH3S9- has been estimated by applying the molar extinction coefficient for Sx2- (ε = 1390 L mol-1 cm-1, pH 8.0). The coefficient is based on the socalled polysulfide excess sulfur concentration. Because of the poor solubility of biosulfur particles, the reaction will essentially take place at the surface of the sulfur particles.11 The kinetics of this heterogeneous reaction depend on the available surface area. Chemicals. A 3.0 M sodium methylthiolate (NaCH3S) solution supplied by Arkema (Rotterdam, The Netherlands) was used as the MT-containing reagent. The solution was made from gaseous MT, dissolved in a high-purity NaOH solution. The gas contained only a few impurities, i.e. < 0.7% methanol, < 0.3% DMS and trace amounts of DMDS. NaHCO3 and Na2CO3 (>99.7% pure) and liquid DMDS, DMTS (all with a purity of more than 98.5%) were purchased from Merck (Darmstadt, Germany). Methyl trifluoromethanesulfonate (methyl triflate) was purchased from Acros Organics, Pittsburgh, PA. Pure K2S5 was kindly provided by Dr. A Kamyshny (Max Planck Institute, Bremen). The specific absorption coefficient of DMDS at 285 nm is 95 L mol-1 cm-1 (data not shown), which is considerably lower than the specific absorption coefficient of polysulfide at 285 nm (1390 L mol-1 cm-1).16

’ RESULTS AND DISCUSSION Determination of Reaction Products. Three batch experiments (30 C) were performed at different MT/S0 ratios, i.e. 1.0 (Exp. A), 0.5 (Exp. B), and 0.25 mol MT/mol S0 (Exp. C) (Table 1). In all experiments, the reactions went to completion as no residual MT and biosulfur particles could be detected after 45 min. After this period, the reaction mixture consists of dissolved sulfides (HS- and Sx2-), DMDS (CH3S2CH3), and DMTS (CH3S3CH3) (Figure 2). Also a number of long-chain dimethyl polysulfides (CH3S4-7CH3) were formed at trace levels (Table 1). In experiment A, the solution colored yellowish during the first 10 min of the experimental run where after it turned colorless.

ARTICLE

Table 1. Final Product Concentrationa in Batch Experiments (pH 8.7, 30C, Biosulfur Particles of Sample 1) reagents

Exp. A

Exp. C

methanethiol (mM)

8.0

8.0

4.0

biosulfur (mM S0)

8.0

16

16

Exp. A

Exp. B

Exp. C

MeS2Me (mM)

2.53

1.96

0.98

MeS3Me (mM)

1.18

3.04

2.1

MeS4Me (mM)

0.06

0.16

0.16

MeS5Me (mM)

0.01

0.04

0.04

MeS6Me (mM)

0.002

0.006

0.01

MeS7Me (mM)

n.d.

0.001

0.001

total sulfide (mM) CH3S2- or S22-

3.50 n.a.

3.34 n.d.

1.59 n.d.

CH3S3- or S32-

n.a.

þ

þ

CH3S4- or S42-

n.a.

þ

þ

CH3S5- or S52-

n.a.

þ

þ

CH3S6- or S62-

n.a.

n.d.

þ

products

a

Exp. B

MeSxMe was determined by HPLC, total sulfide (HS- and Sx2-) was determined by the Dr. Lange cuvette test, and CH3Sx- or Sx2was determined by HPLC after methylation with triflate. n.d. not detected; n.a. not analyzed; þ detected, but only qualitative analysis possible.

Monomethyl polysulfides (CH3Sx-) and polysulfides (Sx2-) are known for their yellow color5,17 and dimethyl polysulfides (CH3SxCH3) are colorless which indicates that first (monomethyl) polysulfides are formed (eqs 1 and 2) which are further methylated, according to Sx 2 - þ 2CH3 SH / CH3 Sx CH3 þ 2HS -

ð7Þ

The total sulfide concentration in the equilibrium mixture at the end of experiment A can be solely attributed to the presence of HS- ions. No polysulfides are present because this would lead to an optically clear yellow solution at concentrations above 0.04 mM S52-. In contrast to experiment A, the yellow color formed during the first 5 min remained until the end of experiment B and C, indicating that (monomethyl) polysulfides remained present during the entire experimental period (120150 min). Table 1 shows that the DMDS/DMTS ratio in the equilibrium mixture is influenced by the initial MT to S0 ratio: higher MT/S0 ratios result in a higher selectivity for DMDS over DMTS (A > B > C). According to Vineyard,5 high dialkyl disulfide and dialkyl trisulfide yields can be obtained at thiol/S0 ratios above 2.5 and 1.25 mol/mol, respectively. The polarity of the solvent, the reaction time, and the temperature play a role in the selectivity of the product formation. If the thiol/S0 ratio is appreciably less than 1.25, like in our experiments, a mixture of (dimethyl) polysulfides is formed. Our results indeed confirm that at MT/ S0 ratios below 1 mol/mol a mixture of (dimethyl) polysulfides is formed, while dimethyl tetrasulfide and higher dimethyl polysulfides are only formed in trace amounts. Any formation of dimethyl tetrasulfide (and higher dimethyl polysulfides) is not expected from the reaction between sulfur and MT.18 The initial MT concentration determines the capacity to methylate polysulfide ions, resulting in the formation of dimethyl polysulfides in the reaction mixture at equilibrium conditions. 1322

dx.doi.org/10.1021/es102987p |Environ. Sci. Technol. 2011, 45, 1320–1326

Environmental Science & Technology

Figure 2. Formation of (dimethyl) polysulfides (30 C, pH 8.7) in a mixture of 16 mM S0 (sample 1) and 4 mM MT (experiment C); -*- MT, -(- total dissolved sulfide (HS- þ Sx2-), -9- DMDS, -2- DMTS, -  DMTeS (CH3S4CH3).

From Table 1, it can be seen that in experiments A and B (initial MT concentration 8 mM) about equal amounts of sulfide (S2-tot) are formed. However, in experiment A more free HS- must have been present, while in experiment B the polysulfide concentration must have been higher because of the subsequent reaction of the formed HS- with residual elemental sulfur particles. In experiment C, the initial MT concentration is 50% lower than in experiments A and B. As a result about 50% lower sulfide concentrations (HS- and Sx2-) were recorded in the reaction mixture. Polysulfides and monomethyl polysulfide were detected by HPLC analyses following a dedicated pretreatment step as described by Kamyshny et al.12 This method was used at the end of experiments B and C. Diluted samples were treated with methyl triflate to convert all polysulfides and monomethyl polysulfide ions to their corresponding dimethyl polysulfides. From a comparison of the HPLC chromatograms before and after methyl triflate treatment, it could be concluded that S32-, S42-, S52-, and S62- or their methyl derivates (CH3Sx-) were formed in experiments B and C (Table 1). Effect of the Biosulfur Concentration on the Initial Reaction Rate. Figure 3 shows the effect of the biosulfur concentration on the absorbance curves at 50 C. The biosulfur concentration ranged from 2 to 16 mM (Ac = 442 to 3533 m2/m3), while MT was kept constant at 8 mM in all experiments. It can be seen that after 30 s the absorbance (A285) rapidly increased and started to decrease after 3 min to reach a plateau value. At increasing sulfur concentrations higher plateau values were found. Based upon experiments A and B (Table 1) it is concluded that at higher biosulfur concentrations, higher (dimethyl) polysulfides concentrations are present at the plateau value (equilibrium). It is most likely that the increase in absorbance is the combined effect of the formation of methyl polysulfides (eq 1) and the subsequent formation of (dimethyl) polysulfides (eqs 2-4). Hereafter, shortchain (dimethyl) polysulfides are formed, with a lower specific absorbance than the initially formed long chain polysulfur compounds.12 Lower initial biosulfur concentrations result in lower concentrations of products and in relatively more short-chain (dimethyl) polysulfides (e.g., DMDS) in the reaction mixture (Table 1). After 18 h, similar absorbance values were measured as after 15 min (data not shown), indicating that an equilibrium situation has been reached. Table 2 shows the effect of the biosulfur concentration on the initial reaction rate at a range of temperatures (30-60 C). This is important information because the new desulfurization processes can be applied to treat hot gases, such as H2S-containing

ARTICLE

Figure 3. Effect of the initial biosulfur (sample 1) concentration [in mM: 16 (), 8 (Δ), 4 (9), and 2 ())] on the absorbance at 285 nm. MT0 = 8.0 mM, pH 8.7, 50 C.

syngas streams from coal fired gasification units. It can be seen that the reaction rate depends on the MT and biosulfur concentration and on the temperature. At 50 and 60 C, all biosulfur particles (sample 1) disappeared within 20 min after addition of MT. In the experiments carried out at 30 and 40 C and high biosulfur concentrations (8 and 16 mM), biosulfur (sample 1) particles were still present. Mixtures containing 4 and 8 mM biosulfur colored yellowish during the experiment, indicating the presence of polysulfides. Higher reaction rates were found when using biosulfur particles from sample 2. Obviously, the size of the biosulfur particles plays an important role in the reaction rates. The larger particles in sample 1 (Figure 1) have a lower specific surface area than those in sample 2. Hence, the reaction rate in the former case will be lower.11 A doubling in biosulfur concentration results in a doubling of the CH3S9- formation rate at low biosulfur concentrations and temperatures, e.g. biosulfur sample 1 at 30 C, 8 mM MT and 2, 4, and 8 mM biosulfur (Table 2). A comparison between biosulfur samples 1 and 2 (Table 2) at 8 mM MT and 4 mM biosulfur (i.e., 24.2 m2/m3 and 63.3 m2/m3, respectively, ratio Ac2/Ac1 = 2.6) shows that the CH3S9- formation rate with biosulfur sample 2 is on average 4 times faster at all temperatures tested. Effect of the MT Concentration on the Initial Reaction Rate. The effect of MT on the initial reaction rate was investigated in the 30-60 C temperature range (Table 2). As an example, absorbance curves at 30 C and pH 8.7 are shown in Figure 4 (biosulfur particles from sample 1). The MT concentration ranged from 4 to 16 mM, while biosulfur was kept constant at 8 mM. As expected for nonzero order reaction kinetics, higher MT concentrations result in higher reaction rates. Higher MT concentrations not only increase the initial reaction rate but also the subsequent reactions. At the highest MT concentration, the absorbance curve rapidly levels off to a lower plateau value (Figure 4). This behavior differs from the behavior of the reaction between biosulfur and H2S as found by Kleinjan et al.11 and can be explained by the formation of short-chain (dimethyl) polysulfides (e.g., DMDS and DMTS) from the initially formed long-chain (dimethyl) polysulfides. The mixture containing 16 mM MT (Figure 4) became clear and colorless after 20 min. This indicates that all biosulfur particles disappeared and dimethyl polysulfides have been formed. However, the mixtures containing 4 and 6 mM MT turned yellowish and still contained some biosulfur particles in suspension. In combination with the high A285 values, this is an indication that polysulfides were still present. 1323

dx.doi.org/10.1021/es102987p |Environ. Sci. Technol. 2011, 45, 1320–1326

Environmental Science & Technology

ARTICLE

Table 2. Estimation of the Initial CH3S9- Formation Rate Rx (μmol CH3S9- L-1 s-1) from the Reaction between Biosulfur and MT (eq 1) Rx (30 C)

Rx (40 C)

Rx (50 C)

Rx (60 C)

MT (mM)

biosulfur (mM)

MT/S0 ratio (mol/mol)

4

8

0.5

0.10

-a

0.39

-a

6 16

8 8

0.75 2

0.35 2.9

-a -a

1.12 4.93

-a -a

8

2

4

0.13

0.38

0.63

1.21

8

4

2

0.27

0.70

1.02

1.59

8

8

1

0.57

1.09

2.25

3.33

8

16

0.5

0.72

1.63

3.25

4.78

Biosulfur Sample 1

Biosulfur Sample 2

a

2

4

0.5

0.19

0.50

0.91

0.93

4

4

1.0

0.70

1.09

1.41

3.34

8 12

4 4

2.0 3.0

1.30 1.66

2.22 3.90

4.50 9.61

6.41 21.1

16

4

4.0

3.61

10.1

14.4

18.1

4

2

2.0

0.30

0.47

1.06

1.33

4

8

0.50

1.53

2.82

3.48

7.31

4

12

0.33

2.47

3.93

5.90

12.0

4

16

0.25

2.58

4.48

6.52

19.9

-: not measured.

Figure 4. Effect of the MT concentration [in mM: 16 ()), 8 (9), 6 (Δ), and 4 (O)] on the absorbance at 285 nm. S0 = 8.0 mM (sample 1), pH 8.7, 30 C.

Proposed Reaction Mechanism. The overall reactions (eqs 3 and 4) between thiols and elemental sulfur were first described some decades ago.5,8 Alkyl tri- and tetrasulfides can be prepared from thiols and elemental sulfur in the presence of catalytic amounts of n-butylamine at 25-63 C.7 However, the reactions described in the current paper are not based on an amine catalyst but on the reaction of MT with biologically produced sulfur particles. The molar thiol/S0 ratio, the polarity of the solvent, the reaction time, the temperature, and the pH as well as the nature of the rest group of the thiol determine the products formed.5,18 In our experiments a pH of 8.7 was applied. At this pH, MT is predominantly (98%) present in its molecular form (pKa = 10.3). An increase in pH results in a higher dissociated fraction resulting in higher reaction rates in the reaction with biosulfur because methylthiolate is a stronger nucleophile than molecular MT. In control experiments with hydrophobic orthorhombic sulfur crystals in a NaHCO3 buffer solution, no conversion was observed because the sulfur particles were dissipated from the

aqueous phase and accumulated, against gravity forces, at the water-air interface. As a result of the hydrophilic properties the biosulfur particles are wetted and therefore completely dispersed in the aqueous phase, which enables a rapid reaction with MT. The presence of polythionates, i.e. sulfonic groups, lead to the overall hydrophilic character, while at alkaline conditions, like in our system, dissociated organic molecules such as carbohydrates cause the hydrophilic character.3 After the reaction between MT and biosulfur proceeded for about 15 to 20 min, a complex mixture of (dimethyl) polysulfides and nucleophilic species is obtained. Besides MT, also other thiophiles (CH3Sx-, Sx2-) may attack the sulfur rings. The relative strength of the various nucleophiles (Sx2- > CH3S- > HS-) determines the position of the equilibrium and is given by their thiophilicity.6 The following reaction mechanism is proposed to explain the formation of the observed end-products in Table 1. After the ring-opening of the S8 crystal by a nucleophilic attack of the MT ion or molecule (eq 1), in a second step, the unstable CH3S92rapidly reacts with MT to form shorter (di)methyl polysulfides CH3 S - þ CH3 S9 - / CH3 S5 CH3 þ S5 2 -

ð9Þ

CH3 S - þ CH3 S9 - / 2CH3 S5 -

ð10Þ

S52-

(eq 9) are more stable products than CH3S5CH3 and CH3S5- (eq 10). Although there is no evidence for the existence of methyl polysulfides that are longer than CH3S2-,7,19 CH3S5can be formed as an intermediate. In subsequent reaction steps, (di)methyl polysulfides and polysulfides of shorter chain length can be formed according to the following reactions CH3 S5 CH3 þ CH3 S - / CH3 S3 - þ CH3 SSSCH3

1324

ð11Þ

S5 2 - þ CH3 S - / CH3 S3 - þ S3 2 -

ð12Þ

S5 2 - þ CH3 SH / CH3 S5 - þ HS -

ð13Þ

dx.doi.org/10.1021/es102987p |Environ. Sci. Technol. 2011, 45, 1320–1326

Environmental Science & Technology

ARTICLE

Table 3. Empirical Reaction Rate Coefficients (Log k) of the Reaction between MT and Biosulfur for Samples 1 and 2

Table 4 acronym

full name

chemical formula

temperature (C)

sample 1

sample 2

MT

methanethiol

CH3SH

30

-5

-6

ET

ethanethiol

CH3CH2SH

40 50

-4 -4

-6 -6

PT

propanethiol

CH3CH2CH2SH

-

methylthiolate

CH3S-

60

-3

-5

DMDS

dimethyl disulfide

CH3SSCH3

DMTS

dimethyl trisulfide

CH3S3CH3

DMTeS

dimethyl tetrasulfide

CH3S4CH3

dimethyl polysulfide organic polysulfide

CH3SxCH3 RSxR

-

þ CH3 S5 CH3

ð14Þ

CH3 S5 - þ CH3 S - / CH3 SSCH3 þ S4 2 -

ð15Þ

-

CH3 S3 - þ CH3 S - / CH3 SSCH3 þ S2 2 -

ð16Þ

-

methyl polysulfide

CH3Sx-

-

polysulfide

Sx2-

CH3 S5

þ CH3 SH / HS

-

CH3 S3 - þ CH3 SH / CH3 SSSCH3 þ HS Sn 2 - þ Hþ / HS - þ ðn - 1Þ=8S8

ð17Þ ð18Þ

After the S8 ring-opening, subsequent reactions (eqs 9-18) already take place before all biosulfur is allowed to react which apparently slows down the increase in A285. The nucleophile CH3SH can break sulfur chains at other places than indicated in eqs 9-17 resulting in a variety of products with a different sulfur chain length. S22- (eq 16) was not detected in experiments B and C (Table 1), either because the equilibrium of this reaction is very much to the left-hand side or S22- reacted fast with other sulfur compounds. The short chain S22- and S32- are the strongest polysulfide nucleophiles and thus the most reactive ones.20 They have only been observed at very high alkalinities (pH > 14).16,21 In contrast to these references, CH3S32- or S32- was formed at pH 8.7 in our experiments (Table 1). Determination of the Reaction Orders and Rate Constants. The rate (Rx) of the heterogeneous reaction between biosulfur and MT (eq 1) can be described by the general rate law     d½CH3 SH dAc Rx ¼ ¼ dt dt t ¼ 0 t¼0 ¼ k½CH3 SHR Ac β

ð19Þ

Because of the poor solubility of the biosulfur particles, the reaction will essentially take place at the surface of the particles (heterogeneous reaction). In expression 19, the concentration of the biosulfur surface, Ac (m2 m-3) is therefore used. Reaction orders R and β were determined for both biosulfur samples, using the differential method by plotting the initial reaction rates and concentrations from Table 2 on a log-log scale (data not shown). The empirical reaction orders appeared to be 2.1 ( 0.4 (R) and 0.76 ( 0.07 (β) for sample 1 and 1.38 ( 0.08 (R) and 1.11 ( 0.11 (β) for sample 2. From the reaction orders R and β the empirical reaction rate coefficients were calculated (Table 3). The unit of k is (mol/L)-1,1 m0,76 s-1 in the case of biosulfur sample 1 and (mol/L)-0,38 m1,11 s-1 in the case of biosulfur sample 2. Differences in the reaction coefficients (R and β) and the log k values between both samples are most likely caused by the polydisperse character of biosulfur samples. This research shows that biologically produced sulfur particles rapidly react with MT to form complex mixtures, consisting of (dimethyl) polysulfides. Biosulfur sampled from the full-scale wastewater treatment plant at different time points can differ in particle size distribution, probably due to small fluctuations in the

operating conditions which can lead to different reaction rates in the reaction with MT. The products formed from this heterogeneous reaction depend not only on the MT and biosulfur concentration but also on the MT/bio-S ratio. The special hydrophilic properties of biosulfur are of importance since it allows a simultaneous removal of thiols and H2S from sour gas streams.

’ AUTHOR INFORMATION Corresponding Author

*Phone: þ 31 6 129 155 43; e-mail: [email protected].

’ ACKNOWLEDGMENT The authors thank Dr. A Kamyshny (Max Planck Institute, Bremen) for his help in developing the HPLC method for measuring dimethyl polysulfides and for instructive discussions. ’ REFERENCES (1) Van Leerdam, R. C.; Bonilla-Salinas, M.; De Bok, F. A. M.; Bruning, H.; Stams, A. J. M.; Lens, P. N. L.; Janssen, A. J. H. Anaerobic methanethiol degradation and methanogenic community analysis in an alkaline (pH 10) biological process for Liquefied Petroleum Gas desulfurization. Biotechnol. Bioeng. 2008, 101 (4), 691–701. (2) Janssen, A. J. H.; De Keizer, A.; Van Aelst, A.; Fokkink, R.; Yangling, H.; Lettinga, G. Surface characteristics and aggregation of microbiologically produced sulphur particles in relation to the process conditions. Colloids Surf., B 1996, 6 (2), 115–129. (3) Kleinjan, W. E.; de Keizer, A.; Janssen, A. J. H. Biologically produced sulfur. Top. Curr. Chem. 2003, 230, 167–187. (4) Van Leerdam, R. C.; De Bok, F. A. M.; Lomans, B. P.; Stams, A. J. M.; Lens, P. N. L.; Janssen, A. J. H. Volatile organic sulfur compounds in anaerobic sludge and sediments: biodegradation and toxicity. Environ. Toxicol. Chem. 2006, 25 (12), 3101–3109. (5) Vineyard, B. D. Versatility and mechanism of n-butylaminecatalyzed reaction of thiols with sulfur. J. Org. Chem. 1967, 32 (12), 3833–3836. (6) Steudel, R. The chemistry of organic polysulfanes R-Sn-R (n > 2). Chem. Rev. 2002, 102, 3905–3945. (7) Steudel, R.; Kustos, M. Sulfur: organic polysulfanes. In Encyclopedia of inorganic chemistry; King, R. B., Ed. Wiley: Chichester, 1994; Vol. 7, pp 4009-4038. (8) Jocelyn, P. C. Chemical reactions of thiols. In Biochemistry of the SH group; Academic Press Inc.: London, 1972; pp 2, 85-87. (9) Clark, P. D.; Derdall, G.; Lesage, K. L.; Hyne, J. B. The formation of dimethylpolysulphides from sulphur and dimethyldisulphide. Alberta Sulphur Research Ltd. Q. Bull. 1983, 12–25. 1325

dx.doi.org/10.1021/es102987p |Environ. Sci. Technol. 2011, 45, 1320–1326

Environmental Science & Technology

ARTICLE

(10) Clark, P. D.; Fitzpatrick, E.; Lau, C. S. C.; Oriakhi, C. O. Studies on the stability of dimethyl disulfide-polysulfide mixtures under sour gas pressure. Energy Fuels 1993, 7 (1), 155–158. (11) Kleinjan, W. E.; de Keizer, A.; Janssen, A. J. H. Kinetics of the reaction between dissolved sodium sulfide and biologically produced sulfur. Ind. Eng. Chem. Res. 2005, 44 (2), 309–317. (12) Kamyshny, A.; Goifman, A.; Gun, J.; Rizkov, D.; Lev, O. Equilibrium distribution of polysulfide ions in aqueous solutions at 25C: a new approach for the study of polysulfides equilibria. Environ. Sci. Technol. 2004, 38 (24), 6633–6644. (13) Rizkov, D.; Lev, O.; Gun, J.; Anisimov, B.; Kuselman, I. Development of in-house reference materials for determination of inorganic polysulfides in water. Accredit. Qual. Assur. 2004, 9 (7), 399–403. (14) Pearson, T. G.; Robinson, P. L. J. Chem. Soc. 1930, 133, 1473– 1497. (15) Van den Bosch, P. L. F.; Van Beusekom, O. C.; Buisman, C. J. N.; Janssen, A. J. H. Sulfide oxidation under halo-alkaline conditions in a fed-batch bioreactor. Biotechnol. Bioeng. 2007, 97 (5), 1053–1063. (16) Kleinjan, W. E.; De Keizer, A.; Janssen, A. J. H. Equilibrium of the reaction between dissolved sodium sulfide and biologically produced sulfur. Colloids Surf., B 2005, 43 (3-4), 228–237. (17) Steudel, R. Mechanism for the formation of elemental sulfur from aqueous sulfide in chemical and microbiological desulfurization processes. Ind. Eng. Chem. Res. 1996, 35, 1417–1423. (18) Vineyard, B. D. Mercaptan-sulfur reaction. Alkyl trisulfides. J. Org. Chem. 1966, 31 (2), 601–602. (19) Steudel, R. The chemical sulfur cycle. In Environmental technologies to treat sulfur pollution - principles and engineering; Lens, P. N. L., Hulshoff Pol, L. W., Eds.; IWA Publishing: London, 2000; pp 1-31. (20) Steudel, R. Inorganic polysulfides Sn2- and radical anions Sn•-. Top. Curr. Chem. 2003, 231, 127–152. (21) Giggenbach, W. Optical spectra and equilibrium distribution of polysulfide ions in aqueous solution at 20. Inorg. Chem. 1972, 11 (6), 1201–1207.

1326

dx.doi.org/10.1021/es102987p |Environ. Sci. Technol. 2011, 45, 1320–1326