Zean Wang,
Hao Liu*,
Kang Zhou,
Peifang Fu,
Hancai Zeng and
Jianrong Qiu
State Key Laboratory of Coal Combustion, Huazhong University of Science and Technology, Wuhan 430074, China. E-mail: liuhao@hust.edu.cn
First published on 7th October 2015
Activated carbon fibers (ACFs) can effectively remove pollutants including nitrogen oxides, sulfur oxides and trace metals due to their rich micropores and large specific surface area. This work aims to investigate the roles of surface carbon–oxygen and nitrogen–oxygen species on the removal of hydrogen chloride (HCl) using viscose-based ACFs. To evaluate the effect of surface carbon–oxygen and nitrogen–oxygen groups, commercial viscose-based ACFs were treated by thermal treatment (900 °C) and chemical impregnation using H2O2, HNO3 and Cu(NO3)2. Pore volumes, average pore sizes and specific surface areas were separately characterized by t-plot method, density functional theory and Brunauer–Emmett–Teller theory. The surface morphology of the ACFs was observed by a scanning electron microscope. An X-ray photoelectron spectroscopy (XPS) technique was applied to determine the specific ratios of surface oxygen and nitrogen groups. Temperature programmed desorption was performed to investigate HCl adsorption behaviors over the ACFs. As experimental results, the thermal process decreased the carbon–oxygen groups while H2O2 impregnation increased the carbon–oxygen groups (especially carbonyl and carboxyl). Nitroso and nitro groups were introduced onto the carbon surface after HNO3 and Cu(NO3)2 treatments. The removal efficiency of HCl was improved slightly by H2O2 modification due to the increase of carbon–oxygen groups, and significantly by HNO3 and Cu(NO3)2 treatments because of newly formed nitrogen–oxygen groups. Nitroso and nitro groups show significant promotion of HCl retention ability over the ACFs surface. Moreover, HCl removal efficiency was much more influenced by nitroso than nitro groups.
For the past decade, activated carbon fibres (ACFs) have been widely used in air purification on the removal of contaminants such as SOX, NOX, VOC and Hg0 due to their low cost, large specific surface area, highly porous structure and various surface functional groups.3–14 It is recognized that the textural properties and surface chemistry are the most important properties of ACFs for applications in adsorption and catalytic processes. Bansal and Suzuki reported that the adsorption capacity of ACFs was closely related to the textural properties as well as surface chemistry, in particular the amount of surface functional groups.6,15 Unfortunately, untreated ACFs only have very small amount of surface functional groups, and hence various treatments, i.e. acid and alkaline modifications, have been adopted to modify the carbon surface.16,17 Park used oxygen plasma to yield oxygen enriched ACFs to capture HCl gas.18 After several times of treatment, the HCl adsorption capacity of ACFs can be improved as much as 300% due to the increase of surface oxygen groups, however, the high cost limits the practical applications. Meanwhile, more cost-effective ammonia modification was applied by Mangun to obtain ACFs of high HCl adsorption efficiency.19 Ammonia treatment significantly enhanced HCl adsorption due to the introduction of nitrogen–hydrogen groups, i.e. NH, –NH2, –NH3, CN, etc. However, nitrogen–oxygen species, i.e. nitroso (–NO2) and nitro (–NO3) groups, deserve much attention since those nitrogen species are generally associated with the catalytic removal of acidic gases, i.e. SO2
20 and NO,21 whereas few relevant work was conducted. Particularly, among the studies regarding HCl capture, ACFs were mostly treated by just one single method which is not sufficient enough to fully understand the influence of surface groups, because single method could not only affect the surface chemistry but also the textural properties of carbon materials.
In this paper, both physical and chemical techniques, i.e. thermal treatment and chemical impregnations in H2O2, HNO3 and Cu(NO3)2 solution, were employed to modify the ACFs surface, and scanning electron microscopy (SEM) was utilized to observe the morphology changes. Moreover, the amounts of surface oxygen and nitrogen groups were thoroughly examined through X-ray photoelectron spectra (XPS), and atom ratios of O/C and N/C were given based on elemental analysis results. Subsequently, HCl adsorption was carried out and temperature programmed desorption (TPD) was used to investigate HCl adsorption behavior over the ACFs surface. Based on such information, effects of surface carbon–oxygen and/or nitrogen–oxygen groups on the removal efficiency of HCl were comprehensively discussed.
η = (1 − Coutlet/Cintlet) × 100% | (1) |
Sample | SBET (m2 g−1) | Vtotal (cm3 g−1) | Vmicro (cm3 g−1) | Dpore (nm) | Atom ratio | |
---|---|---|---|---|---|---|
O/C | N/C | |||||
ACF-AR | 1270 | 0.56 | 0.35 | 1.78 | 0.09 | 0.019 |
ACF-1 | 1263 | 0.55 | 0.36 | 1.70 | 0.07 | 0.024 |
ACF-2 | 1248 | 0.55 | 0.35 | 1.78 | 0.13 | 0.012 |
ACF-3 | 1132 | 0.51 | 0.34 | 1.78 | 0.11 | 0.036 |
ACF-4 | 1098 | 0.49 | 0.30 | 1.77 | 0.09 | 0.038 |
For sample ACF-1, both SBET and the average pore width were diminished while micropore volume almost keep the same. According to Table 1, the average pore size decreased because macropore can be shrunk by thermal treatment.27 In the case of ACF-2, the decrease of SBET can be attributed to the slight decline of pore volume. For ACF-3 and ACF-4, SBET were even reduced by over 10% due to the decline of mesopore and/or macropore volume. Micropore structure in ACF-4 can be damaged by the pore blocking from surface oxygen groups into some micropores.28 In addition, SBET and the total pore volume of ACF-4 declined much more than those of ACF-3 while the average pore size of ACF-4 keep the same as ACF-3, further indicating that Cu(NO3)2 treatment did block the macropore instead of micropore. SEM images of ACF-AR and ACF-4 are presented in Fig. 3 to study the changes of surface morphology. It can be observed that carbon surface was uniformly modified after Cu(NO3)2 modification, leading to the decline of SBET and the total pore volume.
In the case of ACF-1, the specific ratio of C–C increased significantly while the amount of C–O and CO decreased slightly. Thermal treatment reduced the amount of C–O and C
O, and reformed to C–C by inducing secondary reactions at high temperature.27,30,31 For sample ACF-2 and ACF-3, the increase of C
O and COO, as well as the decrease of C–C can be owned to severe oxidation by H2O2 and nitric acid.27,29,32 As for ACF-4, the amount of C
O was slightly increased while the contents of other oxygen species kept the same as ACF-AR, indicating that Cu(NO3)2 has very limited impact on surface carbon–oxygen groups. Moreover, the slight change in surface oxygen groups may associate with the change of total acidity from Cu(NO3)2 treatment.
Fig. 5 displays the specific ratios of nitrogen groups over the ACFs surface before and after modifications, and the ratios of surface nitrogen groups before and after HCl adsorption experiments were simultaneously presented in Fig. 6. Data of ACF-1 and ACF-2 are not given because no obvious change in nitrogen groups were detected after treatments. For ACF-3 and ACF-4 in Fig. 5, the amounts of pyrrole, pyridine and quaternary nitrogen sharply dropped, while the specific ratios of N–O kept the same. As shown in the histogram, new nitrogen species, i.e. –NO2 and –NO3, were newly formed after modification. Overall, nitrogen groups were significantly modified since N treatment could fix nitrogen functional groups on external and/or internal surface.29 From Fig. 6, after HCl uptake the amounts of –NO2 and –NO3 respectively experienced reductions of 73%, 14% for ACF-3, and 61%, 36% for ACF-4, while the specific ratios of other nitrogen groups were all increased. It is notable that the content of –NO2 decreased much more than that of –NO3 after capture HCl.
![]() | ||
Fig. 6 Specific ratios of surface nitrogen groups for the ACFs before (ACF-3, ACF-4) and after (ACF-3–HCl, ACF-4–HCl) HCl adsorption. |
Fig. 7 shows XPS Cl2p spectra of the ACF-3 and ACF-4 with HCl adsorption. The binding energies around 201.1 eV in the spectra can be attributed to the chlorine atoms chemically boned onto the ACFs.
In the case of ACF-1, breakthrough time was slightly shortened after thermal treatment at 900 °C, as thermal treatment can influence not only the carbon textural properties but also the distribution of surface oxygen groups27 and thus reduce the HCl retention ability over the ACFs surface. Interestingly, ACF-1 exhibited different adsorption behavior from other ACFs when the efficiency is below 10%. However, rare research was carried out to investigate HCl adsorption over the ACFs surface and detailed mechanism still need further efforts, though adsorption behaviors of other acidic gases (i.e. SO2, NO and H2S) over thermal treated ACFs have been investigated by many scholars.3,8,13,33
ACF-2 showed an extremely similar profile as ACF-AR except for the extended period of complete breakthrough. Changes in surface chemistry showed that the amounts of C–C and C–O were reduced and relative ratios of CO and COO were increased after H2O2 treatment. Perhaps oxygen groups, such as C
O and COO, should take responsibility for the enhancement of retention ability over the carbon surface, since they can significantly affect the transport of HCl molecules adsorbed into microspores.18,34,35
The breakthrough time of ACF-3 and ACF-4 were extended from 360 to 1050 and 900 s respectively. Unfortunately, changes in the textural properties (in Table 1) could not support the enhancement of HCl removal since both ACF-3 and ACF-4 have a lower specific surface area than that of ACF-2. XPS analysis in Fig. 4 revealed that the specific ratios of surface carbon–oxygen groups, in particular the CO and COO, were increased after modification. Nevertheless, the slight increase of surface oxygen groups may enhance HCl removal but could not precisely interpret why removal efficiencies of ACF-3 and ACF-4 were improved so much. In other words, there must be other factors conducive to HCl adsorption. The newly formed nitrogen–oxygen groups could be the most probable reason for the improvement of removal efficiency, according to Fig. 5. Furthermore, sharp decrease of –NO2 and –NO3 were observed in Fig. 6 after HCl adsorption, suggesting the significant promotion effect of –NO2 and –NO3 over the removal of HCl. Notably, adsorption processes consumed much more –NO2 than –NO3, indicating the more important role of –NO2 than –NO3 during adsorption. This interesting phenomenon can be explained by following assumptions. –NO2 group can be substituted by chlorine atoms in adsorbed HCl (HCl(ads)) molecules, and generated chlorides and adsorbed HNO2 (HNO2(ads)). Meanwhile, HNO2(ads) can be oxidized to HNO3. Both HNO2 and HNO3 are not stable and thus easy to be decomposed to NO and/or NO2. –NO2 was consumed much more than –NO3 due to the lower stability of –NO2 groups. Possible reaction pathways are presented in Fig. 9.
![]() | ||
Fig. 9 Possible HCl adsorption products and mechanisms that are proposed for the consumption of nitrogen groups. |
![]() | ||
Fig. 10 TPD analysis for the ACFs before (ACF-3, ACF-4) and after (ACF-3–HCl, ACF-4–HCl) HCl adsorption. |
In Fig. 10(a), there is only one HCl desorption peak for ACF-AR, whereas two peaks appeared for ACF-3 and ACF-4. In the initial stage (<150 °C), all materials released significant amount of HCl due to gas desorption from the carbon micropores. However, ACF-3 released much more HCl than ACF-AR while actually ACF-3 has a 10% lower specific surface area than ACF-AR. A possible explanation could be that the first desorption peak come from enhanced physical adsorption.19 Subsequently, the outlet HCl concentration started to decline, yet an increase in outlet HCl concentration was noted again for ACF-3 and ACF-4 when temperature was over 400 °C, followed by a gradual decline to zero. Overall speaking, ACF-3 exhibited different desorption behaviors from ACF-4.
According to TPD profiles in Fig. 10(b), ACF-3 desorbed much more NOX than ACF-4 mainly since adsorbed HCl molecules could interact with –NO2/–NO3 groups and generate HNO2 and/or HNO3 over the ACFs surface. Meanwhile, both HNO2 and HNO3 are unstable and thus can be decomposed into nitrogen oxides, i.e. NO and NO2, which can be liberated in the form of NOX at elevated temperature. Moreover, HNO3 treatment could significantly impact the totally acidity of the ACFs surface, which favored the formation of acidic gases. Hence a considerable amount of NOX was emitted for ACF-3 while only a small amount of NOX was released for ACF-4 at lower temperature. However, with the proceeding of desorption, no more NOX was detected after 200 °C probably due to the depletion of surface nitrogen species.
TPD analysis in Fig. 10(c) was provided to understand the desorption behavior of ACF-3 and ACF-4 for comparison. For ACF-3, –NO2/–NO3 over the carbon surface started to decompose at 100 °C and released significant amount of NOX until 400 °C, while in the case of ACF-4, NOX was detected in the effluent gas until 150 °C. Moreover, the temperature area of ACF-4 is much narrower than that of ACF-3. These interesting phenomena can be explained by the thermal decomposition of –NO2 and/or –NO3. –NO2 can be gently decomposed at low temperature while –NO3 can be quickly decomposed at high temperatures.36
Interestingly, after 200 °C no more NOX was desorbed for both ACF-3–HCl and ACF-4–HCl, indicating –NO2/–NO3 groups over the ACFs surface were mostly consumed in HCl adsorption process. However, the reactions among HCl molecules, –NO2/–NO3 groups, copper ion and carbon surface requires further effort since they could interacted with each other over the ACFs surface to form complex compounds.37 Fortunately, no matter what happened over the ACFs surface, nitrogen groups, especially the –NO2 and –NO3, exhibited significant promotion on the removal of HCl gas.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra11705d |
This journal is © The Royal Society of Chemistry 2015 |