Raw oak wood (OW) was successfully converted into activated carbons through a two steps process, i.e. carbonization and activation.
Pyrolysis of wood conducted at high temperature (i.e. 800 °C) resulted in a higher extent of carbonization compared to that obtained by hydrothermal treatment. In fact, fixed carbon found for pyrolyzed wood (OW800) was twice as
100 200 300 400 500 600 90.5 91.0 91.5 92.0 92.5 93.0 93.5 (b) Sample weight Temperature W (%) t (min) (a) 50 60 70 80 90 100 110 120 T (°C) 1 2 3 4 5 6 7 8 9 10 0 20 40 60 80 100 CO2 sorption capacity CO 2 sorptio n ca pa city (%) Cycle number 40 60 80 100 120 140 160 91.0 91.5 92.0 92.5 93.0 Sample weight Temperature W (%) t (min) 50 60 70 80 90 100 110 120 T (°C)
large (66 wt% vs 33 wt%) that measured for hydrochar (OW250). In contrast, raw material was not completely devolatilized after hydrothermal treatment. Consequently, hydrothermally carbonized wood did not show any well-defined porous structure, whereas some degree of microporosity was already formed after pyrolysis. Both carbonization processes were found to be repeatable in terms of char yield. Additionally, pyrolysis of wood was not affected by the type of rig used. On the other hand, mass balance carried out for HTC process water showed a slightly higher variability. Nevertheless, relatively low errors were obtained for all parameters, thus indicating acceptable repeatability of the carbonization processes.
Synthesized chars were then activated by pursuing two different routes. Physical activation of biochars was optimized by maximizing surface area as a function of activation temperature and dwell time. Results revealed a stronger influence of the former parameter on the texture development of activated hydrochar. However, a remarkable effect of the activation holding time on the textural parameters of pyrolyzed wood was observed. In particular, surface area dramatically increased when increasing the dwell time from 0 to 1 h, whereas significantly decreased when prolonging the holding time from 1 to 2 h. Optimal conditions were found to be 600 °C for 30 min and 800 °C for 1 h for hydrochar and pyrolyzed wood respectively. After optimal heat-treatment, hydrochar (OW250) was subjected to a higher carbon burn-off (46 wt%) than that measured for OW800 (34 wt%). This was likely because of a further devolatilization experienced by the hydrothermally carbonized wood during activation. Yet, limited values of surface areas were obtained for physically- activated carbons, i.e. 415 m2∙g-1 for OW250PA and 627 m2∙g-1 for OW800PA.
Conversely, KOH activation followed by acid washing led to far more dramatic texture enhancement than that achieved through physical (CO2)
activation. In particular, the highest surface area (2757 m2∙g-1) was achieved
after chemical activation of hydrochar (OW250CA). In addition, a washing step was found to be essential in order to exploit the porosity created by KOH carbonation. Particularly, porosity of unwashed sample was completely inaccessible. This was ascribed to potassium carbonate residues blocking the pores. The presence of residual potassium carbonate was proved by combining FTIR and EDX results. In addition, after acid washing of OW800CA, surface area increased by 50% compared to the value obtained after washing the sample with water only. This was due to a more efficient removal of potassium- based compounds deposited within the carbon pores after heat treatment. Furthermore, especially for OW800CA, chemical treatment resulted in a higher purity of the carbon matrix, as most of the inorganic impurities were dissolved by acid. On the other hand, chemical activation involved some drawbacks. First, especially after KOH activation of hydrochar, a very low yield was obtained (less than 3 wt%). Therefore, the production of this material on a larger scale would be uneconomical. In contrast, larger activation yield were obtained for
OW800CA (ca. 44 wt%). Yet, it is worth mentioning that a very light powder with low mechanical strength was obtained as final product. This could represent an issue if an adsorption column was to be packed with this material, as mechanically strong particles would be preferred in industrial applications.
Interestingly, activation treatments of hydrothermally synthesized wood led to much more dramatic texture development than that experienced by pyrolysis- derived char. As aforementioned, this was mostly due to the additional devolatilization that occurred when submitting hydrochars to the activation process. Based on observation reported, it could be inferred that physical activation of hydrochar might be a more energetically convenient route for the preparation of porous sorbents, especially when starting from wet feedstock containing low amount of ash.
According to CO2 adsorption isotherms measured at 0 °C, activation
methods gave rise to carbons having a very low proportion of narrow microporosity. NLDFT model was believed to give a more reliable estimation of the ultramicropore volume as a stronger correlation with the volume of CO2
adsorbed at 0 °C and 1 bar was found.
SEM imaging coupled with EDX analysis allowed for detection of Ca- containing particles attached to the carbon structure of oak wood derivatives. XRD patterns confirmed the presence of Ca-based crystalline phases for all samples except KOH-activated hydrochar (OW250CA), as inorganic fractions were mostly removed after HCl washing. In particular, hydrated forms of calcium-containing compounds (calcium oxalate and calcium hydroxide) were identified within the structure of raw and carbonized wood, while physically activated chars exhibited calcium carbonate as main crystalline phase. This might have arisen after carbonation reaction occurring throughout CO2
activation. On the other hand, amorphous fractions of calcium hydroxide (not detected) may have formed following to a slow rate decomposition of calcium carbonate at 800 °C. However, larger proportion of alkali metal-containing species were believed to be responsible for the higher basic character possessed by OW800PA. Indeed, as shown by Boehm titration, this sample exhibited a far larger number of basic functionalities compared to that measured for its chemically-activated analogue (OW800CA) and for a commercial carbon (GAC) included for comparison purposes.
All carbons exhibited fast CO2 sorption kinetics at 35 °C and 1 bar,
reaching saturation within less than 5 min. OW800CA attained the largest sorption capacity (around 70 mg CO2∙g-1). This figure was higher than that
exhibited by the commercial carbon (GAC), but far lower than highest values reported by literature. Lower CO2 uptakes measured in this study were mostly
attributed to the undeveloped ultramicroporosity of the materials tested. This confirms the importance of narrower microporosity in the CO2 uptake process at
suggest that CO2 sorption at 35 °C and 1 bar might be governed by a
combination of surface area (i.e. larger porosity) and narrow microporosity. Unexpectedly, under simulated post-combustion conditions (53 °C, 0.15 bar) the physically activated carbon (OW800PA) was more selective at capturing CO2 than other sorbents with well more developed texture (i.e.
OW800CA and GAC). This finding was ascribed to the larger basicity measured for OW800PA, which in its turn was related to Ca-containing species intrinsically present within the structure of the CO2-activated wood. It seemed that alkali
metal-based inorganic matter ensured a stronger interaction between CO2 and
the physically-activated carbon. This agreed with the higher energy of adsorption measured between the sorbent and CO2 at 0 °C. Inorganic species
were mostly removed after chemical treatment, by acid washing, and were not present within the commercial carbon. Since OW800CA and GAC were not as selective as OW800PA under post-combustion conditions, the findings clearly indicated that texture is no longer the predominant parameter at higher temperature and lower partial pressure (i.e. post-combustion condition). In contrast, basicity appeared to be the key factor to be considered when designing selective sorbents for post combustion capture.
Along with increased selectivity, OW800PA also exhibited a more facile desorption step and was fully regenerated at 100 °C. This suggested that a very low temperature could be applied for the regeneration of the material over multiple RTSA cycles. Accordingly, heat required for the sorbent regeneration would be reduced, thus optimizing energy costs. Furthermore, excepting an initial loss after the first adsorption-desorption cycle, the sorbent capacity seemed to attain a plateau over the remaining 8 cycles, thereby indicating good durability over the time.
In conclusion, traditional pyrolysis followed by chemical activation seemed to be the most suitable route to produce cost-effective CO2 (physi)sorbents at
higher partial pressure of CO2 and lower temperature. In addition to this, raw
oak wood revealed to be an advantageous precursor for an eco-friendly synthesis of selective CO2 sorbents under post-combustion conditions. In
particular, alkali metal (Ca)-containing compounds intrinsically incorporated within the raw material allowed producing highly basic sorbents without applying any chemical modifications. Physical activation is preferred to chemical treatment for preparation of CO2 sorbents at low partial pressure, as it does not
remove basic inorganic functionalities, which are evidently vital under post- combustion conditions. Additionally, physical treatment is less contaminant than KOH activation.
5 Sustainable and regenerable alkali metal-containing
carbons derived from seaweed for post-combustion capture
of CO2
5.1 Outline
This chapter focuses on the use of a widely available feedstock, seaweed, for a sustainable preparation of alkali metal-containing activated carbons (ACs). In particular, attention was directed toward a type of brown macroalgae (i.e.
Laminaria hyperborea), which was selected as precursor for ACs fabrication.
The raw material processing was similar to that pursued for the oak wood (see Chapter 4). A first step entailed obtaining chars either by dry pyrolysis or by hydrothermal carbonization (HTC). Results related to the repeatability of the carbonization of Laminaria hyperborea are reported in section 5.2.
Following this, pyrolyzed and hydrothermally carbonized Laminaria
hyperborea were activated by pursuing two different routes, i.e. either physical
(CO2) or chemical (KOH) activation. In section 5.3, the optimal conditions are
systematically determined for both activation procedures. In particular, the Brunauer-Emmett-Teller (BET) surface area of the carbons is maximized as a function of chosen parameters. As concerns the CO2 activation, in line with
experiments already done for the oak wood in Chapter 4, the influences of activation temperature and dwell time on the textural properties of the carbons were assessed. With respect to the chemical treatment, the effects of activation temperature and KOH to char ratio were examined.
In section 5.4, the textural parameters of the raw macroalgae and all its derivatives (i.e. chars and optimally-activated carbons) are compared. In addition, the CO2 adsorption at 0 °C was also measured in an attempt to
quantify the narrow microporosity of the samples.
The alteration of the chemical properties of the raw Laminaria hyperborea induced by carbonization and activation processes is discussed in section 5.5. However, some of the characterizations (i.e. X-Ray Diffraction (XRD) and Boehm titrations) were conducted only for some of the macroalgae-derived materials (see section 5.5.3). This choice was mostly dictated both by the large amount of material required by these techniques and by the scarcity of the raw feedstock. In addition to this, in accordance with the strategy already described in Chapter 4, the samples of interest were selected according to their CO2
capture performance (see section 5.6). In particular, a first test carried out at lower temperature (35 °C) and higher partial pressure of CO2 (1 bar) was used
uptake and sorption kinetics. Based on results observed, additional characterizations and a post-combustion test were carried out only for pyrolyzed
Laminaria hyperborea (LH_S800) and its activated counterparts (LH_S800PA
and LH_S800CA).
Finally, a summary of the chapter is given in section 5.7.