Insights on the Mechanisms of H2S Retention at Low Concentration

Subscriber access provided by STEPHEN F AUSTIN STATE UNIV

Article 2

Insights on the mechanisms of HS retention at low concentration on impregnated carbons Randreanne L.C.B. Menezes, Karine Oliveira Moura, Sebastiao M.P. Lucena, Diana C. S. Azevedo, and Moises Bastos-Neto Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/acs.iecr.7b03402 • Publication Date (Web): 18 Jan 2018 Downloaded from http://pubs.acs.org on January 18, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Industrial & Engineering Chemistry Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Insights on the mechanisms of H2S retention at low concentration on impregnated carbons

Randreanne L. C. B. Menezes; Karine O. Moura; Sebastião M. P. de Lucena; Diana C. S. Azevedo; Moises Bastos-Neto*

Universidade Federal do Ceará - Departamento de Engenharia Química Campus do Pici, bl. 731, 60760-400, Fortaleza - CE, Brazil

CORRESPONDING AUTHOR e-mail: [email protected], telephone: +55 85 33669240, fax: +55 85 33669610

1 ACS Paragon Plus Environment

Industrial & Engineering Chemistry Research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 51

ABSTRACT Adsorption of H2S onto porous materials is as an attractive technology for fine biogas cleaning. Three activated carbon samples were studied as adsorbents for biogas desulfurization at low concentration (100 ppm), in order to better understand the underlying mechanisms and provide a basis for the development of new materials. One of the carbons is impregnated with NaOH, other with Fe2O3 and the third one is the parent material. Molecular simulation was performed to distinguish the retention mechanism. Textural characterization revealed high surface areas and the existence of ultramicropores with sizes below 4 Å in all samples. The possibility of discriminating the retention regimes emphasized the great influence of the chemisorption in these systems increasing up to 50 times the capacity of retention of H2S for the sodium-impregnated sample (from 0.3 to 15.64 mg g-1). Surprisingly, both physisorption and chemisorption could be unequivocally detected for the non-impregnated sample by evaluating breakthrough curves in different temperatures (up to 423 K). The evaluation of regeneration by heat indicated that the adsorbents can recover about 50% and 20% of their initial capacity for non-impregnated and impregnated samples, respectively.

KEYWORDS desulfurization; adsorption; fixed-bed; activated carbon; molecular simulation


Page 3 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. Introduction Although being an environmentally friendly alternative 1, biogas has a composition highly dependent on the organic substrate from which it is produced 2, with several contaminants such as sulfur-containing compounds, water, ammonia, siloxanes and halogenated organic compounds 3. Among the sulfur-containing compounds, hydrogen sulfide (H2S) is very harmful and the main cause of corrosion in a biogas facility. Hence, it must be removed prior to combustion 4. The removal of H2S (desulfurization) can combine physical, chemical and biological processes in such a way to address economics and application issues. This process is ordinarily carried out in two stages 5. In the first one, a rough cleaning process (bulk separation) is carried out to reduce H2S concentration in raw biogas one order of magnitude (from thousands to some hundreds or dozens ppm) 6-7. In the second stage, fine cleaning (purification) is intended to further reduce H2S concentration to very low values (< 100 ppm) in order to achieve application requirements. The fine desulfurization by adsorption onto porous materials has been pointed out as an attractive technology in many studies aiming at the evaluation of different sorbents with high potential for that purpose 8-16. An efficient sorbent material should present high capacity and selectivity for H2S, but regenerability is also desirable, so that the sorbent may be used in several adsorption/desorption cycles 11. Activated carbons (ACs) are porous materials that frequently match such features and therefore are extensively applied 16-18 for that purpose with or without undergoing any further surface modification after their synthesis. This modification process can be performed via chemical impregnation on the carbon matrix with such agents as NaOH, KOH, K2CO3, Na2CO3,



Page 4 of 51

H3PO4, ZnO, ZnCl2, KI and Fe2O3 17, 19-24. The intended outcome with impregnation is to promote chemisorption of H2S 17, 25-26 in order to enhance bed performance and thus its useful lifetime. When H2S comes in contact with the structure of a carbonaceous material, several phenomena occur simultaneously, which depend on the presence of water, H2S partial pressure in the feed and the solid porosity 17-18, 20. Due to the complex interplay of these interactions, the influence of each of these variables is difficult to assess. A reliable method to discriminate physisorption from chemisorption regimes could be very useful to evaluate the influence of the porosity in the removal of H2S. Studies in the literature are often ambiguous when discriminating if a certain volume of H2S was retained by physical adsorption or chemisorption. The knowledge of this distinction is particularly relevant in the case of activated carbons, because impregnation also produces changes in the distribution of pores, raising doubts whether the material’s performance improvement is due to the impregnated species or due to the new pore structure. It has been evident that, in general, most of the studies with activated carbons for H2S removal has not assessed with precision the regimes of physical and chemical adsorption. Such analysis requires a detailed characterization of the material’s microporous structure. Several reports in the literature adequately address the role of micropores above 7 Å 19, 27-29 but the effects of the presence of ultramicropores (below 4 Å) are still unclear and not properly taken into account. The reason is that the great majority of the textural features were based on N2 isotherms at 77 K. It is well known that nitrogen has diffusion limitations at this temperature and may not adequately characterize activated carbons containing pores below 7 Å. It is recommended that information on ultramicropores be obtained with CO2 adsorption isotherms at 273 K 30. To better understand the relationship between the H2S retention mechanism and the pore 4 ACS Paragon Plus Environment

Page 5 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


structure of the material, a careful pore size distribution calculation was performed in three activated carbon samples with pore sizes ranging from ultramicropores (below 4 Å) to microporore. Molecular simulation was used for the first time as a reliable technique to calculate physical adsorption of H2S in the carbon pores in order discriminate between physical adsorption and chemisorption. An adequate series of activated carbons consisting of two samples impregnated with chemical species (NaOH and Fe2O3) and the third one being the parent material, all of them presenting ultramicroporous, was selected for this study. The H2S retention capacity was assessed by examining experimental breakthrough curves obtained under dry condition and low concentration (100 ppm H2S) and correlating them to theoretical results from molecular simulations. Novel insights on the adsorption regime are presented in order to help understanding the interplay of porosity and surface chemistry. Sorbent regenerability is also evaluated as a criterion for the choice between modified and non-modified carbons.

2. Experimental 2.1 Materials Three commercial activated carbons produced by Donau Carbon GmbH (Germany), named as Desorex K43 (parent material), Desorex K43-Na (impregnated with NaOH), Desorex K43-Fe (impregnated with Fe2O3) were investigated in this work. These carbons were obtained from activation of a mixture of bituminous carbon and coconut shell and were supplied in the form of pellets 31-32. The samples were crushed and sieved down to 18 × 30 mesh to improve column packing. Hydrogen sulfide was supplied by White Martins Praxair diluted in helium at 100 ppm.



Page 6 of 51

2.2 Adsorbent characterization The chemical composition (in metal oxides) of the samples was determined by X-ray ﬂuorescence analysis (XRF) using an ARL ADVANT’XP+ (Thermo, USA) under helium atmosphere. Additionally, pH measurements were carried out according to the literature 33, by using a Digimed DM-22 pH meter. Briefly, 0.40 g dry activated carbon sample was added to 20 mL distilled water, which were kept under mechanical agitation overnight (about 16 h). pH of the solution was then measured at 298 K. X-ray diffraction patterns of the samples were performed on a PANalytical X’PERT High Score’s diffractometer, which was equipped with Cu-Kα irradiation. The diffraction data were collected in step scans, with a step size of 0.02° (2θ) and a count time of 10 s per step between 5 and 80° (2θ). X-ray photoelectron spectroscopic (XPS) measurements were carried out using a Physical Electronics PHI 5700 spectrometer equipped with a non-monochromatized Mg Kα source focused to a spot size of 720 µm. The binding energy scale was calibrated with respect to carbon (C1s) peak at 284.8 eV. Infrared spectra were obtained with a Vertex 70V (Bruker Optics inc., Billerica, MA, USA) with samples diluted in KBr. Textural properties of the activated carbon samples were evaluated from N2 and CO2 adsorption/desorption isotherms at 77 K and 273 K, respectively, using an Autosorb-iQ3 (Quantachrome Instruments, USA). The samples were outgassed at 453 K under vacuum (10-6 bar) during 6 hours. Specific surface area of all materials was calculated using Brunauer-EmmettTeller (BET) equation, total pore volume according to the Gurvich Rule, and micropore volume by Dubinin–Radushkevich (DR) model 34-35. Pore size distribution (PSD) of each sample was obtained using a Non-Local Density Functional Theory (NLDFT) kernel for slit-shaped pores 3438

. 6 ACS Paragon Plus Environment

Page 7 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2.3 H2S retention capacity In order to evaluate and compare the H2S sorption capacity, dynamic tests were performed in a fixed-bed column (0.4 cm of internal diameter and 7.40 cm of length) containing approximately 0.7 g of sample, placed inside a GC oven (Varian CP-3800). The outlet of the column was coupled with a Pulsed Flame Photometric Detector (PFPD) to detect sulfurcontaining species. The column temperature and the detector signal were monitored with a computer, as illustrated in Figure 1. The inlet concentration of H2S was 100 ppm in He, as declared by the supplier and the flow rate was set with the aid of mass flow controllers. The packed column was kept at 423 K for 12 h under He flow (15 mL min-1) prior to each experiment, which consisted of feeding the column with H2S (100 ppm) under a flow rate of 100 mL min-1 at a constant temperature (typically 298, 323 and 348 K). For the regenerability tests, the same procedure was used only for the temperature of 298 K. After the breakpoint was reached, the feed was kept for 5 minutes and then the adsorbent bed was regenerated with the same conditions as reported above (12 h at 423 K under He flow of 15 mL min-1). Breakthrough followed by He purge (1 cycle) were carried out 4 times. The upper detection limit of the FPDP detector (3.5 ppm of H2S) did not allow measuring the complete breakthrough curves (until outlet concentration equals the inlet concentration). Therefore, the retention capacity at breakpoint was the criterion used to compare the performance of the samples adsorbing H2S at different temperatures, instead of the saturation retention capacity (which would require the complete curve). The “breakpoint” time was defined in this work as the time elapsed since the column started being fed with 100 ppm H2S until the detector signal reached 10% of its full range. For systems with negligible mass transfer zones, i.e. relative low mass transfer resistances, the breakpoint time is not expected to be significantly different 7 ACS Paragon Plus Environment


Page 8 of 51

from the stoichiometric time 39 and hence the breakpoint retention capacity should be similar to the saturation retention capacity.

Figure 1. Fixed bed unit for determination of H2S breakthrough curves.

The amount of H2S (mg g-1) retained in the column at the breakpoint (qbt) was calculated according to Equation 01, adapted from Choo et al. (2013) 17: ‫ݍ‬௕௧ ൌ

஼బ ௧್ ொ ெೌ೏ೞ

‫ܯܯ‬ுଶௌ

(01)

Where Co is the feed molar concentration (mmol L-1), corresponding to 100 ppm; tb is the breakpoint time (min); Q is the volumetric flow rate (L min-1); MH2S is the molar mass of H2S (mg mmol-1); and Mads is the mass of adsorbent in the bed (g).

3. Results and Discussion 3.1 Characterization 3.1.1. Surface chemistry 8 ACS Paragon Plus Environment

Page 9 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The amount of metal species present on the surface of the samples was quantified by XRF. Results confirm the presence of impregnated species, as shown in Table 1. Significant amounts of impregnated Fe and Na can be observed in samples Desorex K43-Fe and Desorex K43-Na, especially when compared to the parent carbon matrix (Desorex K43).In addition to sodium, sample Desorex K43-Na shows a relevant content of calcium and iron, which are already present in the precursor (Desorex K43). A similar amount of calcium was observed for the Desorex K43-Fe sample, with a higher content of iron due to Fe2O3 added to the starting matrix upon impregnation. Among the metals known to augment H2S adsorption on impregnated porous solids, three of them (Na, Ca and Fe) are present in relevant amounts in the Desorex K43-Na sample, while only two (Ca and Fe) are in the Desorex K43-Fe sample. The carbon matrix Desorex K43 also presented Fe and Ca, but in much lower levels. The pH values determined for the samples are listed in Table 2 and were correlated with the data obtained by XRF. Table 1. Chemical composition of the samples by XRF. Element (wt.%)

Desorex K43

Desorex K43-Fe

Desorex K43-Na

Si

1.98

3.36

3.34

Na

-

-

4.74

Al

2.04

2.82

3.00

Fe

1.28

5.04

3.86

Ca

1.75

2.42

2.81

S

0.54

0.71

1.01

Mg

0.33

0.51

0.47

Ba

0.09

0.12

0.46



Page 10 of 51

Table 2. pH values of the carbon materials at 298 K. Materials

pH

Desorex K43

7.10

Desorex K43-Fe

7.20

Desorex K43-Na

10.19

While Desorex K43 and Desorex K43-Fe samples presented pH values close to 7.0, the sodium-containing carbon exhibited a high pH value. These results correlate well with XRF data (Table 1), because the alkaline sample presented high levels of sodium and calcium, which should be a determining factor in the removal of acidic gases such as H2S 20. The absence of sodium in the other samples is associated with their neutral surface, as indicated by the pH values. X-ray diffraction patterns (Figure S1) agree with those of typical activated carbons with graphitic carbon diffraction peaks at (002) and (100) planes. The detected peaks could not be assigned to any of impregnated phases and, as observed, the impregnated phases are amorphous 40

. The crystalline phases, if present, are expected to be in very small amounts to be detected by

the technique. IR spectra (Figure S2) are very similar for all samples and the most relevant bands account for presence of adsorbed water (-OH) and oxygen chemically bonded to aromatic rings, matching with XPS spectra for O1s. The low surface concentration of impregnating agents shown in XPS spectra compared with bulk concentration values, reveal good dispersion of the impregnated species in the corresponding samples. Despite the noise attributed to the low concentration, XPS spectrum of Fe2p3/2 (Figure S3) confirmed the form Fe2O3 for the species impregnated in sample Desorex K43-Fe 41. 10 ACS Paragon Plus Environment

Page 11 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Regarding sample Desorex K43-Na, XPS shows a clear peak corresponding to sodium (Figure S4), but from the Na signal in the XPS spectrum alone, it is not possible to determine whether it corresponds to the form of NaOH or Na2CO3. Nevertheless, some evidences from our CO2 adsorption isotherms experiments imply carbonate as the actual form of sodium on the surface. Adsorption isotherms of CO2 at different temperatures (Figure S5) present reversibility and the adsorption capacity decreases with temperature, behaving as a system with no reaction (chemisorption) between CO2 and Na, which would be rather unexpected to happen if the latter was in the form of NaOH on the surface. Furthermore, when compared to the literature, FTIR spectra (Figure S2) does not indicate any evidence for the presence of NaOH through its most representative IR peak around 3600 cm-1 42.

3.1.2 Textural characterization N2 adsorption isotherms Figure 2 shows the N2 adsorption/desorption isotherms at 77 K. All studied samples presented isotherms of type I (a), which is typical of microporous materials, according to the IUPAC classification 35. In this type of solid, a large portion of the gas volume is adsorbed at low relative pressures, which is related to the filling of the smaller pores 34. The existence of discrete hysteresis loops in the isotherms suggests the presence of larger pores, since the phenomenon is generally associated to capillary condensation in mesopores. Table 3 summarizes the textural characteristics of the samples obtained from the isotherms presented in Figure 2. All samples have high surface areas and are potentially interesting for the adsorption of gases. It is observed that the original carbonaceous matrix (Desorex K43) presented the largest surface area as compared to the impregnated samples, which 11 ACS Paragon Plus Environment


Page 12 of 51

is attributed to the chemical impregnation process and eventual pore occlusion in the latter ones. The incorporation of sodium hydroxide and iron oxide into the pores of the carbonaceous matrix has probably led to a decrease in the total available pore volume, apparently more significantly for the Desorex K43-Na sample. Table 3. Textural properties calculated from adsorption of nitrogen at 77 K. Materials

specific surface area (m2 g-1)

total pore volume (cm3 g-1)

micropore volume (cm3 g-1)

microporosity (%)

Desorex K43

1003

0.48

0.37

77

Desorex K43-Fe

952

0.43

0.36

84

Desorex K43-Na

815

0.38

0.30

79

Figure 2. Nitrogen adsorption isotherms of the samples at 77 K.

The microporosity, represented by the ratio of micropore volume to the total pore volume, is also a performance indicator for the adsorption of small molecules 19, 43.This parameter was used to 12 ACS Paragon Plus Environment

Page 13 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


monitor the effects of the continuous adsorption of H2S on the structure of the materials, as discussed later. Pore Size Distribution (PSD) was determined from N2 adsorption isotherms at 77 K by applying a NLDFT slit-shaped pore kernel 44-45, which is considered as the most reliable method for the evaluation of the porous structure of activated carbons 46. Figure 3 shows the PSDs for all studied samples. One may note that all samples present PSD centered in the micropore region, which was already expected based on the shape of the isotherms. Most of the pores are smaller than 15 Å and the volumetric contribution of mesopores (above 20 Å) is relatively low, which is in agreement with microporosity values above 77%.

Figure 3. Pore size distributions for the initial carbons obtained according to N2 adsorption isotherms.



Page 14 of 51

Observing the pore size distributions by N2 adsorption of the studied carbon samples and considering the kinetic diameters of H2S (3.6 Å) 47 and N2 (3.64 Å) 48, one might come to a simplistic conclusion that all pores accessed by N2 should be also accessed by H2S. However, the PSDs obtained from N2 isotherms fail to consistently assess pore sizes below 4 Å, due to the extremely slow diffusion of the N2 molecule in these pores at 77 K. For that reason, CO2 adsorption isotherms of at 273 K, a considerably higher temperature, were obtained and are shown in Figure 4. CO2 has a kinetic diameter of 3.3 Å 48 and, at 273 K, the diffusion problems related to N2 adsorption at 77 K are not present. New PSDs were then evaluated from CO2 data, by applying the corresponding NLDFT slit-shaped pore kernel to verify the existence of smaller pores, as exhibited in Figure 5. The PSDs obtained from both N2 and CO2 isotherms are compared in Figure 6, showing that in the range of larger pores (> 4 Å) the distributions exhibited similar behavior for each sample.

Figure 4. Carbon dioxide adsorption isotherms of the samples at 273 K.


Page 15 of 51 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Unlike N2, Desorex K43-Na sample presented higher CO2 adsorption capacity over the whole measured pressure range, followed by Desorex K43-Fe and Desorex K43 (Figure 4). The data in this particular pressure range have allowed observing a different trend with respect to the pore texture of the adsorbents. The different trend in adsorption presented by CO2 isotherms suggests that N2 was unable to access pores smaller than 4 Å. The PSDs obtained from CO2 isotherms reveal a significant volume of pore sizes within the region of ultramicropores (pore size

Insights on the Mechanisms of H2S Retention at Low Concentration

Recommend Documents