The concept of synthesizing CNTs using solid polymers was first reported about 16 years ago.[152-155] Since then there have been numerous efforts on further advancing such low cost CNT synthesis route. Work have been done with regards to types of plastics, conversion processes, growth conditions, catalysts, and quantities and qualities of the resulting CNTs. Various processes have also been explored for such conversions, and there is one (and the only one up-to-date) review article covering this topic. Nevertheless, that work examined processes based on the reactors types alone (autoclave, crucible, fix bed (tube furnace, muffle furnace), moving bed, fluidized bed, etc.). Instead, the authors hereby focus on both the common and on the differentiating features of the existing waste-plastics-to-CNT conversion processes. Thereby, such conversions may be classified into two categories: (1) one-pot conversion where synthesis of CNTs occurs upon the in situ formation of carbon feedstocks from the solid plastic waste; and (2) stepwise conversion where synthesis of CNTs occurs after the formation of carbon feedstocks from the solid plastic waste. Typical processes are summarized in Table 4, which includes their representative CNT products, related conversion conditions, as well as corresponding references.
One-Pot Conversion of Plastics into CNTs
One-pot synthesis of carbon nanotubes typically starts with solid polymers which are mixed with catalysts. A heat resource is then applied to catalytically decompose (pyrolyze) the plastics. The decomposed products, either in liquid or gaseous phases, serve as carbon sources for the growth of CNTs on the catalysts. Polymers, such as polyethylene (PE),[157-161] polypropylene (PP),[160, 162-168] polystyrene (PS), polyvinyl alcohol (PVA), polyvinyl chlorine (PVC), polytetrafluoroethylene (PTFE), polycarbosilane (PCS), phenol formaldehyde (PF), and polyethylene terephthalate (PET), etc. have been studied using this method (Table 4). Various catalysts have been examined, including transition metals in either elemental form (nickel,[157, 174, 175] iron,[163, 172] etc) or in chemical compound form (nickel oxides,[168, 176-178] ferrocene,[158, 167] ferrous chloride, cobalt acetate, etc.) among others. Heat is supplied by either electric furnaces (fixed beds,[163, 178, 179] autoclaves, and fluidized beds,[160, 164]), or by combustion of fuels.[162, 166, 168, 177, 180]
This one-pot synthesis features simultaneous plastic degradation and the CNT synthesis, and efforts have been made to reveal how it proceeds. Jiang et al. proposed a possible reaction mechanism, using PP as the sample feedstock with nickel-based catalysts, as illustrated in Table 4. With the presence of catalysts, plastics are degraded (decomposed), by active intermediates of carbenium ions, but not free radicals which play a major role in noncatalytic thermal decomposition of plastics. It is suggested that the resulting “free carbons” from catalytic pyrolysis of plastics will then dissolve into the catalyst, diffuse through and then precipitate at its surface to form CNTs. Such one-pot synthesis can be greatly affected by utilizing a combination of metal catalysts with solid acids (such as organically-modified montmorillonite (OMMT)[166, 168] or zeolite), chlorinated compounds (such as CuCl), and/or activated carbon. Solid acids are suggested to provide the intermediate proton acidic sites, which will assist breaking down the molecular chains of the plastics[166, 182] favoring the formation of CNTs. On the other hand, chlorine radicals could promote the dehydrogenation and aromatization during the degradation of the plastics, resulting in more carbon nanofiber and less CNTs. Activated carbon plays multiple roles in the conversion of plastics into CNTs, such as (1) to absorb and to be functionalized by the fragment radicals from the plastics decomposition products, (2) to promote the formation of light hydrocarbons and aromatic compounds (especially the diaromatic and polycyclic hydrocarbons) through the fragment radicals as well as the aromatic intermediates, and (3) to further assist catalyzing the dehydrogenation and aromatization of the aromatic intermediate or PAHs.
Different reaction conditions (temperatures, catalyst compositions/concentrations/loads, reaction durations, etc.) usually lead to CNTs with various quantities and qualities. In addition to existing reviews on such topics, the authors would like to briefly discuss the one of the features this one-pot synthesis processes: the CNT production yield. It is worth mentioning, that there are at least three different definitions in production yield: the mass production rate of CNTs (unit: g/h), the mass of CNTs over the mass of the catalyst (unit: g/g), and the mass of CNTs over the mass of the feedstock (unit: g/g or %).The former two definitions are commonly used in traditional CNT production processes, as the corresponding focuses are either on the end products (CNTs) or on the activities of the catalysts. In waste-to-CNT conversion processes, however, the third definition is predominately used, as the focus of such processes is on conversion efficiency. That is why the data compiled from all different processes is not directly comparable. The yield could be as low as several percent (e.g. 5% in Ref. 178), and this number can reach 50% or higher,[156, 178] under optimum conditions. Such a high yield conversion is achieved typically in a setting where the mix of plastics and catalysts is enclosed within a container, such as a crucible. Comparing with an open system such as in a quartz tube reactor, where the pyrolyzed products (usually in gaseous phase) are carried away by the processing gas, this enclosed environment enables a longer reaction time between the plastics pyrolyzates and the catalysts. Therefore this could be a reason for such high conversion rate. High temperatures (>700°C) and compound-based catalysts are also found to lead to high CNT yields.[158, 178]
The impurities in the CNTs generated by both one-pot processes and stepwise processes usually consist of residual catalysts and amorphous carbon. Impurities typically account for 10 wt % of the as-produced CNTs. The lowest amount of impurity reported was 0.3 wt % as determined by thermogravimetric analysis (TGA), a common method to investigate the impurity and thermal stability of CNTs. In order to keep the number of impurities low, efforts should be made to optimize process parameters as: the components and dispersion of catalysts, the feeding rates of plastics, the reaction temperature, etc., for the catalysts be kept active and produce CNTs with minimum defects.
Stepwise Conversion of Plastics into CNTs
The stepwise conversion process features sequential reactions, typically starting with the thermal decomposition of plastics. As a second step, the resulting gaseous products (hydrocarbons) are then channeled downstream where they react with catalysts to form CNTs. This is the method applied in the first reported explorations of plastics to CNT conversion,[153-155] since then additional works have been done to investigate the underlying fundamentals as well as to advance the technology of such conversion processes. The plastics-to-CNTs conversion mechanism is rather complicated, as it combines the steps of high temperature plastic cracking,[185, 186] high temperature oxidation/combustion,[187, 188] thin film formation through chemical vapor deposition and gas-solid chemical conversion through heterogeneous catalysis.[50, 189, 190] A comprehensive review has been done by Bazargan and McKay, addressing among other topics this formation mechanism, whereas this concise review highlights significant technology developments.
An inherent feature of such a stepwise process is that it enables individual controls over its subprocesses. That is to say, decomposition (pyrolysis) of plastics can be processed without interference with the synthesis of CNTs. For instance, Yang et al. employed a three-stage reactor, where catalysts (ferrocene) was first sublimated, in the first reactor, before entering the second reactor, where plastics (either PE, PP, or PVC) were decomposed. The mixture of the formed catalysts and the gaseous pyrolyzates was then channeled into the third reactor, where CNTs were formed and collected. Their setting enabled independent controls of reaction temperatures (120–140°C for ferrocene sublimation, 450°C for plastics decomposition, and 800–850°C for CNT growth). With reaction time in the order of 10 min, lengths of CNTs were typically in the order of 100 µm. It is not clear, however, whether the iron particles (sublimation products of ferrocene) only catalyzed the growth of CNTs, or they also participated in the thermal decomposition/pyrolysis of plastics.
An improved two-stage stepwise process was investigated by Liu et al. Polypropylene (PP) was catalytically pyrolyzed over HZSM-5 zeolite in a screw kiln reactor, and the resulting pyrolysis gases were subsequently decomposed over nickel catalysts in a moving-bed reactor. CNTs and hydrogen were produced simultaneously, and it was found that 700°C was the optimum decomposition temperature, in terms of maximum productions of both CNTs and hydrogen with a synthesis temperature of 750°C. Similar work was reported by Wu et al., where a fixed-bed two-stage reaction system was used. Waste plastics samples were pyrolysed in N2 in the first stage at 500°C, and the resulting compounds were further used to synthesize CNTs at 800°C in N2. Water was injected into the second reactor, where catalysts (Ni/Ca-Al or Ni/Zn-Al) were also present, to jointly facilitate the CNT formation.
Recently, the authors developed a hybrid reactor, combing sequential feeding, pyrolysis, premixed-combustion, and CNT synthesis process (Zhuo et al.[194-200]) To achieve conversion of solid hydrocarbon fuels to CNTs, quantities of a solid fuel, in pelletized form, were first thermally pyrolyzed into a stream of gaseous decomposition products (pyrolyzates) inside a pyrolyzer furnace at 600 to 800°C, depicted on the left side of Figure 4. Pyrolysis occurred in inert nitrogen, to prevent the ignition and combustion of the pyrolyzates therein. The pyrolyzates then flowed through a mixing venturi where they were thoroughly mixed with preheated oxygen-containing gases. The oxygen/pyrolyzate charge autoignited at the exit of the venturi, forming a premixed flame. The fuel/oxygen ratio in the venturi was set to be fuel-rich (i.e., oxygen deficient) by adjusting the amount of oxygen introduced to the venturi. This ensured that there were large amounts of carbon-bearing compounds (CO and unburned aliphatic hydrocarbons), hydrogen, and water vapor in the combustion effluent, as needed for the efficient growth of CNTs. The combustion effluent then entered the second furnace where fixed catalyst substrates were also preinserted (such as stainless steel screens or honeycombs). Growth of CNTs occurred on the catalyst substrates at 600 to 1000°C. To prevent soot, which may be generated in the flame, from entering the second furnace and inadvertently depositing on the catalyst substrates and contaminating them, a high-temperature ceramic barrier filter had been inserted at the entrance of the CVD synthesis chamber, see Figure 4. This feature allowed operation of the apparatus above and beyond the equivalence ratio φ(defined as (mfuel/moxygen)actual/mfuel/moxygen)stoichiometric)that corresponds to the soot onset thresholds.[140, 201] The ceramic filter may be periodically thermally regenerated in situ, at 1000°C in air, to burn any entrained soot particles.
Figure 4. a: Four-stage laminar-flow, electrically-heated, muffle furnace, used for proposed work; zone 1: steady- state, steady flow, continuous feeding stage, zone 2: pyrolysis stage, zone 3: combustion stage*, and zone 4: CVD-synthesis stage. b: Ceramic (SiC) honeycomb filter manufactured by Ibiden. c: stacked catalyst in the forms of either woven screens or honeycomb structures. *: A flame is present at the combustion stage only when oxygen-containing gases are added to the venturi. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
Download figure to PowerPoint
Another unique feature lie on the design of the venturi section, which enables the introduction of additional materials, such as one or more gases (e.g., oxidizing agents such as oxygen gas, chlorine gas, carbon dioxide, any other gas containing oxygen, and the like), or catalysts (such as ferrocene) to mix with the gaseous decomposition products, which enter the venturi section.
Use of stacked catalyst substrates (screens or honeycombs) will allow not only for high-yield production of CNTs but, also, for selective and detectable depletion of some carbon-bearing gases in the combustion effluents as CNTs form. Sampling the gases before and after the catalyst substrates from sampling ports 1 and 2, as illustrated in Figure 4, can help determine which gases are likely growth agents for CNTs. In summary, this process incorporates the following unique features: (a) It pyrolyzes or gasifies solid wastes to generate gaseous carbon-bearing components. (b) It affords control of the fuel-to-air ratio in the venturi from fuel-lean to fuel-rich and, to the limit, to purely pyrolytic confitions at the absence of oxygen. (c) The pyrolysis and the CNM synthesis occur in two separated stages, which afford individual control of the subprocess conditions. (d) The risks of handling highly reactive gases (such as H2), and/or highly toxic gases (such as CO), are eliminated as they are generated in situ in an inert gas. Finally, (e) this process can generate a hydrogen-enriched syngas.
Moreover, this process can be designed to be energy efficient. For instance, Jinno et al. measured the heat of pyrolysis of polyethylene (PE) to be 254 kJ/kg, independently of the heating rate. Comparing this heat of pyrolysis with the heating value (energy content) of this polymer, which is 46,000 kJ/kg, it becomes evident that a small fraction of the heat released during combustion may be fed back and pyrolyze this solid fuel. Similarly, the energy fluxes necessary to pyrolyze polypropylene (PP) and polystyrene (PS) are 44,000 and 40,200 kJ/kg, respectively. Thus, the energy balance of the proposed process is overwhelmingly positive. A portion of the heat released during the combustion of the polymer pyrolyzates may be fed back to the pyrolyzer through a heat exchanger to gasify incoming precursors, whereas the remaining heat may be used elsewhere as process-heat or to generate electricity. Regarding the environmental impact of this process, it should be stated that the use of premixed flames, high temperature filtration, and possibly a post-process afterburner can drastically reduce unburned or partially burned health-hazardous species, as shown before in the author's laboratory.[204-210] The generation of the greenhouse gas CO2 could be an issue when waste polymers are used as feedstocks. However, by using such waste hydrocarbon-based post-consumer products, the conventionally used, more valuable premium hydrocarbons are merely substituted, and thus, CO2 emissions (that would have been released during their extraction and processing) are avoided. In addition, there are also benefits to cleaning the earth of nonbiodegradable wastes. Generating value-added products, such as CNMs, from recycled waste plastics is thought to enhance the motivation for recycling, lower the cost of the products, and in turn, help overcome the hurdles of their large-scale use in consumer and industrial applications.
Challenges and Possible Solutions
One of the major challenges in plastic-wastes-to-CNT up-cycling comes from the nature of waste feedstocks, where there is a lack of consistent and reproducible supply of carbon feedstock with controlled quality (in terms of compositions and impurities). Feedstocks from recycling streams are typically mixtures of various types of plastics. The resulting pyrolysis products, of different compositions, are known to affect the formed CNTs, while the detail correlations are yet to be revealed. What is more, waste plastics containing fillers and/or other additives could have detrimental effects on the quality of formed CNTs. These two problems get further complicated when there are variances among different batches of waste plastic streams.
Another challenge of such up-cycling processes lies on the process complexity. Comprehensive scientific investigations are still needed in both one-pot and stepwise methods, especially those involved on the driving force behind the proposed waste-to-CNT up-cycling process. In the case of stepwise method, for instance, whereas gaseous feedstocks have been correlated directly to CNT products,[212-216] the fact that these feedstocks pyrolyze at the high reaction temperatures therein (>400°C) to form intermediate products has been largely ignored.[217-219] Therefore, the CNT catalysts therein have experienced a blend of carbon-bearing gases, which are very different from the input feedstocks. Hence, it remains unresolved whether the CNT synthesis process consumes the carbon-bearing gases equally or selectively.[194, 195] Simply building a connection between the initial feedstocks and the CNT products is not sufficient. The thermally-driven chemistry that takes place between the input and the output also needs to be accounted for. It is therefore worth examining what carbon-bearing gases get consumed at the catalyst substrates, and are thereby converted to CNTs. Further understanding of the detail reaction schemes will help engineers control the quality of produced CNTs that will satisfy the commercialization objectives.
There are solutions to overcome such challenges, albeit empirically, by adapting technology advancements from plastic waste recycling processes, as well as from CNT synthesis processes. For example, intelligent waste identification/separations can be assisted by (tribo)electrostatics, laser-induced plasma spectroscopy,[220, 221] laser-induced breakdown spectroscopy,[222-224] density media, artificial neural networks, etc. Plastic waste decompositions, on the other hand, can be benefited from the well-established pyrolysis/gasification fundamentals,[227-233] reaction chemistry (thermodynamics, kinetics,[235-238] mechanisms, etc.), as well as the related technology developments, such as those of catalysts,[239-241] additives,[242, 243] process, or their combinations. Last but not least, CNTs synthesis, especially their yields with controlled quality, could also gain from recent progresses in process engineering at nanoscale.[21, 111, 245-249]