Effect of distribution patterns of refractory overlayers on cyclic high temperature CO₂ capture using waste oyster shell

Abstract

Waste oyster shell powder has been applied in cyclic high temperature CO₂ capture, but the sintering effect during calcium looping (carbonation/calcination) significantly shortens the life span of the adsorbents. The association of waste oyster shell with refractory materials, such as zirconium oxide has been evaluated, using either furnace treatment or atmospheric pressure plasma treatment to improve the thermal stability. It was noted that samples with 5 wt% of ZrO₂, by either furnace treatment or plasma treatment, exhibited the highest CO₂ capture capacity (∼360 mol-CO₂ per mol-CaO in 20 calcium looping cycles). The plasma treatment was found to quite effectively enhance the thermal stability when more ZrO₂ was added; however, evaporation–condensation reactions during plasma treatment led to evenly distributed CaZrO₃ particles, which hindered the access of CO₂ to CaO, and therefore reduced the overall CO₂ capture capacity. In contrast, furnace treatment leads to most of the ZrO₂ functioning as wedges to mitigate the sintering effect, which essentially exposes more CaO for CO₂ access and accounts for its high overall CO₂ capture capacity. Accordingly, the distribution patterns of refractory overlayers on waste oyster shell play an important role in cyclic high temperature CO₂ capture.

---

Introduction

In modern society, waste management is usually complicated because it involves a tradeoff between the economic perspective and public health. Among the options that exist, waste reutilization is of high priority from the perspective of green chemistry because it simultaneously reduces the amount of waste and the cost of waste treatment.¹ Waste oyster shell, for example, is estimated to have about 4200k metric ton production in Taiwan.² The high calcium content allows it to be a high temperature CO₂ sorbent for post-combustion CO₂ capture, in which carbonation by reacting CaO with CO₂ forms CaCO₃, while CaO is recovered by releasing CO₂ in the calcination of CaCO₃.^3–14 The most competitive benefit of calcium looping is its high efficiency in capturing CO₂ from flue gas, and subsequently releasing high-purity CO₂ in the calcination stage. The as-obtained, high-purity CO₂ can be a green, raw material for further CO₂ utilization in several value-added chemicals through (photo)catalytic reactions.^15–17

The weakness of waste oyster shell is its low surface area, which essentially limits the CO₂ capture capacity. The rapid loss in pore volume (and therefore the surface area) resulting from the formation of larger grains was reported to be responsible for deactivating the CO₂ uptake performance.^3,8,11 This explains why the synthesized CaO-based sorbents that often possess porous structures usually exhibit higher CO₂ capture capacity than sorbents obtained from waste reutilization.^4,7,12 However, byproduct recovery, such as high-purity CaO, along with waste reduction, essentially make waste oyster shell a favorable source for preparing high temperature CO₂ sorbents (1). To capture CO₂ in an efficient way, and ultimately obtain a negative carbon footprint process, suppressing the pore-filling and pore-plugging during the carbonation process is important.^8,13 This is achieved by associating high Tammann temperature elements^18–22 such as Zr(5) or Al^7,11 with nanostructured CaO-based sorbents. Our previous results² suggested that waste oyster shell powders could be efficient CaO-based absorbents, but like other synthesized sorbents prepared by waste Ca-rich materials,^6,9,10,13 they also suffer pore reduction in calcium looping. In this study, we mixed waste oyster shell powders with ZrO₂ particles to enhance their thermal resistance against the sintering effect. In addition to conventional thermal treatment using a furnace, the technique of atmospheric-pressure plasma treatment was also introduced and assessed. Our results indicate that the unique environments, such as relatively higher temperature, shortened treatment time and ionized environment, lead to the formation of an evenly distributed surface coating layer.^23–25 This feature allows us to evaluate the effect of the distribution patterns of refractory overlayers on cyclic, high temperature CO₂ capture and hence, develop an effective surface treatment method to improve the performance of CO₂ sorbents prepared by waste reutilization.

Experimental

All chemicals, including ZrO₂ powders (5 μm, 99% trace metals basis), were ACS grade, purchased from Sigma-Aldrich and Merck, and used as-received without any further purification. Waste oyster shells were collected from a local oyster farm located in eastern Taiwan (>98% CaCO₃, quantified by XRD analysis). They were pulverized until they could pass through a 200-mesh sieve (particle size < 0.075 mm).² To prepare zirconium oxide associated CaO-based sorbents, the pulverized waste oyster shells were blended with x wt% amount of zirconium oxide (x = 0, 5, 10, 20 or 30) (Table 1). These mixtures were vortexed for 5 min in polypropylene centrifuge tubes and then the mixtures were spread over ceramic plates for atmospheric-pressure plasma treatment (200 W power radio frequency; refers to P-series samples) for 10 minutes. Another set of mixtures was subjected to thermal treatment in a box furnace at 600 °C for 2 h in ambient conditions (refers to F-series samples). Both thermally treated samples were then cooled at room temperature. The as-obtained CaO-based CO₂ sorbents were manually ground using a pestle and mortar and stored under ambient conditions prior to use.

Table 1 Composition of ZrO₂-stabilized CaO-based waste oyster powders prepared in this study

Treatment methods	Sample label	CaCO3 (g)	ZrO2 (g)	Zr/Ca mole ratio
Plasma (P) or furnace (F)	P0/F0	5	0	0
	P5/F5	4.75	0.25	0.04
	P10/F10	4.50	0.50	0.09
	P20/F20	4.00	1.00	0.20
	P30/F30	3.50	1.50	0.35

---

Characterization of these CaO-based sorbents was performed using X-ray diffraction (XRD) with Cu Kα radiation (D2 Phaser diffractometer, Brucker, Germany), nitrogen physisorption isotherms (Quantachrome Nova 4200 Series), scanning electron microscopy (SEM; JSM700F, JEOL) and high resolution transmission electron microscopy (HRTEM; JEOL JEM-2100). Their performance in cyclic high temperature CO₂ capture was measured using a thermogravimetric analyzer (NETZSCH TG 209 F1 Iris). Calcium looping (20 cycles) was conducted isothermally at 750 °C (temperature ramp: 20 °C min⁻¹) with a flow rate of 20 mL min⁻¹: 20 min in an atmosphere composed of 15% CO₂ and 85% N₂ for CO₂ uptake (carbonation) and 10 min in pure N₂ for CO₂ release (calcination). The CO₂ uptake ratio was defined as the gained weight in the carbonation step per gram of sorbents. Quality control was performed by conducting four batches of cyclic high temperature CO₂ capture experiments, and the one exhibiting the total CO₂ uptake capacity (sum of 20 cycles) closest to the average capacity was selected for characterization, and is presented in this study (RSD was 6–9%).

Results and discussion

Characterization of zirconium oxide associated CaO-based waste oyster shell powders

To examine the conversion degree of CaCO₃ in waste oyster shell powders to CaO at the end of the furnace and plasma treatment, their XRD spectra were collected and the mineral phases in these thermally treated samples were identified. At the end of the 10 minutes of plasma treatment, it seemed that a short period of plasma treatment was insufficient to completely convert waste oyster shells into calcium oxide. As shown in Fig. 1a, a visible diffraction peak appears at 29.4° 2θ, which is assigned to the (104) diffraction of calcium carbonate (JCPDS 04-007-8659), and its intensity gradually increases with increasing amount of zirconium oxide additives. Also, intense (111) and (200) diffraction peaks of CaO (JCPDS 82-1691) centered at 32.2° and 37.4° 2θ are noted to exist concurrently with CaCO₃. At the same time, diffractions from secondary phase CaZrO₃ (JCPDS 75-0358) appear when the amount of ZrO₂ additives exceeds 10 wt%. Several minor diffractions originating from ZrO₂ (marked with * in Fig. 1a) are also noted in sample P10, P20 and P30. Concurrently existing CaO, CaCO₃, ZrO₂ and CaZrO₃ phases reveal the unique treatment environment created by atmospheric pressure plasma. First, atmospheric pressure plasma is sufficient to activate zirconium diffusing into CaCO₃ particles and reacting with calcium. This solid state reaction leads to the formation of refractory CaZrO₃ perovskite secondary phase on the surface of the waste oyster shell powders.^13,26 This is a very interesting feature because it only takes 10 minutes to convert the CaO–ZrO₂ mixture into the CaZrO₃ phase during plasma treatment. Conversely, experiment and ab initio simulation suggest that a two-hour thermal treatment at 600 °C is required to activate the solid chemistry reaction between CaO and ZrO₂, leading to the formation of several refractory phases, such as CaZr₄O₉ and CaZrO₃ (depending on the CaO/ZrO₂ molar ratio).^27,28 Since sub-stoichiometric amounts of zirconium oxide were mixed with waste oyster shell powders, we therefore suspect the major phase in our samples would be the CaZrO₃ phase over the surface of the CaO particles, which is also supported by XRD analysis, as shown in Fig. 1a. The technique of atmospheric pressure plasma has been reported as capable of significantly changing/tuning the surface properties of metal oxides in a much shortened period of treatment time.^29,30 Our results unambiguously support these research findings, and further demonstrate that the surface property of CaO particles is significantly changed by plasma treatment, due to the formation of the surface CaZrO₃ phase.

Fig. 1 XRD pattern of the as-obtained ZrO₂ associated CaO-based sorbent prepared by (a) plasma treatment and (b) furnace treatment. * refers to the monoclinic ZrO₂ phase (JCPDS no. 37-1484) and the scale bars stand for intensity of 100 cpm.

Another interesting feature induced by plasma treatment is that the formation of the surface CaZrO₃ phase hinders, to certain degree, the further conversion of calcium carbonate into calcium oxide, which is realized because of the extremely high melting point and low thermal expansion of the refractory CaZrO₃ phase over the CaCO₃ surface.²⁸ This also explains the diffractions observed in the XRD pattern from some unreacted ZrO₂ particles (* marks in Fig. 1). Note that a similar XRD pattern is observed even after an elongated period of time of plasma treatment (30 minutes, data not shown due to superimposition), which means that the plasma treatment might not be able to completely convert calcium carbonate into the calcium oxide phase. However, this also emphasizes that the plasma treatment has a pronounced influence on the surface, rather than on the bulk of particles, because one of the most important energy transfer mechanisms during plasma treatment is through the collision between high energy ions (plasma) and the surfaces of objects (waste oyster shell powders in this study).

In contrast, thermal treatment using a conventional furnace seems to be able to completely convert waste oyster shell powders into calcium oxide, as no diffractions from calcium carbonate appear in F-series samples (Fig. 1b). Interestingly, diffractions from the CaZrO₃ phase start to appear when the Zr/Ca molar ratio exceeds 0.2. Our results substantially support the simulations^27,28 that a thermal treatment condition of 600 °C for 2 hours will develop the CaZrO₃ phase. Similar to results observed in P-series samples, some unreacted ZrO₂ was also found in samples F20 and F30. However, we speculate that the mechanism accounting for the ZrO₂ residue is somehow different for the two different treatment methods. In the plasma treatment, high-energy ions keep bombarding and evaporating the surfaces of the CaO and ZrO₂ particles. These evaporated Zr and Ca ions first mix into the plasma media and then homogeneously condense over the surface of these reactants. This evaporation–condensation reaction^24,25 highly limits the reaction occurring only on the particle surface, and this explains the relatively high amount of unreacted ZrO₂ and CaCO₃ observed in the P-series samples. The conventional high temperature thermal treatment leads to the formation of the condensed phase of CaO and ZrO₂ and their viscosity is rather different, given their approximately 200 °C difference in melting point (2715 °C vs. 2572 °C). This means that the ZaZrO₃ phase would only occur at the interface between individual CaO and ZrO₂ phases, and this is expected to be the origin of those unreacted ZrO₂ phases appearing in the XRD patterns (Fig. 1b). Note that the complete conversion from ZrO₂ to CaZrO₃ is always found (no unreacted ZrO₂ noted in XRD patterns) when samples are prepared through the sol–gel method, particularly for samples mixed with sub-stoichiometric amounts of Zr additive.^13,20 Indeed, incomplete conversion (i.e., relatively lower purity) is always the major flaw when choosing waste substances as the raw materials for further application. However, for the purpose of sustainability, waste reutilization in an appropriate way is always a high priority, since it would greatly reduce the amount of waste production and minimize the associated environmental impact, such as material acquisition and energy required during the waste treatment process. Aside from the incomplete conversion, the different distribution patterns of the CaZrO₃ phase resulting from the plasma and furnace treatments are shown in Fig. 2. It is of note that the plasma treatment allows dense nanoparticles with size of around 10 nm to homogeneously deposit over the surface of CaO particles (Fig. 2a), while conventional thermal treatment would lead to the development of coarse CaZrO₃ particles depositing on the CaO particles (Fig. 2b). Note that TEM images themselves do not give information about the chemical composition; additional elemental mapping studies in the same image region may be a more suitable alternative. Accordingly, additional EDX analyses were conducted and the results obtained substantially suggest that these dense particles possess high Zr content (data not shown). In addition to these dense nanoparticles, plasma treatment seems to induce the formation of a film-like appearance, while those subjected to furnace treatment would keep their solid particulate morphology at the end of the thermal treatment. According to TEM images, it is therefore expected that the waste oyster shell powders that were subjected to plasma treatment would have higher thermal resistance, due to the CaZrO₃ phase being evenly distributed over their surface.

Fig. 2 TEM images of sample with 5 wt% of ZrO₂ additives after treatment using (a) plasma (b) furnace.

Although the plasma treatment would result in a finer CaZrO₃ phase over the surface of CaO-based sorbents, it decreases the surface area, as suggested by the disappearance of the hysteresis loop in the N₂-BET isotherm (Fig. 3a). This is because the origin of a hysteresis loop is the capillary condensation of surface-bound nitrogen gas in the mesopores, and this mesoporous property is essential for a high-surface-area particle (otherwise, the surface area of an adsorbent is simply equal to the spherical surface area that is proportional to its radius). For P-series samples, it is noted the addition of ZrO₂ would gradually transform the porous environment of sorbents from a type II to a type IV isotherm. As indicated by XRD patterns shown above, the introduction of ZrO₂ would somehow inhibit the transformation between calcium carbonate and calcium oxide. Additional TEM images suggest that the plasma treatment would lead the surface of CaO sorbent to possess a film-like morphology. Based on these observations, the higher surface area of sample P0 compared to others is speculated to be as a result of the stacking of fine CaO particles produced by the evaporation–condensation reaction. On the other hand, evaporated Ca/Zr during plasma treatment would condense on the surface of CaO and would block the pores of the CaO sorbent. This accounts for the disappearance of the hysteresis loop (Fig. 3a) and significant surface area reduction with the addition of ZrO₂ (Fig. 4a). Unlike P-series samples, all F-series samples keep their hysteresis loops, independent of the amount of ZrO₂ added (Fig. 3b). As mentioned above, the CaZrO₃ phase only developed at the CaO/ZrO₂ interface, which explains the porous environment (owing to the stacking of waste oyster shell powder) being retained at the end of thermal treatment. Note that the determined surface areas and pore volumes between the P- and F-series samples are essentially comparable (Fig. 4), which was realized by considering the nonporous surface of the waste oyster shell powders; therefore, the determined surface areas and pore volumes are essentially from the stacking of particles.

Fig. 3 N₂ adsorption/desorption isotherms of ZrO₂-stabilized waste oyster shell powders subjected to (a) plasma and (b) furnace treatment. Inserted scale bars refer to 2.0 cm³ g⁻¹ STP of adsorbed N₂. Surface area measurements were conducted using the N₂-BET method with Micromeritics ASAP 2020. These samples were degassed at 110 °C for 24 hours to remove adsorbed moisture, prior to surface area measurement.

Fig. 4 Surface area and pore volume of ZrO₂-stabilized waste oyster shell powders subjected to (a) plasma and (b) furnace treatment.

Performance of ZrO₂-stabilized waste oyster shell powders

Fig. 5 shows representative TGA results of 20 cycles of high temperature CO₂ capture by 5 wt% ZrO₂-stabilized waste oyster shell powders subjected to plasma and furnace treatment. Note that while there was approximately 5% weight loss when the samples were heated up to 750 °C (emphasized by red rectangle), the patterns of weight loss were rather different (stepwise for sample P5 and one-step for sample F5). In 30 minutes (temperature ramp: 20 °C min⁻¹), the temperature had already reached 600 °C, which is a temperature that was reportedly high enough to initialize the conversion of CaCO₃ to CaO.³¹ The first step of weight loss of sample F5 (<120 °C) was due to the loss of adsorbed water, while the second was attributed to the conversion of CaCO₃ residue into CaO (>600 °C, insert, Fig. 5b). In contrast, three stages of weight loss appeared in sample P5, which were attributed to the release of adsorbed moisture, formation of CaZrO₃ overlayer and retarded conversion of CaO, respectively (insert, Fig. 5a). This results in two interesting features of plasma treatment. Firstly, the relatively homogeneously distributed surface CaZrO₃ phase enhances the thermal stability of sample P5, resulting in the retarded conversion of CaO (since both 2^nd and 3^rd stage of weight loss occurred at 750 °C). Secondly, this again indicates that thermal treatment using the furnace is more capable of completely converting CaCO₃ into CaO than the plasma treatment does. To quantitatively evaluate the results of 20 cycles of isothermal calcium looping at 750 °C, the results shown in Fig. 5 were rearranged based on the amount of CO₂ uptake in each cycle (Fig. 6a and b), and fitted using a first-order kinetic equation, y = A exp(−kt), where A refers to the content of adsorbent and k stands for the decay constant (Fig. 6c) (Table 2).

Fig. 5 Representative TGA results of 20 cycles of high temperature CO₂ capture by 5 wt% ZrO₂-stabilized waste oyster shell powders subjected to (a) plasma and (b) furnace treatment.

Fig. 6 Results of CO₂ uptake in each cycle of the calcium looping of samples subjected to (a) plasma and (b) furnace treatment; (c) decay constants derived from fitting (a and b) with a first-order kinetic reaction y = A exp(−kt); (d) proposed mechanism accounting for different distribution patterns by plasma- and furnace-treatment.

Table 2 The CO₂ uptake capacities (wt%-CO₂ per g sorbent) in each cycle and determined first-order kinetic reaction constants of CaO-based waste oyster powders prepared in this study

Fitting with first-order kinetic reaction, y = A exp(−kt)
	Plasma treatment					Furnace treatment
	P0	P5	P10	P20	P30	F0	F5	F10	F20	F30
R^2	0.896	0.880	0.916	0.963	0.890	0.852	0.882	0.862	0.896	0.907
A	27.182	25.317	21.429	18.726	16.928	26.066	26.534	24.959	2.787	20.815
k	0.045	0.039	0.038	0.039	0.038	0.042	0.043	0.042	0.043	0.043

---

Number of cycles	CO2 uptake capacity (wt%-CO2 per g sorbent) in each cycle
1	32.79	30.30	24.67	19.40	19.93	34.32	33.13	32.28	26.82	24.71
2	28.88	26.91	21.17	18.61	17.83	26.92	27.48	25.78	24.17	20.62
3	24.66	23.31	19.84	16.82	15.57	23.52	24.07	22.52	21.06	19.65
4	22.00	21.22	18.84	16.34	14.26	21.21	21.78	20.43	18.96	17.56
5	20.53	19.71	17.34	15.55	13.36	19.71	20.19	18.96	17.46	16.18
6	19.25	18.61	16.23	14.46	12.66	18.51	18.91	17.76	16.37	15.10
7	18.17	17.71	15.44	13.60	12.06	17.61	17.92	16.88	15.47	14.11
8	17.38	17.01	14.74	12.88	11.55	16.80	17.13	16.19	14.78	13.51
9	16.59	16.41	14.14	12.27	11.15	16.20	16.54	15.50	14.18	12.93
10	16.00	15.91	13.74	11.75	10.85	15.60	15.94	15.00	13.66	12.53
11	15.51	15.41	13.32	11.33	10.45	15.20	15.45	14.61	13.18	12.04
12	15.02	15.01	12.93	10.94	10.25	14.80	14.96	14.22	12.87	11.75
13	14.53	14.71	12.63	10.61	9.95	14.39	14.56	13.83	12.47	11.35
14	14.13	14.31	12.33	10.61	9.75	14.09	14.16	13.53	12.18	11.05
15	13.77	14.10	12.03	10.03	9.55	13.79	13.87	13.23	11.89	10.86
16	13.47	13.81	11.84	9.80	9.35	13.50	13.57	12.94	11.69	10.57
17	13.18	13.51	11.64	9.57	9.25	13.29	13.28	12.64	11.39	10.37
18	12.89	13.30	11.44	9.36	9.05	12.99	12.99	12.43	11.18	10.18
19	12.69	13.11	11.24	9.17	8.95	12.79	12.78	12.23	11.00	9.98
20	12.39	12.90	11.05	8.99	8.75	12.59	12.59	12.03	10.79	9.79

---

At the end of 20 cycles of carbonation/calcination, it was apparent that the introduced ZrO₂ had no significant benefit in enhancing CO₂ uptake capacity, because the surface CaZrO₃ phase would retard CO₂ diffusion into the CaO core. On the other hand, introducing ZrO₂ enhances the thermal stability of CaO sorbents, as indicated by the variations in the decay constant (k value), particularly for plasma treated samples. As indicated in Fig. 6c, the decay constants of P-series samples are significantly lower than the decay constants obtained for the furnace thermal treatment. Since a greater decay suggests the sorbent is more vulnerable to deactivation in a carbonation/calcination cycle, it is clear that plasma treatment is more effective in enhancing thermal stability. Importantly, observed decay constants of P-series samples were rather comparable with one another (they were all lower than the decay constant of samples without spiking any ZrO₂, (P0)). This is likely attributed to the fact that the plasma treatment produced a relatively homogeneous surface CaZrO₃ phase (overlayer, Fig. 2a). Well-dispersed CaZrO₃ nanoparticles prepared by flame spray pyrolysis have been reported to function as a barrier against sintering to prevent the growth of CaO grain up to 1200 cycles²⁶ and this characteristic likely accounts for the enhanced thermal stability (lower decay constants) of the ZrO₂-stabilized CaO sorbents prepared using the plasma treatment. In contrast, no profound difference in the determined decay constants of F-series samples was noted, indicating that introducing ZrO₂ had no significant influence on enhancing thermal stability when using the furnace treatment. Accordingly, we speculated that the introduced ZrO₂ likely functioned more as wedges separating individual CaO particles. Based on literature and our observations, proposed mechanism accounting for different distribution patterns resulting from plasma- and furnace-treatment is presented in Fig. 6d.

Although the surface CaZrO₃ phase has no direct benefit for CO₂ diffusing into the CaO core, it effectively prevents the collapse of the porous environment during the carbonation/calcination cycles, which is one of the most important properties of an efficient CO₂ sorbent. The benefit of the prevention of the collapse of the porous environment can be clearly realized when we take into the account the overall CO₂ uptake in 20 carbonation/calcinations cycles (Fig. 7). From the first glance in Fig. 7, it is noted that the total CO₂ uptake at the end of 20 carbonation/calcination cycles decreases with increasing the amount of ZrO₂ added. However, considering the total CO₂ uptake per gram of CaO-based sorbent, it is obvious that introduction of ZrO₂ enhances the CO₂ adsorption capacity of CaO sorbents. For instance, sample P5 and F5 exhibit a CO₂ adsorption capacity of approximately 370 g-CO₂ per g-sorbent at the end of 20 cycles, which is equal to 8.41 mol-CO₂ per g-sorbent. This is actually a remarkable performance for a sorbent prepared through waste reutilization, when compared with that obtained through the sol–gel autocombustion method (8.5 mol-CO₂ per g-sorbent).¹³ Importantly, this clearly indicates that a good CO₂ sorbent can be prepared using waste materials, as long as an appropriate treatment is adopted. Furthermore, although the difference in CO₂ uptake capacity between the two treatment methods may look small (i.e., sample P5 and F5, Fig. 7), it is still expected to have essential influence in industrial application in the future, based on the process engineering consideration. This is because the difference in CO₂ uptake capacity will lead to different operation conditions and consequently, the associated operation cost. Apart from the observed capacity, samples with 5 wt% ZrO₂ (P5 and F5) exhibiting the highest CO₂ uptake capacity were selected for the following discussions, focusing in particular on their surface evolution in carbonation/calcination cycles.

Fig. 7 Results of the overall CO₂ uptake (gram of CO₂) in 20 cycles of calcium looping uptake (a) per gram and (b) per mole of sorbent.

Surface evolution of samples with 5 wt% ZrO₂ additives at the end of calcium looping

In order to uncover the mechanism controlling the enhanced CO₂ uptake capacity of ZrO₂-stabilized CO₂ sorbent (focusing on 5 wt% of ZrO₂ additives), their surface evolution based on SEM images is discussed in this section. At the end of thermal treatment, the surface morphologies of P5 (Fig. 8a) and F5 (Fig. 8b) are rather similar, and no visible CaZrO₃ particles are noted. This is because the sizes of the as-formed CaZrO₃ particles lie within the scale of several nanometers, as indicated by the TEM images (Fig. 2). It is generally believed that the malfunction of high temperature CO₂ sorbent is usually due to the collapse of the porous environment during calcination that greatly reduces the surface area and consequently, the number of active carbonation sites. Recent in situ synchrotron radiation X-ray powder diffraction studies suggest that the number of active carbonation sites increases linearly with increasing values of surface area.³¹ The authors identified that these active surface area are related to the nucleation points where calcium carbonate crystallites/product islands initially form, and subsequently grow.³¹ Note that the amount of active surface area is not necessarily proportional to the measured surface area (N₂-BET surface area), as the CaO exposed surface along the preferred crystallographic direction is thermodynamically unfavorable for carbonation.³² According to these studies, retaining the amount of active surface area instead of maintaining the surface area during carbonation/calcination cycles is essential for keeping the reactivity of the CaO sorbents. With the deposition of high refractory nanoparticles, such as CaZrO₃ in this study, the CaO grain growth during calcination is interfered with, and it does not grow along the preferred crystallographic direction. Along with the fact that the stresses in the intermediate state of CaO are indicated to be 20 times higher than its strength as concluded by in situ X-ray diffraction coupled with Rietveld refinement study,³³ the release of stress results in the disintegration of bulk CaO particles and consequently, the generation of nano-sized, crystallite CaO. This mechanism explains the high CO₂ uptake capacity of ZrO₂-stabilized waste oyster shell powders despite their surface areas being slightly reduced during thermal treatment (Fig. 4).

Fig. 8 Surface morphologies of sample (a) P5 and (b) F5 obtained at the end of thermal treatment, and their morphologies at the end of 20 cycles of calcium looping (c) P5 and (d) F5.

Although the evenly distributed surface CaZrO₃ phase would greatly enhance the thermal stability of the CaO-based sorbent, it is noted that the total CO₂ uptake capacity by the plasma-treated sample is lower than that of the furnace-treated sample (Fig. 7), but the former has a finer surface CaZrO₃ phase (Fig. 2). This is because evenly distributed surface CaZrO₃ particles produced by the evaporation–condensation reaction in the plasma treatment hinder the access of CO₂ to CaO and therefore reduce the overall CO₂ capture capacity; accordingly, CaO surfaces that were deposited with CaZrO₃ nanoparticles would remain in their morphology. Instead, sintering occurs on the rest of the CaO sorbents that have no CaZrO₃ coverage, resulting in a fused and interconnected CaO bulk, as shown in the SEM image (Fig. 8c). The balance between the degree of surface cracking during calcination and the extent of sintering during carbonation was suggested to be responsible for changes in uptake during cycling.⁶ Although CaZrO₃ nanoparticles only form at the CaO/ZrO₂ interface in furnace-treated samples, they can function as wedges to reduce the sintering effect, as evidenced by the stacking-flake morphology at the end of 20 cycles of calcium looping (Fig. 8d). This feature allows the access to CO₂ and accounts for the high overall CO₂ capture capacity.

Lastly, it is noted that in general, samples prepared from the furnace treatment have higher CO₂ uptake capacity than those obtained from the plasma treatment. This is an inevitable tradeoff between the thermal stability and CO₂ uptake capacity. In other words, enhancing thermal stability by introducing surface CaZrO₃ nanoparticles would somehow compromise the CO₂ uptake. This is because in essence, this reduces the amount of active carbonation sites on the CaO surface that are responsible for the CO₂ uptake. Also, the surface CaZrO₃ nanoparticles would inhibit the CO₂ molecules diffusing into the CaO core. This emphasizes the important role of the distribution of surface refractory nanoparticles in the CO₂ uptake capacity of CaO-based sorbents. Importantly, our results clearly show that the performance of waste oyster shell powders is only slightly lower than those prepared using purified chemicals through the sol–gel method (8.5 mol-CO₂ per g-sorbent,¹³ estimated to be 7.8 mol-CO₂ per g-sorbent, based on their reported data²⁰), or thermal pyrolysis method (estimated to be 12.0 mol-CO₂ per g-sorbent, based on their reported data²⁶). Our samples further exhibit better thermal resistance than CTAB-assisted synthesized CaO sorbents⁹ and slightly lower CO₂ capture capacity than CaO-molecular sieve-alumina ceramic sorbents.¹⁰ This gives rise to an additional incentive for waste reutilization, as it can simultaneously solve the issue of waste disposal and reduce the environmental impact from mineral exploitation in order to acquire high purity chemicals.

Conclusion

In this study, ZrO₂-stabilized CaO-based sorbents for cyclic high temperature CO₂ capture were prepared using waste oyster shell powders. Two thermal treatment methods, namely atmospheric pressure plasma and conventional furnace treatment were adopted and evaluated. It is noted that atmospheric pressure plasma could produce evenly distributed CaZrO₃ nanoparticles over CaO surfaces, which greatly enhance the thermal stability of the as-obtained sorbents. However, they would also reduce the access of CO₂ to the sorbents and afterwards diffuse into CaO cores. In contrast, CaZrO₃ nanoparticles produced using furnace treatment would function as wedges to reduce the sintering effect, resulting in a stacking-flake look at the end of 20 cycles of calcium looping. This feature allows the access of CO₂ and accounts for its high overall CO₂ capture capacity. Importantly, the overall CO₂ uptake capacity of sorbents prepared using waste oyster shell powders is only slightly lower than those prepared using pure chemicals (sol–gel or pyrolysis method). This provides an opportunity for the reutilization of waste materials to prepare CaO-based sorbents for cyclic high temperature CO₂ capture in a low environmental impact manner.

Acknowledgements

The authors are grateful for Ministry of Science and Technology, Taiwan (MOST 105-2119-M-007-013 and MOST 104-2811-M-007-097).

References

1. S. C. Tian, J. G. Jiang, F. Yan, K. M. Kaimin, X. J. Chen and V. Manovic, Highly efficient CO₂ capture with simultaneous iron and CaO recycling for the iron and steel industry, Green Chem., 2016, 18(14), 4022–4031.
2. T. H. Wang, D.-C. Xiao, C.-H. Huang, Y.-K. Hsieh, C.-S. Tan and C.-F. Wang, CO₂ uptake performance of CaO-based sorbent made from waste oyster shells blended with PMMA nanosphere scaffold and its life cycle assessment, J. Hazard. Mater., 2014, 270, 92–101.
3. C. Chen, S.-T. Yang and W.-S. Ahn, Calcium oxide as high temperature CO₂ sorbent: effect of textural properties, Mater. Lett., 2012, 75, 140–142.
4. Y. Wang, W. Y. Zhang, R. Li, W. B. Duan and B. Liu, Design of Stable Cage-like CaO/CaZrO₃ Hollow Spheres for CO₂ Capture, Energy Fuels, 2016, 30(2), 1248–1255.
5. H. R. Radfarnia and M. C. Iliuta, Development of zirconium-stabilized calcium oxide absorbent for cyclic high-temperature CO₂ capture, Ind. Eng. Chem. Res., 2012, 51(31), 10390–10398.
6. W. Liu, B. Gonzalez, M. T. Dunstan, D. S. Sultan, A. Pavan, C. D. Ling, C. P. Grey and J. S. Dennis, Structural evolution in synthetic, Ca-based sorbents for carbon capture, Chem. Eng. Sci., 2016, 139, 15–26.
7. M. Broda and C. R. Muller, Synthesis of highly efficient, Ca-based, Al₂O₃-stabilized, carbon gel-templated CO₂ sorbents, Adv. Mater., 2012, 24(22), 3059–3064.
8. V. Manovic and E. J. Anthony, Parametric study on the CO₂ capture capacity of CaO-based sorbents in looping cycles, Energy Fuels, 2008, 22(3), 1851–1857.
9. N. N. Hlaing, K. Vignesh, S. Sreekantan, S. Y. Pung, H. Hinode, W. Kurniawan, R. Othman, A. A. Thant, A. R. Mohamed and C. Salim, Effect of Cetyl Trimethyl Ammonium Bromide Concentration on Structure, Morphology and Carbon Dioxide Adsorption Capacity of Calcium Hydroxide Based Sorbents, Appl. Surf. Sci., 2016, 363, 586–592.
10. Z. Sun, C. C. Xu, S. Y. Chen and W. G. Xiang, Improvements of CaO-Based Sorbents for Cyclic CO₂ Capture Using a Wet Mixing Process, Chem. Eng. J., 2016, 286, 320–328.
11. M. Broda, A. M. Kierkowska and C. R. Muller, Influence of the calcination and carbonation conditions on the CO₂ uptake of synthetic Ca-based CO₂ sorbents, Environ. Sci. Technol., 2012, 46(19), 10849–10856.
12. F. D. M. Daud, K. Vignesh, S. Sreekantan and A. R. Mohamed, Improved CO₂ adsorption capacity and cyclic stability of CaO sorbents incorporated with MgO, New J. Chem., 2016, 40(1), 231–237.
13. A. Antzara, E. Heracleous and A. A. Lemonidou, Improving the stability of synthetic CaO-based CO₂ sorbents by structural promoters, Appl. Energy, 2015, 156, 331–343.
14. J. M. Valverde, A. Perejon, S. Medina and L. A. Perez-Maqueda, Thermal decomposition of dolomite under CO₂: insights from TGA and in situ XRD analysis, Phys. Chem. Chem. Phys., 2015, 17(44), 30162–30176.
15. S. Perathoner and G. Centi, A New Scenario for Green & Sustainable Chemical Production, J. Chin. Chem. Soc., 2014, 61(7), 719–730.
16. A. S. Al-Fatesh, M. A. Naeem, W. U. Khan, A. E. Abasaeed and A. H. Fakeeha, Effect of Nano-support and Type of Active Metal on Reforming of CH₄ with CO₂, J. Chin. Chem. Soc., 2014, 61(4), 461–470.
17. L. T. M. Nguyen, H. Park, M. Banu, J. Y. Kim, D. H. Youn, G. Magesh, W. Y. Kim and J. S. Lee, Catalytic CO₂ hydrogenation to formic acid over carbon nanotube-graphene supported PdNi alloy catalysts, RSC Adv., 2015, 5(128), 105560–105566.
18. C. L. Qin, W. Q. Liu, H. An, J. J. Yin and B. Feng, Fabrication of CaO-based sorbents for CO₂ capture by a mixing method, Environ. Sci. Technol., 2012, 44(3), 1932–1939.
19. H. L. Ping and S. F. Wu, Preparation of cage-like nano-CaCO₃ hollow spheres for enhanced CO₂ sorption, RSC Adv., 2015, 5(80), 65052–65057.
20. N. N. Hlaing, S. Sreekantan, R. Othman, S.-Y. Pung, H. Hinode, W. Kurniawan, A. A. Thant, A. R. Mohamed and C. Salime, Sol–gel hydrothermal synthesis of microstructured CaO-based adsorbents for CO₂ capture, RSC Adv., 2015, 5(8), 6051–6060.
21. Y. D. Xia, R. Mokaya, G. S. Walker and Y. Q. Zhu, Superior CO₂ adsorption capacity on N-doped, high-surface-area, microporous carbons templated from zeolite, Adv. Energy Mater., 2011, 1(4), 678–683.
22. W.-J. Liu, H. Jiang, K. Tian, Y.-W. Ding and H.-Q. Yu, Mesoporous carbon stabilized MgO nanoparticles synthesized by pyrolysis of MgCl₂ preloaded waste biomass for highly efficient CO₂ capture, Environ. Sci. Technol., 2013, 47(16), 9397–9403.
23. D. Han and S. Y. Moon, Development of superhydrophobic surface on glass substrate by multi-step atmospheric pressure plasma treatment, Thin Solid Films, 2015, 587, 34–38.
24. V. T. Bui, X. Y. Liu, S. H. Ko and H. S. Choi, Super-amphiphilic surface of nano silica/polyurethane hybrid coated PET film via a plasma treatment, J. Colloid Interface Sci., 2015, 453, 209–215.
25. C. W. Kan, C. H. Kwong and S. P. Ng, Surface modification of polyester synthetic leather with tetramethylsilane by atmospheric pressure plasma, Appl. Surf. Sci., 2015, 346, 270–277.
26. R. Koirala, K. R. Gunugunuri, S. E. Pratsinis and P. G. Smirniotis, Effect of Zirconia Doping on the Structure and Stability of CaO-Based Sorbents for CO₂ Capture during Extended Operating Cycles, J. Phys. Chem. C, 2011, 115(50), 24804–24812.
27. P. Stoch, J. Szczerba, J. Lis, D. Madej and Z. Pędzich, Crystal structure and ab initio calculations of CaZrO₃, J. Eur. Ceram. Soc., 2012, 32, 665–670.
28. S. Jonas, F. Nadachowski and D. Szwagierczak, A new non-silicate refractory of low thermal expansion, Ceram. Int., 1998, 24, 211–216.
29. V. Medvecka, D. Kovacik, A. Zahoranova, M. Stupavska and M. Cernak, Atmospheric pressure plasma assisted calcination of organometallic fibers, Mater. Lett., 2016, 162, 79–82.
30. A. A. El Mel, N. Stephant, J. Hamon, D. Thiry, A. Chauvin, M. Chettab, E. Gautron, S. Konstantinidis, A. Granier and P. Y. Tessier, Creating nanoporosity in silver nanocolumns by direct exposure to radio-frequency air plasma, Nanoscale, 2016, 8(1), 141–148.
31. A. Biasin, C. U. Segre and M. Strumendo, CaCO₃ Crystallite Evolution during CaO Carbonation: Critical Crystallite Size and Rate Constant Measurement by In Situ Synchrotron Radiation X-ray Powder Diffraction, Cryst. Growth Des., 2015, 15(11), 5188–5201.
32. J. M. Valverde, P. E. Sanchez-Jimenez and L. A. Perez-Maqueda, Limestone Calcination Nearby Equilibrium: Kinetics, CaO Crystal Structure, Sintering and Reactivity, J. Phys. Chem. C, 2015, 119(4), 1623–1641.
33. R. Molinder, T. P. Comyn, N. Hondow, J. E. Parker and V. Dupont, In situ X-ray diffraction of CaO based CO₂ sorbents, Energy Environ. Sci., 2012, 5, 8958–8969.

---