Waste and Biomass Valorization https://doi.org/10.1007/s12649-018-0226-9

ORIGINAL PAPER

Effect of Operating Conditions on Separation of ­H2S from Biogas Using a Chemical Assisted PDMS Membrane Process Ebrahim Tilahun1 · Erkan Sahinkaya2 · Bariş Çalli1 Received: 28 October 2017 / Accepted: 5 February 2018 © Springer Science+Business Media B.V., part of Springer Nature 2018

Abstract Hydrogen sulfide (­ H2S) is an undesirable impurity that has to be removed from biogas to avoid the corrosion of co-generation units. In the present study, we evaluated the potential of a gas–liquid membrane contactor process for selective removal of ­H2S from biogas. The effects of biogas retention time (GRT), membrane thickness and liquid absorbent pH were investigated. A dilute sodium hydroxide solution was used as absorbent. The results revealed that ­H2S removal efficiency (RE) improved with increasing GRT and absorbent pH, and decreased with increasing membrane thickness. When GRT reduced from 19 to 3.4 min, the RE of H ­ 2S and ­CO2 decreased by over 2.5 and 5.2 times, respectively. In contrast, a higher desulfurization selectivity was observed with lower GRT and thicker membranes. The ­CH4 content of the treated biogas increased along with increasing GRT and was enriched from 60% to a maximum of 87% with only 4.68% loss. The SEM–EDS analysis confirmed the deposition of inorganics such as Ca, Mg, S and Si on the membrane surface. However, any membrane clogging and fouling problem was not observed. In summary, the novel gas–liquid polydimethylsiloxane membrane contactor tested in this study has performed well in selective removal of ­H2S from biogas and is expected to be a promising alternative to conventional desulfurization processes. Keywords  Biogas · Desulfurization · Inorganics deposition · CH4 content · Membrane separation

Introduction Biogas produced from anaerobic digestion of organic wastes is a favorable energy source that plays a vital role in the emerging renewable energy market, because of its substrate flexibility and energy yield [1]. Biogas composition mainly depends on the raw materials used and process conditions. Biogas consists of mainly methane (­ CH4) and carbon dioxide ­(CO2) together with minor impurities such as hydrogen sulfide ­(H2S), siloxanes and aromatic, halogenated and other volatile organic compounds (VOCs) and * Ebrahim Tilahun [email protected] Erkan Sahinkaya [email protected] Bariş Çalli [email protected] 1



Department of Environmental Engineering, Marmara University, 34722 Istanbul, Turkey



Department of Bioengineering, Istanbul Medeniyet University, 34700 Istanbul, Turkey

2

ammonia [2]. Biogas can be used directly to generate power and heat; however, the high ­CO2 content reduces its heating value and increases the compression and transportation costs restricting its use at the point of production [3]. The ­H2S is known to cause critical problems such as; toxicity, corrosion in pipes, turbines or other units, as well as adverse environmental pollution (sulfur dioxide emissions) [4, 5]. In order to obtain energy from biogas in a more productive and cost-efficient way, the biogas has to be enriched and the impurities should be removed. There is an increasing demand for technologies capable of separating ­H2S and ­CO2 from the biogas and enable the delivery of a purified gas product suitable for use in different applications [6, 7]. Conventional gas–liquid absorption processes such as bubble column, packed and spray towers are dependent on direct contact between the gas and liquid phases, and cheap process when water is available. However, these processes require large space, high pressure and energy consumption, low temperature, high investment cost [8, 9]. The membrane gas–liquid contacting process, which integrates the advantage of conventional absorption devises, can potentially overcome those operational limitations and

13

Vol.:(0123456789)



is considered as a better alternative for the treatment of both ­H2S and C ­ O2 [10]. Membranes used in a gas–liquid contacting process act as barrier between the gas and the liquid phases without dispersing one phase into another. In the membrane contactor, gas diffuses from the gas side across the membrane and reaches into the gas–liquid interface, where gas is absorbed in the liquid. Compared with conventional gas absorption devices, the use of membrane contactor offers a number of advantages such as high surface area per unit contact volume, ease of adjusting the flowrate of liquid and gas phases independently, high compactness and easy to scale up or down [10–12]. Because of the advantages offered by membrane contactors, numerous studies have been carried out. Qi and Cussler [13, 14] first used microporous hollow fiber membranes for gas absorption. Following their innovative findings, many others have focused on gas–liquid membrane contactors for removal of ­H2S, ­CO2 and other gases from gas streams [15, 16]. The prolonged contact of microporous membrane with the liquid phase can generate detrimental effects on the membrane material due to swelling or gradual wetting of the hydrophobic membrane pores, which then may decrease the absorption flux significantly [17–19]. Similarly, other researchers also attributed the performance deterioration of the membrane contactors due to wetting problem. Wetting significantly affects the mass transfer coefficients of the membrane module, thus membrane resistance increases sharply and operation performance declines soon [20, 21]. The possibility to ensure nonwetting conditions for long operational period is of drastic importance. Hence, to overcome the wetting problem and avoid any penetration of the liquid into the membrane, a material which represents a barrier for penetration of liquid into the membrane is suggested. One possible solution is to use a nonporous or dense polymeric membrane, which is highly permeable to H ­ 2S and absolutely impermeable to the liquid permeation. Besides, the nonporous membrane can enhance selectivity for ­H2S over ­CO2 and ­CH4 since separation is mediated by diffusion through the polymeric membrane material [22]. Nevertheless, it was concluded in a limited number of studies that the nonporous membrane layer could not provide sufficient mass transfer compared to microporous membranes [23, 24]. Therefore, a high membrane permeability is necessary to increase the overall mass transfer capacity of the nonporous membrane contactors. As a result, reactive liquid absorbent such as quicklime, slaked lime, sodium hydroxide, or amine solutions are preferably used to increase the driving force for better absorption rate and capacity. The high concentrations of C ­ O2 in the biogas make the ­H2S removal difficult because the ­CO2 consumes the alkalinity in the liquid phase. Due to its low cost and rapid reaction rate with ­H2S compared to other gases, a sodium hydroxide solution was ideally used as an absorbent liquid. In addition, hydrophobic nonporous membranes like

13

Waste and Biomass Valorization

polydimethylsiloxane (PDMS) is suitable for use in membrane contactors due to its low cost and availability. So far, many studies investigated the effects of different liquid absorbents and membrane materials on the performance of ­H2S and ­CO2 absorption. However, studies evaluating the effects of different operating parameters on the selective ­H2S removal performance of membrane contactor was very limited. In this study, the effects of gas retention time (GRT), membrane thickness and pH of the liquid absorbent on the membrane contactor based biogas desulfurization process were investigated experimentally.

Experimental Materials A gas cylinder containing H ­ 2S (1%) and C ­ O2 (39%) in balance of C ­ H4 (60%) was purchased from Hat Industrial Gases PLC, Kocaeli, Turkey. Sodium hydroxide was purchased from a local supplier (TEKKİM Kimya Sanayi, Bursa, Turkey) and diluted as required with distilled water to prepare the stock alkaline solution. All chemicals were of the highest available purities and used without further purification. A commercial tubular PDMS (polydimethylsiloxane) membrane having 1 mm thickness and 7 mm internal diameter was purchased from EUROFLEX (Germany), while the PDMS having thickness of 1.5 and 2 mm with internal diameter of 7 mm were purchased from DEUTSCH & NEUMANN (Germany).

Membrane Contactor Setup A laboratory scale gas–liquid membrane contactor was designed and manufactured to perform biogas desulfurization experiments under different operational conditions. The schematic diagram and the details of the desulfurization setup was shown in our previous paper [25]. The internal diameter and height of the glass reactor were 120 and 200 mm, respectively. The reactor active working volume and the tubular membrane length immersed in the liquid were 1.5 L and 3 m, respectively. The reactor was totally filled with tap water to minimize the volatilization of the sulfur compounds. To reduce the oxygen concentration inside the reactor the liquid absorbent was flushed with nitrogen gas. Due to the high toxicity of H ­ 2S, all experiments were performed in a fume hood to confine any accidental leakage of ­H2S gas. The gas flow rate was adjusted by a stainless steel mass flow controller on the line between the gas cylinder and the set-up. Digital gas counter was used to measure the inlet and outlet biogas volumetric flow rate (MGC, Ritter, Germany). A pH range of 7–10 was maintained by automated addition of 1N NaOH into the liquid absorbent as needed. Almost all of H ­ 2S and some of C ­ O2 were removed

Waste and Biomass Valorization

from the simulated gas mixture by diffusing through the membrane and then absorbing/reacting with the mildly alkaline solution. For uniform distribution of gas and alkaline solution in the liquid phase magnetic stirrer was used throughout the experiment. The experiments were carried out at ambient pressure (approximately 1 atm), and the liquid temperature was kept at 20 ± 1 °C using a heating blanket equipped with electric coils wrapped on the wall of the reactor. The conductivity, oxidation reduction potential (ORP) and dissolved oxygen (DO) were monitored with electrodes on-line using a digital multimeter (Multi 9430, WTW, Germany).

Techniques and Test Methods Scanning Electron Microscopy with Energy Dispersive X-ray Spectroscopy (SEM-EDS) analyses were carried out at Yildiz Technical University Science and Technology Application and Research Center (Istanbul, Turkey). The changing morphologies of the membrane were directly observed using a SEM after Au–Pd coating. The semi quantitative elemental analyses of the inorganics deposited on the external surface of the membrane were performed using an EDS coupled with SEM at the end of the experiment. The concentration of sulfide in the liquid phase was determined spectrometrically (WTW PhotoLab 6100VIS) following the method described by Cord-Ruwisch [26]. Sulfate (4500-SO42− E) concentration was measured according to standard methods [27]. The contents of H ­ 2S, ­CO2 and ­CH4 gas in the entrance and exit of the membrane contactor were analyzed by gas chromatography (GC-2014; Shimadzu, Tokyo, Japan) equipped with a thermal conductivity detector and a stainless steel column. Argon gas was used as the carrier gas. The temperatures of injection port, column and detector were 150, 40 and 150 °C, respectively. When the ­H2S concentration of the biogas in the exit was below the detection limit of the GC-TCD, a second scrubbing liquid was used at the end of the gas line to absorb all traces of ­H2S before venting, the detail of the measurement on the second scrubber was described in a former study [28]. All samples were collected and analyzed in triplicate following each change in the process operation conditions and the averages were reported.

Calculations In this study, H ­ 2S, ­CO2 or C ­ H4 removal efficiencies (RE) from the gas phase were calculated according to Eq. 1. ( in ) ( ) Qg × Cg in − Qg out × Cg out RE (%) = × 100 ( in ) (1) Qg × Cg in where ­Qgin and ­Qgout are inlet and outlet biogas flowrates ­(m3/days), respectively. ­Cgin and ­Cgout are the inlet and outlet concentrations in gas phase (mg/l). In the case of

simultaneous removal of ­H2S and ­CO2, selectivity of the desulfurization could be represented by selectivity factor or separation factor (S), which could be expressed by Eq. 2. ( ) Xi ∕Xj S= ( ) (2) Yi ∕Yj where ­Xi,j are the mole fraction of components i and j in the liquid phase and Y ­ i,j are the mole fraction of components i and j in the feed gas. Surface removal rate or flux (J) of the membrane contactor is another performance indicator, which can be estimated as Eq. 3. ( in ) ( ) Qg × Cg in − Qg out × Cg out (3) J= A where J is the flux of the gas components (g/m2/days) and A is the membrane surface area ­(m2).

Results and Discussions Impact of GRT and Membrane Thickness on Desulfurization Performance The gas retention time (GRT) is one of the crucial parameters determining the biogas desulfurization efficiency. The influence of GRT on the removal efficiency (RE) of each gas component is presented in Table 1. The pH of the liquid absorbent was constant (pH 10) in the experiments. The result shows that, the decrease in GRT reduces the membrane surface contact per unit volume of gas, which in turn reduces RE. The ­H2S RE of the process was superior, especially when the GRT laid in the range of 19–10.4 min. The ­H2S was completely removed at that GRT ranges. Moreover, at GRT of over 10 min, an effluent ­H2S concentration below 300 ppmv was achieved, which is safe to use in cogeneration units [29]. However, lower GRT resulted in higher effluent ­H2S concentrations which needs further treatment before using the gas for various applications. In the same way, ­CO2 removal declined significantly with the decreased GRT. At all GRT tested, the RE of H ­ 2S was much higher than that of ­CO2. Because ­H2S has relatively high critical temperature or gas condensability and expected to permeate faster through the dense membrane than ­CO2. Baker [30] stated that the permeation of gas components through rubbery polymers depends mainly on the gas condensability. Over all, the RE of H ­ 2S and C ­ O2 improved by more than 2.5 and 5.2 times, when the GRT was raised from 3.4 to 19 min, respectively (Table 1). This finding is in agreement with the results presented by Wang et al. [21]. They reported that gas removal capacity decreased as the GRT declined because of the shorter contact between the gas and liquid

13



Waste and Biomass Valorization

Table 1  The membrane contactor performance at different GRT and membrane thicknesses at liquid pH of 10

Thickness (mm)

Gas retention time (GRT) (min)

1

19 10.4 5.6 3.4 19 10.4 5.6 3.4 19 10.4 5.6 3.4

1.5

2

100

H2S

CO2

CH4

Gas effluent (%)

80

60

40

20

0

3

6

9

12

15

18

21

GRT (min)

Fig. 1  Impacts of GRT on the effluent gas contents at pH 10. (Color figure online)

phases. Nevertheless, C ­ H4, having a low diffusivity through the PDMS membrane and low solubility in the liquid absorbent, could not permeate across the membrane. Figure 1 also illustrated that, ­CH4 content in the effluent stream increased along with the GRT and enriched from 60% to a maximum of 87% with only 4.68% loss. Heile et al. [31] performed a similar study for upgrading of a biogas containing 20% of ­CO2 and 80% of ­CH4 using a PDMS membrane and they improved the ­CH4 content up to 88% in the outlet. In addition, experiments with varying membrane thicknesses of 1, 1.5 and 2 mm were also performed. According to the results shown in Table 1, thicker membrane reduces the ­H2S and ­CO2 transfer rates across the membrane due to longer diffusion time. The negative impact of higher membrane thickness in overall mass transfer of other permeates in silicone membranes also reported by Raghunath and Hwang [32]. They found significant mass transfer resistance when the thickness of a silicone membrane was above 1.16 mm.

13

Removal efficiency (RE) (%)

Selectivity (S)

H2S

CO2

CH4

H2S/CH4

H2S/CO2

100 98.3 63.6 40.0 99.9 97.8 60.6 38.2 99.7 97.0 56.9 36.2

79.3 56.4 34.9 15.3 68.7 43.0 22.8 12.1 60.5 35.8 18.0 8.92

4.68 3.44 1.84 0.69 4.12 2.86 1.56 0.63 3.58 2.4 1.23 0.58

21.4 28.6 34.6 59 24.2 34.2 38.8 60.6 27.8 40.2 46.3 62.4

1.26 1.74 1.82 2.6 1.45 2.27 2.66 3.06 1.65 2.72 3.16 4.06

In a similar manner, Brookes and Livingston [33] used silicone based membranes and they reported a reduction by a factor of 1.5 in the overall mass transfer coefficient for phenol when the membrane thickness was increased from 0.3 to 0.5 mm. Therefore, with thinner membranes, higher loading rates could be attained. As a general trend, when the membrane thickness increases, the ­CO2 removal efficiency decreases more significantly than that of H ­ 2S even at high GRT. The reduction was much more pronounced for lower GRT. Nii and Takeuchi [34] also used PDMS hollow fıber modules for the removal of ­CO2 and achieved a higher performance with thinner membranes. It can be concluded that thicker membrane had introduced a considerable resistance to the gas diffusion, which decreased the absorption flux inevitably. In the previous studies, separation factor has been extensively used to indicate gas separation efficiency of membrane based processes [35, 36]. In the gas–liquid membrane contacting process used here ­H2S in the feed gas diffused through the membrane and was absorbed in the mildly alkaline absorbent. For that reason, the permeate selectivity was used to describe the process efficiency. As shown in Eq. 2, the permeate selectivity depends on the difference between the inlet and outlet gas concentrations which are influenced by the membrane thickness, GRT, and pH of the absorbent used. In order to examine the effect of different membrane thicknesses and GRT on the selectivity of the membrane ­ O2 and contactor, the overall mass transfer ratio of H ­ 2S to C ­CH4 were calculated. During each experimental tests, pH of the liquid was adjusted to 10. As presented in Table 1, a higher selectivity for the separation of ­H2S from the gas mixture was observed at lower GRT. A change in membrane thickness from 1 to 2 mm had a positive influence on the selectivity due to the higher permeability of H ­ 2S in thicker membrane compared to those of other gases. However, in that case a lower RE was observed as a result of additional

Waste and Biomass Valorization

resistance of the membrane. Consequently, for a gas liquid membrane contactor process used here, these two parameters can considerably influence the H ­ 2S selectivity and RE. Specific to the experimental results of this work, the maximum separation factor for ­H2S/CO2 and H ­ 2S/CH4 were reached up to 4 and 62, respectively, in the case of using thicker membrane (2 mm). Stern and Bhide [37] pointed out that H ­ 2S is more permeable than C ­ O2 through PDMS by approximately a factor of 1.8. Chatterjee et al. [38] also used cellulose acetate membrane to clean the biogas having 6% of ­H2S, 29% of ­CO2 and 65% of C ­ H4 at 10 bar, and they reported the ­H2S/CH4 separation factor as only 19. In another work on separation of gases using Poly (ether urethane) membrane, separation factor of H ­ 2S/CO2 and H ­ 2S/CH4 were reported as 3 and 21, respectively [38].

Impact of Liquid Absorbent pH on Desulfurization Performance The gas-liquid-membrane contactor was also tested with liquid absorbent having different pH values for the removal of ­H2S from the biogas stream. As displayed in Table 2, the reduction of GRT negatively affected the RE of the gas components. However, when the pH of the liquid absorbent increased, the permeation of ­H2S across the membrane improved due to the increased concentration differences between the two sides of the membrane, i.e. the driving force. According to the report of earlier studies charged ions existing in the liquid absorbent are unable to permeate through the silicone rubber membrane [22, 39, 40]. Indeed, when the pH decreased, the amount of undissociated H ­ 2S increased, resulting in an increase of the H ­ 2S back diffusion [25]. Hence, the pH is a key parameter in controlling H ­ 2S mass transfer through the membrane. The mildly alkaline Table 2  Impact of different liquid pH on gas RE at 1 mm membrane thickness pH

10

8.5

7.0

Gas retention time GRT (min)

Removal efficiency (RE) (%) H2S

CO2

CH4

19.0 10.4 5.6 3.4 19.0 10.4 5.6 3.4 19.0 10.4 5.6 3.4

100 98.3 63.6 40.0 99.9 91.5 52.8 32.0 97.9 82.0 41.8 20.0

79.3 56.4 32.9 15.3 71.1 51.6 28.9 14.8 69.9 45.8 27.1 12.0

4.68 3.44 1.84 0.69 4.44 3.25 1.55 0.59 3.96 3.01 1.15 0.50

solutions give the greater fractional removal of H ­ 2S compared to ­CO2 and ­CH4. This is owing to the instantaneous reaction rates ­OH− with ­H2S rather than the other gases [13, 14]. At the highest GRT (19 min) changing of pH from 7 to 10 were not significantly affect H ­ 2S RE (Table 2). To establish the limiting process conditions, the transition in GRT from 19 to 3.4 min was examined, and upon decreasing of the GRT, ­H2S RE dropped sharply. In general, at higher liquid pH (> 8.5) a better RE was achieved. Because at high pH values almost all of the dissolved ­H2S was dissociated to hydrosulfide ion, which increased the RE by continually re-establishing the higher ­H2S concentration gradient. Smet et al. [41] also observed a sharp increase in hydrosulfide ion concentration when the pH was higher than 7.04. Similarly, González-Sánchez et al. [42] reported that a slightly alkaline pH could improve the H ­ 2S mass transfer from the gas phase to the liquid phase. According to the result shown in Table 2, the absorption of ­CO2 into the liquid absorbent was considerably higher than that of ­CH4, as it chemically reacted with ­OH−. The absorption of both ­H2S and ­CO2 in alkaline liquid was assisted by agitation. The turbulence in the liquid increases the diffusion of the molecules into the liquid phase due to reduction of boundary layer resistance between membrane and liquid interface. Under the tested conditions, the highest RE of ­CO2 was achieved at GRT of 19 min as 69.9, 71.1 and 79.3% at liquid pH of 7, 8.5 and 10, respectively. On the contrary, at the lowest GRT (3.4 min), the RE of ­CO2 at liquid pH of 7 and 8.5 were 12 and 14.8%, respectively, while as the liquid pH raised to 10 its RE increased to 15.31%, which is due to the high concentration of O ­ H− ion at higher liquid pH values. Likewise, the positive effect of absorbent concentration on the mass transfer rate were formerly reported [43, 44]. In general, the pH of the liquid in this study was not increased above 10, to avoid much carbonate formation ­(CO2 absorption) and ­CH4 loss. The conductivity of a solution depends on the concentration of all the ions present, the greater their concentrations, the higher the conductivity. During operation of the system used here, the major cations are, ­H+, ­Na+, ­Ca2+, and ­Mg2+. The major anions are, ­OH−, ­HS−, ­S2−, ­HCO3− and ­CO32−. Hence, acidic or basic solution resulted with high conductivity. Moreover, the conductivity is the sum of the contribution of all ions present in the solution. For soft water samples, a pH of 7 will have the least conductivity. On other hand, samples with a pH above 7 are likely to be higher conductivity. Figure 2 also confirmed that when the pH of the liquid absorbent increased from 7 to 10, proportionally the conductivity raised up. This happened due to higher dissociation of ­H2S and ­CO2 into sulfide and carbonate. Moreover, to kept constant pH sodium hydroxide has been supplied to the liquid phase.

13



Waste and Biomass Valorization

Fig. 2  ORP and conductivity observation at different pH and GRT. (Color figure online)

ORP (mV)

-100

-200

-300

-400 2000

pH 7

pH 8.5

pH 10

pH 7

pH 8.5

pH 10

Conductivity (µS/cm)

1500

1000

500

0 5

10

15

20

Biogas retention time (min) Our results were consistent with the finding of leveling [45]. Figure 2 demonstrated that ORP decreased as the pH of the liquid absorbent increased, which was due to higher accumulation of sulfide. ORP values was around − 200 mV at low sulfide accumulation, whereas at higher loadings significant sulfide accumulation caused ORP to drop below − 380 mV. Besides, at pH 10, in particular with a GRT of 3.4 min, the ORP value decreased sharply as an indication of sulfide accumulation. This behavior proves a correlation between the ORP and H ­ 2S absorption capacity of the slightly alkaline liquid absorbent.

13

Membrane Morphology and Inorganics Deposition The surface morphology of PDMS membrane was examined using a scanning electron microscope (SEM). SEM images of virgin membrane and dirty membrane are presented in Fig. 3a and b, respectively. The SEM images confirms there is no visible variation between the surface morphologies of the two membranes samples. However, when the surface of the membrane was magnified, the layer of deposits on the membrane surface could be clearly realized. As illustrated in Fig. 3b, it is observed that after each experimental work a white crystal substances appeared on the surface of used

Waste and Biomass Valorization Fig. 3  SEM image of the membrane surface a before (virgin) and b after the experiments (used)

membrane, which should be inorganic matters as confirmed by EDS analysis, whereas the virgin membrane surface viewed clean and smooth (Fig. 3a). The SEM images also attributed to the fact that the pattern of inorganic matters is an unevenly distributed over the membrane surface. Despite minor deposition of inorganics on the membrane surface, the structure was not suffered after being exposed to sodium

hydroxide aqueous solutions. In addition, the membrane used here was less sensitive to wetting since there is no pores that supports the liquid to penetrate through the membrane. Thus, membrane wetting was not observed in our PDMS membrane. However, other researchers attributed performance deterioration of the microporous membrane contactors due to wetting and blockage of the membrane pores.

13



They reported significant drop in mass transfer capacity of the contactors due to the developed membrane resistance [20, 46]. Keshavarz et al. [19] used a microporous hollow fiber membrane contactor to investigate the simultaneous absorption of C ­ O2 and ­H2S into the aqueous solution of diethanolamine and they found that the RE of both gases significantly decreased due to membrane wetting compared with the non-wetted mode. Wang et al. [21] also studied on ­CO2 absorption using a diethanolamine solution as an

Waste and Biomass Valorization

absorbent in a polypropylene microporous membrane contactor. They reported that with only 5% membrane pores wetting the overall mass transfer coefficient of the contactor may reduce by 20%. With the use of scanning electron microscopy (SEM) associated with an energy dispersive X-ray spectrometer (EDS) often focus on the top surface deposits as indicated in Fig. 4a and b, and it is possible to detect the existence of metal sulfide and carbonate salts deposition. The

Fig. 4  The semi quantitative elemental analyses of the inorganics deposited on the membrane surface a SEM image b EDS analysis after the experiment

13

Waste and Biomass Valorization

semi-quantitative EDS composition analysis indicated certain amounts of inorganic elements were accumulated on the membrane surface. The presence of multivalent metal ions in the tap water, which used in the system as an absorbent liquid, was the origin of inorganics deposition. Thus, in the presence of metal ions such as C ­ a2+ and M ­ g2+ that act as cationic effects, there would be a strong interaction with anionic sulfide, and carbonate, which were generated during the dissociation of ­H2S and ­CO2, to form precipitates. Furthermore Fig. 2 also approves that, when the pH of the liquid phase increased the conductivity of the liquid absorbent linearly increased proportionally. The reason for this observation was the greater accumulation of sulfide and carbonate in the reactor as the pH of the liquid raised up (when the pH increased to 10). It can be observed that carbon (C) and oxygen (O) were the major elements present in the spot samples. The C and O peaks were likely due to the chemical structure of the membrane and entrapment of inorganic matters. As discussed above, the multivalent ions have been shown to form precipitates with sulfide and carbonates could contribute to the calcium (Ca), magnesium (Mg) peaks and their weight percent were about 36.1 and 0.5%, respectively. The sulfur (S) peaks possibly indicating that S containing compounds were able to penetrate through the membrane with some retained on the membrane surface. Silicon (Si) detected on the used membrane surface was mainly due to its presence in the polydimethylsiloxane membrane layer. The EDS spectra of the virgin membrane has also strong Si and O peaks which originated from the membrane itself (data not shown). Gold (Au) and palladium (Pd) were used during coating procedures, due to its irrelevant to the elemental analysis of the membrane foulants their peaks were removed. In the present study regardless of inorganics deposition discussed above, the abiotic experiments show selective removal ­H2S than ­CO2 and ­CH4. Besides, in this gas–liquid membrane contactor applications, membrane fouling and clogging were not observed, which resulted in almost stable flux during the operation of the membrane. Previous studies also demonstrated nonporous membrane was more resistant to fouling [47]. However, in the long run operation, an excess deposition may have reduced the mass transfer efficiency of the PDMS membrane. It was assumed that the accumulation of inorganics on the membrane surface may decreased the cross sectional area, subsequently reduced the gas retention time and increased the pressure drop of the gas permeation. Moreover, it may create an additional film resistance for the gas to reach the liquid phase, which likely resulted in limitation of ­H2S transfer. Chuichulcherm et al. [40] used silicone membrane combined with biological sulfide production for the treatment of metal-containing wastewater. They reported that the chemical reaction between sulfide and multivalent ions in the wastewater enhanced the sulfide mass transfer. However, accumulation of metal precipitates

on the membrane surface limited sulfide transfer and the resistance due to the metal sulfide precipitates even exceeded the membrane resistance. Other studies also demonstrated the mass transfer reduction of membrane based reactors due to the blockage of the membrane pores by the organic and inorganics accumulation [46, 48, 49].

Conclusions Results of the study performed to evaluate the effects of operating parameters on H ­ 2S removal performance of a gas–liquid PDMS membrane contactor showed that the ­H2S removal declined severely, when GRT decreased below 10 min, due to mass transfer limitation. In addition, increasing membrane thickness reduced the rate of ­H2S diffusion through the membrane. By using a slightly alkaline absorbent, more ­H2S was removed with respect to C ­ O2 and C ­ H 4. On other hand, lower GRT and higher membrane thickness significantly increased the mass transfer resistance against ­CO2, but showed a marginal influence on H ­ 2S removal; hence it favors a higher selectivity for ­H2S. The maximum selectivity for H ­ 2S/CO2 and H ­ 2S/CH4 were 4 and 62, respectively. Despite some loss, the ­CH4 content of the biogas enriched up to 87%. Results of SEM-EDS analysis showed that at higher pH values, sulfide and carbonate salts of Ca, Mg, and Si deposit were observed on the membrane surface. In spite of inorganics deposition, no membrane clogging and fouling problems were observed. However, it is supposed that in long run, the chemical precipitates on the outer surface of membrane may behave as a secondary barrier and reduce the diffusion of gasses. Finally, it is concluded that the gas–liquid PDMS membrane contactor process tested in this study is a promising alternative to conventional biogas desulfurization processes. Acknowledgements  This study was financially supported by YTB (Presidency for Turks Abroad and Related Communities) and Marmara University Scientific Research Committee BAPKO (Project No. FEN-C-DRP-070317-0109).

References 1. Holm-Nielsen, J.B., Al Seadi, T., Oleskowicz-Popiel, P.: The future of anaerobic digestion and biogas utilization. Bioresour. Technol. 100, 5478–5484 (2009). https​://doi.org/10.1016/j.biort​ ech.2008.12.046 2. Zhang, R., Brown, R.C., Suby, A., Cummer, K.: Catalytic destruction of tar in biomass derived producer gas. Energy Convers. Manag. 45, 995–1014 (2004). https​://doi.org/10.1016/j.encon​ man.2003.08.016 3. Marzouk, S.A.M., Al-Marzouqi, M.H., Teramoto, M., Abdullatif, N., Ismail, Z.M.: Simultaneous removal of ­CO2 and H2S from pressurized ­CO2–H2S–CH4 gas mixture using hollow fiber

13



4.

5. 6.

7.

8. 9.

10. 11. 12.

13. 14.

15. 16. 17.

18.

19.

20.

Waste and Biomass Valorization membrane contactors. Sep. Purif. Technol. 86, 88–97 (2012). https​://doi.org/10.1016/j.seppu​r.2011.10.024 Poloncarzova, M., Vejrazka, J., Vesely, V., Izak, P.: Effective purification of biogas by a condensing-liquid membrane. Angew. Chem. Int. Ed. 50, 669–671 (2011). https ​ : //doi. org/10.1002/anie.20100​4821 Panza, D., Belgiorno, V.: Hydrogen sulphide removal from landfill gas. Process Saf. Environ. Prot. 88, 420–424 (2010). https​:// doi.org/10.1016/j.psep.2010.07.003 Asri, O., Hafidi, I.E., Afilal, M.E.: Comparison of biogas purification by different substrates and construction of a biogas purification system. Waste Biomass Valoriz. 6, 459–464 (2015). https​ ://doi.org/10.1007/s1264​9-015-9378-z Kurchania, A.K., Panwar, N.L., Pagar Savita, D.: Improved biogas stove with scrubbing unit for household use. Waste Biomass Valoriz. 2, 397–402 (2011). https​://doi.org/10.1007/s1264​ 9-011-9080-8 Awe, O.W., Zhao, Y., Nzihou, A., Minh, D.P., Lyczko, N.: A review of biogas utilisation, purification and upgrading technologies. Waste Biomass Valoriz. 8, 267 (2017) Ryckebosch, E., Drouillon, M., Vervaeren, H.: Techniques for transformation of biogas to biomethane. Biomass Bioenerg. 35, 1633–1645 (2011). https​: //doi.org/10.1016/j.biomb​ ioe.2011.02.033 Gabelman, A., Hwang, S.-T.: Hollow fiber membrane contactors. J. Memb. Sci. 159, 61–106 (1999). https​://doi.org/10.1016/ S0376​-7388(99)00040​-X Klaassen, R., Feron, P.H.M., Jansen, A.E.: Membrane contactors in industrial applications. Chem. Eng. Res. Des. 83, 234–246 (2005). https​://doi.org/10.1205/cherd​.04196​ Belaissaoui, B., Claveria-Baro, J., Lorenzo-Hernando, A., Albarracin Zaidiza, D., Chabanon, E., Castel, C., Rode, S., Roizard, D., Favre, E.: Potentialities of a dense skin hollow fiber membrane contactor for biogas purification by pressurized water absorption. J. Memb. Sci. 513, 236–249 (2016). https​:// doi.org/10.1016/j.memsc​i.2016.04.037 Qi, Z., Cussler, E.L.: Microporous hollow fibers for gas absorption: I—mass transfer in the liquid. J. Memb. Sci. 23, 321–332 (1985). https​://doi.org/10.1016/S0376​-7388(00)83149​-X Qi, Z., Cussler, E.L.: Microporous hollow fibers for gas absorption: II—mass transfer across the membrane. J. Memb. Sci. 23, 333–345 (1985). https​://doi.org/10.1016/S0376​-7388(00)83150​ -6 Karoor, S., Sirkar, K.K.: Gas absorption studies in microporous hollow fiber membrane modules. Ind. Eng. Chem. Res. 32, 674– 684 (1993). https​://doi.org/10.1021/ie000​16a01​4 Poddar, T.K., Majumdar, S., Sirkar, K.K.: Membrane-based absorption of VOCs from a gas stream. AIChE J. 42, 3267–3282 (1996) Mavroudi, M., Kaldis, S.P., Sakellaropoulos, G.P.: A study of mass transfer resistance in membrane gas-liquid contacting processes. J. Memb. Sci. 272, 103–115 (2006). https:​ //doi.org/10.1016/j.memsc​ i.2005.07.025 Keshavarz, P., Fathikalajahi, J., Ayatollahi, S.: Analysis of C ­ O2 separation and simulation of a partially wetted hollow fiber membrane contactor. J. Hazard. Mater. 152, 1237–1247 (2008). https​ ://doi.org/10.1016/j.jhazm​at.2007.07.115 Keshavarz, P., Fathikalajahi, J., Ayatollahi, S.: Mathematical modeling of the simultaneous absorption of carbon dioxide and hydrogen sulfide in a hollow fiber membrane contactor. Sep. Purif. Technol. 63, 145–155 (2008). https​://doi.org/10.1016/j.seppu​ r.2008.04.008 Atchariyawut, S., Jiraratananon, R., Wang, R.: Separation of C ­ O2 from ­CH4 by using gas-liquid membrane contacting process. J. Memb. Sci. 304, 163–172 (2007). https:​ //doi.org/10.1016/j.memsc​ i.2007.07.030

13

21. Wang, R., Zhang, H.Y., Feron, P.H.M., Liang, D.T.: Influence of membrane wetting on C ­ O2 capture in microporous hollow fiber membrane contactors. Sep. Purif. Technol. 46, 33–40 (2005). https​://doi.org/10.1016/j.seppu​r.2005.04.007 22. Tilahun, E., Sahinkaya, E., Çalli, B.: A hybrid membrane gas absorption and bio-oxidation process for the removal of hydrogen sulfide from biogas. Int. Biodeterior. Biodegrad. 127, 69–76 (2018). https​://doi.org/10.1016/j.ibiod​.2017.11.015 23. Al-Marzouqi, M.H., Marzouk, S.A.M., El-Naas, M.H., Abdullatif, N.: ­CO2 removal from ­CO2–CH4 gas mixture using different solvents and hollow fiber membranes. Ind. Eng. Chem. Res. 48, 3600–3605 (2009). https​://doi.org/10.1021/ie800​977z 24. Al-saffar, H.B., Ozturk, B., Hughes, R.: A comparison of porous and non-porous gas-liquid membrane contactors for gas separation. Chem. Eng. Res. Des. 75, 685–692 (1997). https​://doi. org/10.1205/02638​76975​24182​ 25. Tilahun, E., Bayrakdar, A., Sahinkaya, E., Çalli, B.: Performance of polydimethylsiloxane membrane contactor process for selective hydrogen sulfide removal from biogas. Waste Manag. 1–8 (2017). https​://doi.org/10.1016/j.wasma​n.2017.01.011 26. Cord-Ruwisch, R.: A quick method for the determination of dissolved and precipitated sulfides in cultures of sulfate-reducing bacteria. J. Microbiol. Methods 4, 33–36 (1985). https​://doi. org/10.1016/0167-7012(85)90005​-3 27. AWWA A.: WEF, standard methods for the examination of water and wastewater. (2005) 28. Bayrakdar, A., Tilahun, E., Calli, B.: Biogas desulfurization using autotrophic denitrification process. Appl. Microbiol. Biotechnol. 100, 939–948 (2016). https​://doi.org/10.1007/s0025​ 3-015-7017-z 29. Ramos, I., Fdz-Polanco, M.: Microaerobic control of biogas sulphide content during sewage sludge digestion by using biogas production and hydrogen sulphide concentration. Chem. Eng. J. 250, 303–311 (2014). https​://doi.org/10.1016/j.cej.2014.04.027 30. Baker, R.W.: Membrane Technology and Application. Wiley, New York (2004) 31. Heile, S., Rosenberger, S., Parker, A., Jefferson, B., McAdam, E.J.: Establishing the suitability of symmetric ultrathin wall polydimethylsiloxane hollow-fibre membrane contactors for enhanced CO2 separation during biogas upgrading. J. Memb. Sci. 452, 37–45 (2014). https​://doi.org/10.1016/j.memsc​i.2013.10.007 32. Raghunath, B., Hwang, S.T.: General treatment of liquid-phase boundary layer resistance in the pervaporation of dilute aqueous organics through tubular membranes. J. Memb. Sci. 75, 29–46 (1992). https​://doi.org/10.1016/0376-7388(92)80004​-4 33. Brookes, P.R., Livingston, A.G.: Aqueous-aqueous extraction of organic pollutants through tubular silicone rubber membranes. J. Memb. Sci. 104, 119–137 (1995). https​://doi.org/10.1016/03767388(95)00020​-D 34. Nii, S., Takeuchi, H., Takahashi, K.: Removal of ­CO2 by gas absorption across a polymeric membrane. J. Chem. Eng. Jpn. 25, 67–72 (1992) 35. Stern, S.A.: Polymers for gas separations: the next decade. J. Memb. Sci. 94, 1–65 (1994) 36. Mudler, M.: Basic Principles of Membrane Technology. Springer, New York (1996) 37. Stern, A., Bhide, B.D.: Permeability of silicone polymers to ammonia and hydrogen sulfide. J. Appl. Polym. Sci. 38, 2131– 2147 (1989) 38. Chatterjee, G., Houde, A.A., Stern, S.A.: Poly(ether urethane) and poly(ether urethane urea) membranes with high H ­ 2S/CH4 selectivity. J. Memb. Sci. 135, 99–106 (1997) 39. Livingston, A.G.: Extractive membrane bioreactors: a new process technology for detoxifying chemical industry wastewaters. J. Chem. Technol. Biotechnol. 60, 117–124 (1994). https​://doi. org/10.1002/jctb.28060​0202

Waste and Biomass Valorization 40. Chuichulcherm, S., Nagpal, S., Peeva, L., Livingston, A.: Treatment of metal-containing wastewaters with a novel extractive membrane reactor using sulfate-reducing bacteria. J. Chem. Technol. Biotechnol. 76, 61–68 (2001) 41. Smet, E., Lens, P., Van Langenhove, H.: Treatment of waste gases contaminated with odorous sulfur compounds. Crit. Rev. Environ. Sci. Technol. 28, 89–117 (1998) 42. González-Sánchez A., Revah S., Deshusses M.A.: Alkaline biofiltration of ­H2S odors. Environ. Sci. Technol. 42, (2008) 43. Wang, R., Li, D.F., Liang, D.T.: Modeling of C ­ O2 capture by three typical amine solutions in hollow fiber membrane contactors. Chem. Eng. Process. Process Intensif. 43, 849–856 (2004). https​://doi.org/10.1016/S0255​-2701(03)00105​-3 44. Yan, S.P., Fang, M.X., Zhang, W.F., Wang, S.Y., Xu, Z.K., Luo, Z.Y., Cen, K.F.: Experimental study on the separation of C ­ O2 from flue gas using hollow fiber membrane contactors without wetting. Fuel Process. Technol. 88, 501–511 (2007). https​://doi. org/10.1016/j.fupro​c.2006.12.007 45. Leveling, T.: The relationship between pH and conductivity in a lithium contaminated, de-ionized water system. fermilab-pbarnote-674, 1–11 (2002)

46. Attaway, H., Gooding, C.H., Schmidt, M.G.: Comparison of microporous and nonporous membrane bioreactor systems for the treatment of BTEX in vapor streams. J. Ind. Microbiol. Biotechnol. 28, 245–251 (2002). https​://doi.org/10.1038/sj/jim/70002​35 47. Côté, P., Bersillon, J.L., Huyard, A.: Bubble-free aeration using membranes: mass transfer analysis. J. Memb. Sci. 47, 91–106 (1989). https​://doi.org/10.1016/S0376​-7388(00)80862​-5 48. Yurtsever, A., Çinar, Ö, Sahinkaya, E.: Treatment of textile wastewater using sequential sulfate-reducing anaerobic and sulfide-oxidizing aerobic membrane bioreactors. J. Memb. Sci. 511, 228–237 (2016). https​://doi.org/10.1016/j.memsc​i.2016.03.044 49. Alvarez-Hornos, F.J., Volckaert, D., Heynderickx, P.M., Van Langenhove, H.: Performance of a composite membrane bioreactor for the removal of ethyl acetate from waste air. Bioresour. Technol. 102, 8893–8898 (2011). https​://doi.org/10.1016/j.biort​ ech.2011.06.064

13