Introduction

The fossil fuel burning processes produce greenhouse gases, including CO2, N2, and CH4. Global warming caused by these gases leads apparently to shore floods, hot atmosphere, soil droughts, and damage of eco system. In this scenario, carbon dioxide holds the most significant portion of the flue gases released to the atmosphere [1]. Thus, extensive efforts have been made by scientists, institutions, countries, and environmental organizations to reduce the CO2 emissions. The principal source of CO2 is a combustion processes that used fossil fuel. However, the utilization of fossil fuel is still very for satisfying energy demands. Hence, the feasible solution to continue using fossil fuel with mitigation of climate change is Carbon Capture and Storage (CSS).

A massive number of researchers have already studied CO2 capture using both experimental and simulation approaches as well as synthesizing novel adsorbents [2]. Using adsorption for CO2 separation is advantageous by the ease of regenerating the adsorbent by being exposed to heat and/or vacuum [3]. The most common known adsorbents are activated carbons and zeolites being exploited for CO2 separation and storage. Zeolites could adsorb a higher quantity of CO2 than does by activated carbon at low operating pressures (< 20 kPa) [4, 5], whereas carbon-based adsorbents are better for CO2 storage applications [5, 6]. Conversely, the obvious merits of carbon adsorbents over zeolites are cost penalties, hydrostability, lower regeneration energy, and easiness of production on a commercial scale [7]. Despite, zeolite-based adsorbents have the relatively higher CO2 adsorption capacity, especially at lower adsorption pressures (10–30 kPa (abs.) at T = 30 °C), the CO2 uptake is greatly reduced in case of CO2/H2O mixture and requires significantly higher heat of regeneration [8, 9].

Since 20 years ago, a novel class of reticulate adsorbents has been discovered and called metal organic frameworks (MOFs) [10, 11]. In this context, the highest adsorption CO2 adsorption capacity (1470 mg g−1) was reported for MOF-177 at 35 bar [12]. Over time, a large number of MOFs have been developed by scientists for maximize the CO2 uptake and selectivity.

A certain class of MOFs incorporating functionalized and open metal sites has shown high separation efficiency at ambient conditions like HKUST-1, Mg-MOF-74, and NH2MIL-53 (Al) [13]. With regard to maximum CO2 adsorption capacity uptake, a few MOFs have shown a reasonably high CO2 uptake, such as CPM-5, MIL-53 (Al), UMCM-150, and Ni-STA-12, while others have been evaluated for comparatively lower uptakes like MOF-5 and MOF-177 [13]. A nickel-based MOF, Ni/DOBDC, was also investigated for CO2 capture and was measured to exhibit a fairly high adsorption capacity [14]. A nickel isonicotinate-based ultramicroporous MOF, 1 [Ni-(4PyC)2·DMF], showed lower parasitic energy than Mg-MOF-74 and good stability with humidity [15].

MOF-74 is the most currently available MOFs with high CO2 adsorption uptake as well as an excellent CO2 over N2 selectivity [16, 17]. More specifically, Mg-MOF-74 was quantitatively identified as the highest adsorbent adsorbs CO2 at low-pressure conditions (350 mg/g at 298 K) [17]. It is also reported that Mg-MOF-74 showed higher H2O hydrophilicity (593 ml/g at 298 K) than zeolite [18]. Despite the high CO2 uptake of Mg-MOF-74, the existence of H2O reduced the CO2 capture capacity, unlike HKUST-1, MIL-101(Cr), and MIL-100 (Fe) types [19]. The study [19], moreover, showed the reduction of CO2 adsorption at different conditions. For example, at 1 bar and 298 K, the CO2 adsorption capacity of the dry Mg-MOF-74 was about 8.4 mmol/g of CO2, while with hydration 6.5 and 13%, the CO2 adsorption capacity values were 6.7 mmol/g and 5.4 mmol/g, respectively [19]. MIL-100(Fe) is remarked as the hydro- and thermal-stable adsorbent. It showed that some increase in CO2 uptake recorded with increasing RH, (up to 105 mg g−1 for CO2 at 40% RH), with a large decrease in adsorption heat [20].

Less than 2 decades, carbon nanotubes (CNTs) have confirmed a capability towards CO2 separation [21,22,23,24,25]. Specifically, the functionalized CNTs by amine-groups showed a significant improvement in the CO2 uptake [26,27,28,29]. The CO2 uptake was reported around 2.59 mmol/g at 293 K using APTES functionalized CNTs. Incorporating CNTs with adsorbents could considerably improve the thermal diffusion of the adsorbents, thereby enhancing the CO2 adsorption uptake [30,31,32]. For example, the compound of CNT@13X/CaCl2 exhibited higher thermal conductivity and then more CO2 uptake than did by 13X/CaCl2 and pristine 13X [30, 31]. Furthermore, the thermal stability along with the adsorption uptake was improved by adding CNT to MIL-101-68 (Al) [32]. Incorporating CNT and lithium ions with MOF Cu3(btc)2 resulted in improving the CO2 capacity by about 305% compared with those of the base adsorbent (Cu3(btc)2) [33]. In addition, the adsorption capacity of CO2 at high pressure and room temperature (10 bars and 298 K) was improved by adding MWCNT to MIL101 [34].

A considerable number of experimental and numerical research attempts based on CO2 capture and separation have been conducted so far in terms of breakthrough and pressure and temperature swing adsorption [35,36,37,38,39,40,41,42,43,44,45,46,47,48,49]. Nevertheless, the majority amount of research chemically concentrated on development of novel adsorbent materials targeted for acquiring high CO2 adsorption and selectivity. At this point, the poor thermal conductivity shown by a vast majority of these adsorbents has been experienced as a major barrier in improving the CO2 uptake of these materials in real applications. In addition, a very limited number of research attempts have conducted the enhancement of CO2 capture and storage via improving the thermal properties of the base adsorbent. For this reason, this paper aims at investigating the effects of incorporating multi-walled carbon nanotubes (MWCNTs) with Mg-MOF-74 and MIL-100(Fe), with the aim of enhancing the thermal properties of the adsorbents, and investigating the influence of the MWCNT addition on the CO2 adsorption capacity and breakpoint of the resulting MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe) composites. The synthesized and activated Mg-MOF-74 and MIL-100(Fe) have been physically incorporated with different percentages of MWCNTs. They are characterized for the degree of crystallinity, CO2 adsorption isotherms, and CO2 adsorption breakthrough. Moreover, the dynamic separations of breakthrough tests are carried out to address the actual CO2 separation from CO2/N2 (20% v/v. CO2 and 80% v/v. N2 for MWCNT/Mg-MOF-74, and 15% v/v. CO2 and 85% v/v. N2 for MWCNT/MIL100(Fe)) mixture and then computing the level of enhancements on CO2 uptake and separation.

Experimental work methodology

MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe) samples preparation

We have followed a successful procedure for synthesizing Mg-MOF-74 as described in [50]. Briefly, 0.337 g 2,5-dihydroxyterephthalic acid and 1.4 g Mg(NO3)2·6H2O were dissolved in a solution of 135 ml dimethylformamide, 9 ml ethanol, and 9 ml water with sonication for 10 min. The resulting stock solution was decanted into twelve 50 ml bottles. The bottles were tightly capped and heated at 398 K for 26 h. The mother liquor was, then, decanted. Following this, the products were washed with methanol, and then, left immersed in methanol. The products were combined to one bottle and exchanged into fresh methanol daily for 4 days. The activation process was carried out by evacuating the product to dryness and then heated under vacuum at 523 K for 6 h.

The synthesis of MIL-100(Fe) was performed in accordance with a previously reported procedure [51]. We first dissolved Fe(NO3)3·9H2O (4.04 g, 0.01 mol) in de-ionized water (50.2 ml, 2.8 mol) and the mixture was completely put in a 125 ml Teflon-liner containing BTC (1.4097 g, 0.00671 mol). After that, the Teflon-liner was tightly sealed inside a stainless steel autoclave and was kept at 383 K for 14 h. After heating, the autoclave was slowly cooled to ambient temperature, after which the “as-synthesized” dark orange solid was recovered using a centrifuge that was operated at 8000 rpm for about 45 min. The as-synthesized MIL-100(Fe) was washed with copious amounts of water and ethanol and finally with an aqueous NH4F solution in purpose removing any unreacted species. Specifically, the dried solid was, first, immersed in deionized water (60 ml per 1 g of solid) and the resulting suspension was stirred at 70 °C for 5 h. Again, the suspension was centrifuged and the wash process was repeated using ethanol (60 ml) at 65 °C for 3 h. This two-step purification was repeated until the decanted solvent following centrifugation became completely colorless, after which the solid was immersed in a 700 ml aqueous NH4F solution and stirred at 70 °C for 5 h. The suspension was again centrifuged and the solid was washed 5 times DI water at 60 °C, and finally dried in air at 75 °C for 2 days followed by 95 °C for 2 days.

The MWCNT/Mg-MOF-74 sample designations were based according to the weight percentage of powder form of MWCNTs which was physically mixed with as-synthesis powder form of Mg-MOF-74 as follows. Mg-MOF-74, 0.1 wt% MWCNT/Mg-MOF-74, 0.25 wt% MWCNT/Mg-MOF-74, 0.5 wt% MWCNT/Mg-MOF-74, 0.75 wt% MWCNT/Mg-MOF-74, 1 wt% MWCNT/Mg-MOF-74, and 1.5 wt% MWCNT/Mg-MOF-74 and shortly named as Mg-MOF-74, MFC1, MFC2, MFC3, MFC4, MFC5, and MFC6, respectively. The incorporation of powder form of MWCNT with as-synthesis powder form of MIL-100(Fe) has produced MIL-100(Fe), 0.1 wt% MWCNT/MIL-100(Fe), 0.25 wt% MWCNT/MIL-100(Fe), and 0.5 wt% MWCNT/MIL-100(Fe) which were named as MIL-100(Fe), MMC1, MMC2, and MMC3, respectively. The particle size distribution for selective composites has been conducted using Particles-Size Analyzer, Model S3500, Microtrac, USA.

Powder X-ray diffraction (PXRD) analysis

To determine the crystallinity of composites, PXRD patterns of MWCNT/Mg-MOF-74 were collected using a Bruker D8-Advance (Cu Kα λ = 1.54056 Å). The operating power of the PXRD system was 30 kV/30 mA and the step-counting method (step = 0.02o, time = 3 s) was used to collect data at range 2θ = 3–45o and 298 K. For MIL-100(Fe); the diffraction data were collected between 3 and 45° (2θ) with a total scan time of 3 h.

Scanning electron microscopy (SEM)

Scanning electron microscopy (SEM) was carried out using a TESCAN LYRA3 FEG microscope to test structure of Mg-MOF-74, 1.5 wt% MWCNT/Mg-MOF-74, MIL-100(Fe) and 0.5 wt% MWCNT/MIL-100(Fe). SEM samples were prepared by placing a powder form of samples on Al tapes and sputter coated with gold. The images were obtained at voltage of 20 kV.

Gas physisorption measurements

The first step in the physisorption measurements of CO2 and N2 is the sample degassing to remove any guest molecules within the pores of each material. Typically, 50–200 mg of each sample was transferred to pre-weighed empty sample cell with a 9 mm diameter. Degassing was conducted at 150 °C under vacuum for about 17 h for MWCNT/MIL-100(Fe) and 220 °C under vacuum during about 5 h for MWCNT/Mg-MOF-74 using an Autosorb degasser (Quantachrome Instruments, Inc.). Nitrogen adsorption isotherms at 77 K were first recorded to estimate the Brunauer–Emmett–Teller (BET)-specific surface area (SBET), average pore radius, and total pore volume. The interesting equilibrium adsorption isotherms for CO2 at 273, 298, and 313 K, and for N2 at ambient temperature (298 K) were recorded. The CO2 heat of adsorption was evaluated using the adsorption isotherms measured at 273 and 313 K in accordance with the Clausius–Clapeyron equation.

Breakthrough experiments of binary gas mixture (CO2 + N2)

A dynamic CO2/N2 breakthrough setup was constructed to separate CO2 from a CO2/N2 mixture (representing a flue gas), Fig. 1. The home-made setup is composed of a bed column with specifications of Inner diameter = 4 mm, outer diameter = 6 mm, and length = 7 cm. The column was filled with the MWCNT/Mg-MOF-74 composite (about 0.26 g), or the MWCNT/MIL-100(Fe) composite (about 0.74 g). The system includes CO2 and N2 cylinders, two MFC (calibrated for CO2 and N2 flow rates), two check valves, and bypass tube (for calibrating the gases concentration detected by mass spectrometer from the inlet gas compositions). It, moreover, includes two bourdon absolute pressure (manufactured by Baumer, accuracy ± 1.6%), mass spectrometer (to measure the outlet gases composition leaving from the bed), vacuum pump and electric heater jacket (for regeneration purpose), and interconnecting stainless steel fittings and tubes to regulate the flow of carrier gas within the system.

Fig. 1
figure 1

Scheme of CO2/N2 adsorption separation breakthrough setup

Full size image

The first step in the operation of breakthrough setup has involved the degassing of the activated MWCNT/MIL-100(Fe) or MWCNT/Mg-MOF-74 composite sample at 423 K under vacuum for 1 day period to remove any guest gases inside the pores of MWCNT/MIL-100(Fe) or MWCNT/Mg-MOF-74 frameworks. The breakthrough experiments were conducted at 297 K and 101.3 kPa. The flowrate of the feed gas, a mixture of 20% CO2 and 80% N2 (v/v) for MWCNT/Mg-MOF-74 samples and 15% CO2 and 85% N2 (v/v) for MWCNT/MIL-100(Fe) samples, was kept constant at 20 and 10 sccm for MWCNT/Mg-MOF-74 and MIL-MWCNT/100(Fe) composites, respectively. The full breakthrough capacity of CO2 and N2 was measured by evaluating the ratio of compositions of the downstream gas and the feed gas. The CO2 adsorption capacity of the adsorbents is evaluated from the CO2 molar flowrate inters and leaves the bed using the expression [42]:

$$q_{{{\text{co}}_{ 2} }} = \frac{1}{m} \left[ {\mathop \int \limits_{0}^{t} \left( {Q_{\text{F}} C_{0 } - C\left( t \right)Q\left( t \right)} \right) {\text{d}}t - \varepsilon V{\text{C}}_{0 } } \right],$$
(1)

where \(q_{{co_{2} }}\) (mmo/g) represents the CO2 uptake, QF and Q (t) (m3/s) are the input and output volumetric flow rates, C0 and C(t) (mol/m3) are the influent and effluent CO2 concentrations, t (s) is the time, ε is the bed porosity, and V (m3) is the bed volume.

The term (\(\varepsilon V{\text{C}}_{0 } )\) in Eq. 1, which represents the CO2 amounts still in the bed void (without being adsorbed), has very small values comparing to the other terms, so that, it can be ignored. However, we have considered it in our calculations.

Results and discussion

Powder X-ray diffraction (PXRD) analysis

Figure 2 shows the PXRD patterns of MWCNT/Mg-MOF-74 compounds as well as MIL-100(Fe). It can be seen (Fig. 2a) that the PXRD pattern of MWCNT/Mg-MOF-74 and Mg-MOF-74 samples is in good agreement with the simulated pattern. The incorporation of MWCNTs has not decreased the crystallinity of the framework, as all the intensity peaks locations have represented the Mg-MOF-74 structure. Hence, it can be concluded that the incorporation of less than 1.5 wt% MWCNTs using physical mixing preserves the characteristic lattice structure of the Mg-MOF-74 framework. The same conclusion was drawn for MWCNT/MIL-100(Fe) as reported in the recent work [51]. Figure 2b is to exhibit that the synthesized MIL-100(Fe) was in good agreement with the simulated pattern.

Fig. 2
figure 2

PXRD patterns for a MWCNT/Mg-MOF-74 and b MIL-100(Fe)

Full size image

Electron microscopy analysis

The morphologies of Mg-MOF-74, MIL-100(Fe), 1.5% wt MWCNT/Mg-MOF-74, and 0.5 wt% MWCNT/MIL-100(Fe) composites observed by SEM are shown in Fig. 3. The physical incorporation of MWCNTs does not change the shape of the Mg-MOF-74 and MIL-100(Fe) crystals. The crystal size of Mg-MOF-74 is much bigger than that of MIL-100(Fe).

Fig. 3
figure 3

SEM micrographs of a Mg-MOF-74, b MIL-100(Fe), c 1.5 wt% MWCNT/Mg-MOF-74, and d 0.5 wt% MWCNT/MIL-100(Fe)

Full size image

Adsorption equilibrium isotherms of carbon dioxide and nitrogen

The N2 equilibrium isotherms for the MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe) composites have been measured at 77 K. Table 1 lists the important porosity-related parameters estimated from the N2 adsorption data MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe) composites. The measured BET surface area was almost close to each other in MWCNT/Mg-MOF-74 compounds between 1470 and 1590 m2/g. In addition, the total pore volume measured at 95% relative pressure (P/P0) and the pore size measured were determined to be almost the same for all the samples by around 0.63–0.71 cc/g and 19 Å, respectively. The Mg-MOF-74 BET surface area and total pore volume values are in good agreement with those reported in the literature [17, 50, 52]. It can be deduced from the data shown in Table 1 that the addition of MWCNTs does not result in substantial differences concerning its influence on the porosity-related parameters evaluated for the MWCNT/Mg-MOF-74 compounds. Similarly, the incorporation of MWCNTs with MIL-100(Fe) did not change the porosity-related parameters of pristine MIL-100(Fe). There was almost slight increase in the surface area from 1083 m2g−1 for base MIL-100(Fe) to 1464 m2g−1 for MMC2. The total pore volume at 0.95 relative pressure was around 0.61 and 0.69 cc/g for MMC1 and MMC2, respectively, in comparison to 0.55 cc/g for the pristine MIL-100(Fe). Regarding the pore size, it is clearly that the pore diameter for all the MIL-100(Fe) and composite samples was around 20 Å. These porous-property values are close to those reported for MWCNT/MIL-100(Fe) composites [51]. However, the synthesized MIL-100(Fe) in the present work is not the best qualitative fashion of this adsorbent due to the method of synthesis and purification we have followed. BET and pore volume of MIL-100(Fe) could be varied between 1090 and 2050 m2g−1 (for BET) and from 0.65 to 1.15 cc/g (for average pore volume) [53].

Table 1 Pores characterization of the MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe) composites for N2 at 77 K
Full size table

The CO2 adsorption isotherms for MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe) composites, measured at 273, 298, and 313 K, are exhibited in Figs. 4 and 5. It is obvious that the adsorption uptake increased sharply in the region below 15 kPa and increased gradually with increasing adsorption pressure greater than 20 kPa. This behavior gives a good advantage for CO2 capturing in low-pressure applications including the CO2 separation from the flue gas (PCO2 = 10–20 kPa). However, as expected, an increase in the measured temperature showed an adverse effect on the recorded uptakes for each material. As obvious from Fig. 4, the highest CO2 uptake has been measured for pristine Mg-MOF-74 followed by MFC1 at all the measured temperatures (273, 298, and 313 K). For MWCNT/MIL-100(Fe) compounds (Fig. 5), the adsorption uptake increased more or less linearly with increasing adsorption pressure. It is obvious that MMC2 showed optimal adsorbed amounts, and MMC1 resulted in the second highest uptake even greater than the pristine MIL-100(Fe) and MMC3 composites, as shown in Fig. 5a–c. It is worth mentioning here that the CO2 uptake for MWCNT/Mg-MOF-74 composites was much higher than that adsorbed by MWCNT/MIL-100(Fe) compounds. Another point is that the best version of MIL-100(Fe) could adsorb about 1 and 1.52 mmol/g at 0.6 bar for outgassing pretreatments 150 and 250 °C, respectively [54]. Consequently, the reduction of CO2 capacity of the present MIL-100(Fe) (q = 0.6 mmol/g at 0.6 bar and 298 K) is due to both the followed synthesis method (using Fe(NO3)3·9H2O) and the activation treatments (75–95 °C).

Fig. 4
figure 4

CO2 adsorption isotherms of MWCNT/Mg-MOF-74 composites at: a 273 K, b 298 K, and c 313 K

Full size image
Fig. 5
figure 5

CO2 adsorption isotherms of MWCNT/MIL-100(Fe) composites at: a 273 K, b 298 K, and c 313 K

Full size image

The N2 adsorption isotherms for MWCNT/Mg-MOF-74 composites, measured at 298 K, are displayed in Fig. 6a. It is evident that the pristine Mg-MOF-74 exhibited the largest uptake amount, followed by MFC4, MFC1, MFC6, MFC2, MFC5, and MFC3, respectively. Figure 6b shows N2 uptake by MWCNT/MIL-100(Fe) compounds. It is clear that MMC1 had higher adsorbed values in comparison to the other compounds. For both adsorbent composites [MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe)], the maximum uptake measured for N2 was observed to be significantly smaller than that measured earlier for CO2. In other words, all the samples have been noticed to exhibit preferential selectivity of CO2 over the N2.

Fig. 6
figure 6

N2 adsorption isotherms at 298 K for a MWCNT/Mg-MOF-74 and b MWCNT/MIL-100(Fe) composites

Full size image

To represent isotherms in mathematical models, Toth fitting (Eq. 2) is satisfied. For instance, the CO2 and N2 values obtained using Toth fitting of MFC6 and MMC1 show an excellent agreement with those of the experimental isotherms as plotted in Fig. 7. The respective fitting parameters at 95% level of confidence are tabulated in Table 2:

$$q_{i} = \frac{{q_{m,i} K_{{{\text{eq}},i}} P_{i} }}{{\left[ {1 + \left( {K_{{{\text{eq}},i}} P_{i} } \right)^{ni} } \right]^{1/ni} }}.$$
(2)
Fig. 7
figure 7

Equilibrium isotherms and Toth fittings for MFC6 and MMC1 at 298 K. a MFC6 and b MMC1

Full size image
Table 2 Toth fitting parameters from experimental CO2 and N2 isotherms for MFC6 and MMC1 at 95% level of confidence
Full size table

Here, q is the equilibrium adsorption amount (mmol/g) of species i. q m , Keq, and n are the Toth fitting constants.

Figure 8a, b depicts the variation of heat of adsorption for CO2, Qst, against the instantaneous CO2 uptake for MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe) composites. For MWCNT/Mg-MOF-74, the Qst_CO2 values were observed to exhibit a more or less curvilinear correlation with the instantaneous CO2 uptake, as shown in Fig. 8a. In general, Q st _CO2 values locate between 25 and 40 kJ/mol; the high values were sequentially associated with Mg-MOF-74, MFC4, MFC2, and MFC5. In contrast, MFC5 and MCF3 showed lower values of CO2 adsorption heat. Mg-MOF-74 heat of adsorption profile is close to that reported by Simmons et a. [55].The MWCNT/MIL-100(Fe) composites have been measured to have CO2 adsorption heat values between 18 and 28 kJ/mol [for q > 0.3 mmol/g, Fig. 8b]; the highest values were measured for MIL-100(Fe) about 27.3 kJ/mol (for q > 0.3 mmol/g) and the lowest obtained by MMC1 about 18 kJ/mol (for q > 0.3 mmol/g). The CO2 adsorption heat of MIL-100(Fe) show a close values with those of the same adsorbent reported by Mei et al. [56].

Fig. 8
figure 8

CO2 heat of adsorption (Qst) a MWCNT/Mg-MOF-74 and b MWCNT/MIL-100(Fe) composites

Full size image

Experimental adsorption breakthrough test for MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe) composites

Breakthrough experiments have been performed for the binary gas (CO2/N2) to quantify the improvements in CO2 adsorption uptake and breakpoint as a result of the incorporation MWCNTs inside Mg-MOF-74 and MIL-100(Fe). In a typical procedure, predetermined amounts of MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe) composite samples were first transferred to a tube (length L = 7 cm, inner diameter ∅ = 4 mm). All breakthrough experiments have been performed at ambient temperature of 297 K.

Particle size used in breakthrough was measured for some selective composites including pure materials and the maximum MWCNTs contents composites, as shown in Fig. 9. Mg-MOF-74 and MFC6 have almost close particle size distribution where the particle size values allocated between 1 and 88 µm. It is also shown that percentage of large sizes have been decreased by adding 1.5 wt% MWCNT. For MWCNT/MIL-100(Fe) particle size distribution, it is obvious from Fig. 9b that the both tested samples (MIL-100(Fe) and MMC3) have similar distribution (between 1 and 1400 µm) with almost the same percentages. This is attributed to the small percentages of MWCNT added to the adsorbents.

Fig. 9
figure 9

Particle size distribution (microns) for a Mg-MOF-74 and MFC6, and b MIL-100(Fe) and MMC3

Full size image

For systematic tests, the pressure drop has been diminished to a round zero as monitored by two bourdon meters at the inlet and outlet of the bed. The adsorbent bed was packed with almost the same packing density by about 0.292 ± 0.005 and 0.842 ± 0.002 g/cc for MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe) compounds, respectively. The samples have been pre-treated by heating process for 20 h at about 423 K under vacuum. The experimentally measured CO2 and N2 adsorption breakthrough curves for MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe) composites are displayed in Fig. 10. As evident, the outlet concentration ratios calculated each of these two gases have been plotted against the measurement time. In general, it was observed in all the tested samples that the concentration ratio evaluated for CO2 at the bed outlet remained constant at zero for some time (e.g., about 6–7 min for MWCNT/Mg-MOF-74 and 2–3 min for MWCNT/MIL-100(Fe) compounds), whereas the concentration ratio for N2 increased up to about 1.3 (almost molar fraction = 1) owing to the absence of CO2 which was pre-adsorbed into the Mg-MOF-74 or MIL-100(Fe) composite adsorbent bed. Following the first adsorption minutes of measurement time, the CO2 concentration ratio was observed to increase up to 1, whereas the concentration ratio of N2 was evaluated to gradually drop to a value close to 1. For MWCNT/Mg-MOF-74 composites, the optimal value of the breakpoint, a time at which the concentration ratio of the bed outlet has been evaluated to be less than 5%, was measured to be about 8.16 min (28.4 min/g) for MFC6 against 7.5 min (27.67 min/g) for Mg-MOF-74. This was followed by the value measured for MFC4 of about 8.1 min, and then by 7.96 min for MFC1 (Fig. 10a). In the same manner, the highest breakthrough breakpoint obtained by MWCNT\MIL-100(Fe) was associated with MMC2 by about 3.21 min (4.33 min/g) (Fig. 10b). The next breakthrough point was obtained by MMC1 at about 3.19 min (4.32 min/g), and, then, by pristine MIL-100(Fe) at about 2.9 min (3.69 min/g).

Fig. 10
figure 10

Breakthrough curves for a MWCNT/Mg-MOF-74 composites for CO2/N2 (0.2/0.8 v/v) and b MWCNT/MIL-100(Fe) composites for CO2/N2 (0.15/0.85 v/v), measured at 297 K and 101.3 kPa

Full size image

Adsorption breakthrough and separation processes can be also investigated numerically as described in our previous works [57,58,59]. In addition, adsorption breakthrough curves, can be analytically represented by fitting the experimental curves using some approaches reported in the literature; one of these approaches is expressed in the following equation [60, 61],

$$\frac{{C\_{\text{out}}}}{{C\_{\text{in}}}} = \frac{1}{2} {\text{erfc}}\left( {\sqrt \xi - \sqrt \tau - \frac{\sqrt \xi }{8} - \frac{\sqrt \tau }{8}} \right),$$
(3)

where

$$\xi = k K \frac{L}{v} \left( {\frac{1 - \varepsilon }{\varepsilon }} \right),$$
(4)
$$\tau = k \left( {t - \frac{L}{v}} \right).$$
(5)

Here, k an K are fitting constant; k (1/s), called adsorption time constant, could be used to determine the diffusion coefficient (\(k = \frac{15 D}{{r^{2} }}\), D (m2/s) is the diffusion coefficient and r (m) is the adsorbent particle radius), L (m) is the bed length, and v (m/s) is the flow velocity. The breakthrough time (t) is taken in seconds.

For example, Fig. 11 shows the analytical and experimental adsorption breakthrough of carbon dioxide adsorbed by MWCNT/Mg-MOF-74 composites. The analytical results using Eq. (3) provide a good indication to the whole adsorption time and behavior that are as close as to those of the experimental curves. The time constant values (k) obtained from analytical curves are 0.258, 0.248, 0.164, 0.392, 0.433, 0.368, and 0.255 1/s for Mg-MOF-74, MFC1, MFC2, MFC3, MFC4, MFC5, and MFC6, respectively.

Fig. 11
figure 11

Analytical representation of CO2 breakthrough curves of MWCNT/Mg-MOF-74 composites; solid lines are analytical results; symbols are experimental data

Full size image

To evaluate the improvements of CO2 adsorption capacity and breakpoint by adding MWCNT to Mg-MOF-74 and to MIL-100(Fe), the adsorbed amounts of CO2 have been calculated from the experimental breakthrough curves using Eq. 1. The maximum CO2 adsorption capacity for base Mg-MOF-74 and MIL-100(Fe) calculated from the respective breakthrough curve was estimated to be about 5.46 and 0.37 mmol/g, respectively. The maximum CO2 uptakes along with the adsorption breakpoint ratios for Mg-MOF-74 as well as each of the six MWCNT/Mg-MOF-74 composites are displayed in Fig. 12a. As evident, each of the six composites, except MFC2 and MFC3, exhibited a good improvement over pristine Mg-MOF-74 with regard to both the adsorption capacity and the adsorption breakpoint ratio values. More specifically, the most optimum combination of adsorption capacity and breakpoint ratio values have been evaluated for MFC6 which has shown an improvement of 7.35 and 8.03% over pristine Mg-MOF-74 for adsorption capacity and breakpoint ratio, respectively. This pair of statistics was followed by MFC1, MFC4, and MFC5 composites for which the corresponding improvements in adsorption capacity and breakpoint values over pristine Mg-MOF-74 have been evaluated to be 4.43 and 5.71, 2.21 and 7.3%, and 1.49 and 4.98%, respectively. It is worth mentioning here that each of the MFC6, MFC1, and MFC4 composites has already been characterized for lower values of heat of adsorption for CO2 in comparison with pristine Mg-MOF-74, as shown earlier in Fig. 8a. This, theoretically, implies that each of these composites should not only exhibit higher CO2 uptake values than pristine Mg-MOF-74, but also require comparatively lower energy for regeneration process (recycling recovery).

Fig. 12
figure 12

CO2 uptake (cubic bars) and adsorption breakpoint (cylindrical bars) enhancements for a MWCNT/Mg-MOF-74 and b MWCNT/MIL-100(Fe) composites measured at 297 K

Full size image

Figure 12b shows the improvement in both adsorption capacity and breakpoint due to adding MWCNT to the pristine MIL-100(Fe). As evident, MMC1 exhibited an optimal improvement reaches 12.02 and 9.21% for CO2 adsorption capacity and breakpoint, respectively. This improvement was followed by MMC2 for measured adsorption uptake and breakpoint about 8.74 and 9.47%, respectively, comparing with the base adsorbent (MIL-100(Fe)). On the contrary, the evaluated adsorption uptake and breakpoint improvement values for MFC2, MFC3, and MMC3 showed lower performance than the base adsorbents. This attribute indicates that there is no a uniform improvement can be obtained for the incorporation of CNT with MOFs. The detected improvement in the CO2 adsorption capacity and breakpoint primarily refers to an improvement in the thermal properties of Mg-MOF-74 and MIL-100(Fe) frameworks upon the incorporation of MWCNTs [30,31,32]. As the thermal conductivity of CNT is significantly high (2000–5000 W m−1K−1) [62], the MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe) composites’ effective thermal conductivity values could accordingly be higher than those of pristine adsorbents (0.2–0.4 W m−1K−1). Therefore, the heat diffusion across the bulk composites is enhanced during adsorption processes, which helps in cooling down the adsorbent and enhances CO2 adsorption uptake. Furthermore, the enhancement of effective thermal conductivity of adsorbents helps in quickly heating the adsorbent particles during and desorption process which in turns accelerating the CO2 evacuation from the adsorbent.

This research confirms a comparative dynamic CO2 uptake compared to the published data shown in Table 3. The MIL-100(Fe) shows the lowest values of adsorption in a comparison to AC, 13X and Mg-MOF-74, while Mg-MOf-74 and MFC6 have the highest CO2 uptake. Mg-MOF-74 dynamic CO2 uptake was about 5.46 mmol/g which is greater than 4.06 mmol/g reported in the literature [17, 50], because it was, in this study, measured at 20% CO2 molar fraction. The cost of adding very low quantity of MWCNT (< 1.5 wt%) to the adsorbents is believed to be neglected in a comparison to the CO2 separation improvements.

Table 3 Dynamic CO2 uptake of adsorbents
Full size table

In the literature, chemists usually use the adsorption isotherm data to compare the CO2 capacities of different adsorbents. However, we found out by carrying both adsorption isotherm measurements and adsorption breakthrough experiments that they can give different ratings of adsorption capacity. Keeping in mind that adsorption isotherm measurements are taken under constant temperatures, while the breakthrough measurements are not, as the breakthrough bed is allowed to vary its temperature due to the heat dissipation from the adsorbent to the ambient or surrounding environments. The improved thermal diffusion cools down the adsorbent quickly. Therefore, the cooler is adsorbent, the higher is CO2 uptake which is also confirmed in the isotherms. The most accurate adsorption capacity if we are joining to use a PSA/VSA/TSA is that measured in a breakthrough setup.

Conclusions

Two types of MOFs, Mg-MOF-74 and MIL-100(Fe), were synthesized and incorporated with MWCNTs. In total, seven compounds of Mg-MOF-74 materials containing 0, 0.1, 0.25, 0.5, 0.75, 1, and 1.5 wt% MWCNTs and four compounds of MIL-100(Fe) involving 0, 0.1, 0.25, and 0.5 wt% MWCNT have been characterized for the degree of crystallinity, intrinsic porosity, CO2 adsorption capacity and separation, and dynamic adsorption breakthrough tests. The powder X-ray diffraction patterns as well as the porosity-related parameters for each of the composites did not include any substantial variation in peak intensities and peak locations, BET surface area, and pore volume and size, indicating that the crystal lattices of Mg-MOF-74 and MIL-100(Fe) were unaffected by the incorporation of MWCNTs using the physical mixing (up to 1.5 wt% MWCNT for Mg-MOF-74 and 0.5 wt% MWCNT for MIL-100(Fe).

Equilibrium adsorption isotherms of CO2 measured at 273, 298, and 313 K, and N2 adsorption isotherms measured at 298 K confirm that the highest adsorption capacities for each of these two gases are exhibited by Mg-MOF-74 and 0.25 wt% MWCNT/MIL-100(Fe) (MMC2). Overall, the MWCNT/Mg-MOF-74 composites have much larger adsorption uptake values than those of MWCNT/MIL-100(Fe) composites.

The key performance evaluation of the MWCNT/Mg-MOF-74 and MWCNT/MIL-100(Fe) composites has been achieved through the measurement of actual time-variant CO2 breakthrough curves, which have revealed a good improvement in CO2 adsorption capacity as well as adsorption breakpoint due to the incorporation of MWCNTs in the Mg-MOF-74 and MIL-100(Fe) frameworks. The most optimum combination of these characteristics has been observed for an incorporation of 1.5 wt% MWCNTs in Mg-MOF-74, MFC6, which has adsorption capacity of 5.86 mmol/g with corresponding heat of adsorption about 28.27 kJ/mol against 5.46 mmol/g and 30.02 kJ/mol for pure Mg-MOF-74. The estimated improvements of CO2 adsorption capacity and breakpoint, obtained from breakthrough using 1.5 wt% MWCNTs/Mg-MOF-74, were about 7.35 and 8.03% over pristine Mg-MOF-74. The incorporation of 0.1 wt% MWCNTs in MIL-100(Fe), MMC1, moreover, improves the adsorption capacity to about 0.414 mmol/g with low corresponding heat of adsorption about 18.34 kJ/mol in comparison to 0.370 mmol/g and 27.20 kJ/mol for MIL-100(Fe). The adsorption uptake and breakpoint enhancements of 0.1 MWCNT/MIL-101(Fe) over pristine MIL-100(Fe) were about 12.02 and 9.21%, respectively.