This section is related to the analysis of the correlation between the selected sample’s textural properties measured at the DISAT—Applied Science and Technology Department (Polytechnic of Turin, IT)—laboratory and the sample’s adsorption capacity obtained through experimentation performed at DENERG—Energy Department (Polytechnic of Turin, IT)—laboratory and at the Edmund Mach Institute (Trento, IT).
2.1. Materials
Five selected commercial materials for micro-contaminants adsorption were analyzed to determine the performances of the sample in terms of H2S and D4 adsorption capacities:
CKC is a carbon of mineral origin extruded to obtain small cylinders with a particle diameter of 4 mm, steam activated, and impregnated with potassium bicarbonate at 5%. CKI is a carbon of mineral origin extruded to obtain small cylinders with a particle diameter of 4 mm, steam activated, and impregnated with potassium iodide at 2%. C64 is a carbon of mineral origin extruded to obtain small cylinders with a particle diameter of 4 mm, and steam activated with an alkaline pH. CKC and CKI are suggested by the producer for sulfur removal, while C64 is recommended for siloxane removal.
R8G is an activated carbon with a highly dispersed mixed metal oxide active phase with modifiers prepared over a porous support [
42].
R7E is a metal oxide that contains copper oxide. The information on the composition was not available from the producer but was obtained through energy dispersive X-ray spectroscopy (EDS) at the DISAT laboratory. This sample also contained manganese, aluminum and silicon (detailed results are presented in the
Appendix A). R8G and R7E are both recommended for sulfur removal from biogas.
2.2. Evaluation of Textural Properties
A gas sorption analyzer Quantachrome Autosorb 1 (Boynton Beach, FL, USA) was adopted for the determination of adsorption isotherms for N
2 at 77 K. Samples were outgassed at 423 K overnight before the adsorption measurements. The experimental equipment allows measurement of the relative pressure until 10
−6 p/p
0. Micropore volumes were determined using the
t-Plot method in the relative pressure range of 0.15–0.3. For carbon-based materials, the pore size has been evaluated through the DFT method (density functional theory), using the NLDFT (nonlocal density functional theory) equilibrium model for slit/cylindrical pores. The BET method is “the most widely used procedure for evaluating the surface area of porous materials” [
43].
The t-Plot method determines the presence of micropores in a solid by comparing the material adsorption isotherm with a reference one, specific for the material under investigation. This method allowed to determine the micropore volume—V micropore (cm3/g)—of the samples.
The DFT method provides a “reliable approach to pore size analysis over the complete nanopore range” (up to 100 nm); there are different pore shape models developed for various material classes, such as carbons [
43]. The total pore volume—
Pore volume (cm
3/g)—of the samples was obtained through this method.
A scanning electron microscopy (SEM) (FEI Inspect, Philips 525 M) coupled with EDS analysis (SW9100 EDAX) was adopted to characterize the sorbent samples. The localized chemical analysis was achieved by EDS (energy dispersive X-ray spectrometry). Qualitative analysis was achieved through the identification of the lines reported in the spectrum, while quantitative analysis (determination of the percentage elemental composition) was performed through the measurement of line intensities for each element present in the sample. The results of this analysis are presented in the
Appendix A.
Figure 1 shows a series of SEM images of R8G with increasing magnification: 5000, 25,000, 50,000, 100,000 times. The SEM images illustrate the porous structure of the AC, highlighting the presence of discontinuity in the sample’s surface. The presence of high porosity is necessary to limit the energy losses from the blower system to filter the biogas. As reported in some of our previous studies, SEM images coupled with an EDS analysis can be useful to highlight the sulfur adsorption and retention within the pores and metal-based catalytic sites [
15,
44].
2.3. Evaluation of the Adsorption Capacity
Laboratory tests were performed to evaluate the adsorption capacity of the chosen materials. The tests were performed using three different reactors filled with carbon:
Microreactor 4 mm: the first set of experiments was made using micro-reactors with an internal diameter of 4 mm, filled with ground carbon, sieved to select particles with size of 50 µm–70 µm; the carbon retainment was obtained with medical gauze. Closing of the reactor was obtained with a nut and washer made of steel with sulfur-inert coating.
Large reactor: the last set of experiments was made using a reactor constructed in the laboratory, with an internal diameter of 25 mm, filled with carbon as-received from the supplier; carbon blocking was obtained through using cotton wool.
The reactors were fed with simulated biogas from cylinders: the biogas composition was chosen based on average values from measurements performed at the DEMOSOFC site: 62.5 %vol. methane and 37.5 %vol. carbon dioxide. Hydrogen sulfide (H2S) was fed into the system using methane cylinders with the fixed concentration of the contaminant (usually between 100 and 1000 ppmv).
The sensor used for the detection of hydrogen sulfide was an electrochemical device that could measure from 0 to 200 ppmv (Transmitter MECCOS eTR H2S 0–200 ppmv, Siegrist GmbH, Karlsruhe, DE); the error of measurement increases linearly from 0 to 200 ppmv at full scale. To find the percent error, first, we averaged all our measurements. Then, we found the difference between our average and true values. The sensor has flow limits between 500 mL/min and 1000 mL/min, so all the experiments were performed considering this flow range. In all published manuscripts, authors refer to the volumetric flow rate defined under normal conditions. A calibrated sensor was used to verify the value of the measured H2S level, calculating also the signal delay. Compressed air was used to purge the sensor at the end of the tests.
The aim of these experiments was to calculate the quantity of contaminant that was adsorbed by the carbon filters, using the experimental apparatus presented in
Figure 2. The experiments allow the determination of the adsorption curve, which is the evolution of the contaminant concentration over time at the outlet of the fixed bed. This concentration will rise, following a certain trend, from zero (when the filter is completely removing the contaminant) to the inlet contaminant value (saturation, when the sorbent is no longer filtering any contaminants and the outlet concentration is equal to the inlet one). The area below the adsorption curve is the amount of contaminant adsorbed by the carbon material. From this knowledge, the adsorption capacity
Ads Cap (mg/g), defined as an evaluator of the sorbent’s performance, can be calculated.
The adsorption capacity can be typically evaluated at breakthrough or saturation.
Breakthrough adsorption capacity: This value refers to the moment when the concentration of the contaminant after the filter reaches the C/C
0 fixed threshold value (usually 1% in the literature), where C (mol/m
3) is the actual H
2S concentration in the gas phase and C
0 (mol/m
3) is the H
2S inlet concentration. The C/C
0 ratio is fundamental and should be set accurately when the cleaning system aims to protect a device downstream of the cleaning section (for example a fuel cell system that could be seriously damaged by contaminants at very low levels, <1 ppmv). The adsorption performance is influenced by the slope of the saturation curve profile. If the slope is shallow, the saturation is reached earlier even if the ideal linear front is far from the bed outlet. The length of the mass-transfer zone (MTZ), where adsorption takes place [
45], is responsible of the breakthrough time.
Saturation adsorption capacity: This value is related to the maximum value of adsorbed contaminant into the filter; it is evaluated at the moment when the amount of contaminant at the exit of the bed is the same as the inlet value (C = C0). This value could be useful in case a series of beds are available in the system. In this case, the first cleaning vessel could be operated even after breakthrough and until saturation (to fully exploit the material) because other beds in series are available.
If the slope of the curve is steep, the difference between these two adsorption capacities is low. In the experiments, we fixed the breakthrough concentration limit for the fuel cell feeding and evaluated the adsorption capacity at this breakthrough condition. The adsorption capacity was evaluated as:
where
is the biogas volumetric flow rate used for the experiment
CS is the sulfur volumetric concentration
,
MWS is the molecular weight of sulfur (g/mol),
tBT (s) the breakthrough time (retrieved from the experimental adsorption curve) and
mAC is the mass of sorbents contained in the reactor.
The time considered in the formula—tBT (s)—is the time at which there is a breakthrough, defined as the moment in which the concentration downstream of the filter starts to be different from zero. The breakthrough time was measured considering the value of the outlet concentration measured by the sensor. This was determined as the first point of the outlet concentration measured by the sensor that was higher than zero.