WO1993023331A1 - Plasma method for the production of fullerenes - Google Patents

Abstract

A process for the production of fullerenes using plasma is disclosed. Carbon powder is supplied through line (214) to a plasma flame located in plenum (206) between cathode (202) and anode (204). The plasma gases leave plenum (206) and proceed to pathway (216). Pathway (216) is lined with refractory insert (217) which maintains the temperature sufficiently high to allow annealing of fullerene precursors into fullerene molecules.

Description

PLASMA METHOD FOR THE PRODUCTION OF FULLERENES

FIELD OF THE INVENTION This invention is a process for the production of fullerenes, including buckminster-fullerene, using a plasma. The process involves the introduction of particulate carbon-containing material into a plasma jet reactor using a suitable plasma gas to produce fullerenes or fullerene derivatives.

BACKGROUND OF THE INVENTION

Fullerenes, recently discovered by Smalley, Curl, Kroto, Heath and O'Brien [Nature, 318, 162 (1985)], are representative of a set of carbon molecules which have been shown to have both aromatic and olefinic character. They are the third allotropic form of carbon or rather a family of allotropes. The other two allotropic forms of carbon are diamond and graphite. The simplest of the fullerene molecules is a spherical C^ molecule, called buckminster-fullerene, having the geometry of a truncated icosahedron - a polygon with 60 vertices and 32 faces, 12 of which are pentagons and 20 are hexagons. Other fullerene molecules have been identified and include C₇₀, C₇₆, C₇₈, C_M, Cgo, C^, and others up to C₂₆₆. See, Parker et al. J.Am. Chem. Soc. 1991, 113, 7499-7503. Methods for the production of fullerenes have been described in a number of journals. For instance, in J. Phys. Chem. 1990, £4., 8634-8636, Haufler et al describe an apparatus and a process for producing fullerenes in which a graphite rod is vaporized in an arc using a 100-200A current in a helium atmosphere held near 100 Torr. The resulting soot is extracted with boiling toluene to form a dark red-brown liquid predominantly of C^ and higher fullerenes. The liquid extracts about 10% of the weight of the originally recovered soot.

A similar description is made by Kratchmer et al in Nature, 347, 354 (1990)].

Laser ablation methods for the production of fullerenes are also described in the scientific literature (J. Burgess; CHEMISTRY AND INDUSTRY (London), 22, 733, 1990) as have combustion processes using aromatic hydrocarbons as feedstocks (D. H. Parker et al. JOURNAL OF THE AMERICAN CHEMICAL SOCIETY, 113, 7499, 1991).

A few processes have been described which use optimized flames as the source of heat for production of the fullerenes (J.B. Howard et al.

NATURE 1991, 352, 139-141). Similarly, Mckinnon et al, COMBUSTION AND FLAME 88: 102-112 (1992) shows the production of low yields of fullerenes using low pressure benzene/oxygen/argon flames. In Peters et al. Angew. Chem. Int. Ed. Engl. 31 (1991) 223-4, a process using inductive heating of various carbons (pure graphite, pyrolytic graphite, and glassy carbon produced from pyrolyzing polymeric precursors) at 2700°C in a 150 hPa helium produced significant amounts of fullerenes, particularly higher molecular weight fullerenes.

Nevertheless (and probably because of the research nature of the work described) , the majority of published process descriptions for the production of fullerenes involve the use of electric arcs in a helium or argon atmosphere. In an arc-induced fullerene process, an electric current is ^"passed through a gap between two carbon rods or electrodes. The gas in the vicinity of the electrode ends ionizes and allows the formation of the arc. The carbon electrodes are consumed by the process and are clearly directly heated by the arc as a consequence of the arc process.

In contrast, in its most basic aspects, my process introduces particulate carbon into a plasma, such as might be found in a plasma jet or plasma spray torch. The plasma permits the process to operate in a continuous fashion, indirectly heats the particles to allow control of the rate of the process, and further permits variation of the feed material composition. Special to this invention is the use of a refractory insert at the plasma arc which creates a long reactor space in which the temperature declines slowly and allows the fullerenes to slowly self-synthesize.

Plasma jet devices are widely used industrial devices. For instance, typical uses include the coating of a substrate with a metal layer produced through the melting and transporting of melted particles in a plasma jet emanating from a torch¬ like mechanism. Torches of similar design are also used in cutting metals.

Typical designs for plasma torches used in coating substrates with metals are shown in PLASMA SPRAYING, Smart et al. Mills & Boon Ltd. London, 1990. In these torches, a voltage is applied between a water-cooled anode (often made of thoriated tungsten) and a copper anode in the form of a nozzle. The anode is typically water-cooled. The space between the cathode and anode is annular. The plasma gas is introduced into this annulus. Plasma gases for these applications are chosen for a variety of criteria, e.g., they are not reactive with the particles introduced into the plasma, their ionization potential, and their abilities to perform the specific task required by the operator of the device. Typical plasma gases used in this service are hydrogen, helium, nitrogen, argon and their mixtures. Helium has a relatively high ionization potential and consequently its plasma is of a fairly high peak temperature, i.e., as much as 20,000°K. In a plasma sprayer, the particles which are eventually to make up the coating are introduced radially into the plasma downstream of the arc where they are melted (if the desired coating is metal) , reacted with ancillary materials (if the coated material is an oxide, nitride, etc.), and propelled to the substrate surface. Careful design of the torch and sizing of the particles is needed to insure that the particles are neither vaporized nor only partially melted.

The use of plasma arcs to produce fullerenes is described in Yoshie et al, Appl. Phys. Lett. 61 (23) pp 2782-3 and in published European Patent

Application 527,035, to Eguchi et al. These disclosures teach a reaction zone using a hybrid plasma employing a superimposed radio frequency plasma on a dc arc jet. The zone is externally heated.

In my process, carbon particles are introduced into the plasma, indirectly heated (likely partially or completely vaporized) in a reaction zone having a slowly declining temperature, and the resulting fullerenes and byproduct soot are collected at a lower temperature.

Others have introduced carbon particles into plasma torches without producing fullerenes. A process for the production of carbon monoxide is shown in Giacobbe et al, ADVANCED POWDER FEEDING DEVICE FOR USE IN GAS/SOLID PLASMA SYNTHESIS AND PROCESSING APPLICATIONS, Met. Res. Soc. Symp. Proc. , Vol. 98, pp. 405-416, (1987). In that process, powdered carbon (in particular, carbon black) is introduced into a plasma which is produced in a device using carbon dioxide as the plasma gas. The carbon dioxide is also used as the carrier for the fine carbon powder. No mention of fullerene production is made.

Plasma treatment of coal particles for a variety of purposes are known: Scott et al. CHEMISTRY AND INDUSTRY, 17, 739-40, 1976; Nicholson et al. NATURE, 236, 397-400, (1972); Chakravartty et al, FUEL, 55, 43-46, 1976; Chakravartty et al. FUEL, 55, 254-5, 1976.

A similar process is shown in Kong et al. Beta- SiC SYNTHESIS IN A THERMAL ARGON PLASMA JET REACTOR, Met. Res. Soc. Symp. Proc, Vol. 98, pp. 377-384, (1987) . There, ultrafine beta-SiC powders are produced by the introduction of silicon monoxide particles into a plasma jet reactor using methane. Fullerene is not mentioned as a product.

Use of a plasma torch in conjunction with methane to reduce metal oxides to the metal form is shown in Detering et al. REDUCTION OF SELECTED METAL OXIDES IN A THERMAL PLASMA PRODUCED BY A NONTRANSFERRED ARC TORCH, Met. Res. Soc. Symp. Proc, Vol. 98, pp. 359-364, (1987). Again, the production of fullerene is not shown.

None of the prior art shows the production of fullerenes using indirect heat transfer to particulate carbon as described here, particularly by the use of plasma jets or torches. None of the prior art processes is particularly useful as one which can be used on a commercial or larger scale.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 shows a generalized cutaway drawing of a plasma jet reactor suitable for use with the inventive process.

Figure 2 shows a cutaway drawing of a plasma jet reactor having a plasma chamber insert especially suitable for use with the inventive process. Figure 3 shows an NMR spectrograph of the product.produced in the Example.

SUMMARY OF THE INVENTION This invention is a process for the production of fullerenes , including buckminster-fullerene and metallo-fullerenes, using a plasma. The process involves the introduction of particulate carbon into a plasma jet reactor having an insert in contact with the plasma and using a suitable plasma gas to produce fullerene products.

DETAILED DESCRIPTION OF THE INVENTION

This invention is a process for the production of fullerene-containing products. The process generally involves the introduction of particulate carbon-containing feedstock into a plasma jet reactor having an insert which contacts the plasma and maintains the reaction volume at a high temperature level and which uses a plasma gas which preferably is one or more of the noble gases.

The plasma jet reactor design is important to the invention. Use of the desired plasma jet reactor having an insert which contacts the plasma and is therefore heated by the plasma is most desirable and provides for improved yields over that obtainable without the insert. Such an insert or its analogs which are resistively heated form a portion of this invention. A plasma gas of a noble gas, i.e., neon, argon, helium, or mixtures of those gases is desired. The plasma gas preferably comprises or consists essentially of helium, argon, or their mixtures.

The carbon feedstock is particulate and desirably of a small and reasonably consistent diameter. The carbon feed should be fairly pure if a pure fullerene product is desired. That is to say, that if a fullerene consisting essentially of carbon (and having only trivial amounts of, e.g., metals) is the desired product, the feed carbon should be of similar purity. It is highly desirable that the feed be quite dry. Appropriate carbons include graphite powder, carbon black, amorphous carbon, soot, carbon fibers, coal powder, active carbon powder, charcoal powder, coke powder, and other carbons which do not contain substantial amounts of other atoms, particularly oxygen. The size of the carbon particles is desirably quite small. Functionally speaking, it is desirable that the particles be of a size which could be vaporized in the plasma stream to allow self-synthesis of the fullerenes to take place. When using graphite particles, the upper size limit desirably is about - 300 mesh.

When producing metallo-fullerenes, metal- containing compounds such as metals, metal oxides, metal carbides, metal nitrides, metal oxycarbides, metal oxynitrides, metal halides (preferably chlorides) , metallo-organics, and metallo -organic complexes are introduced into the plasma stream with the carbon-containing material. Metals in these compounds may be any metal having an electronegativity of less than about 2.2 Pauling Units. Especially preferred are the alkali metals, the alkaline earth metals, and the rare earths, both the actinide and the lanthanide series. Introduction of these materials into the plasma arc results in the formation of fullerene cages with endoskeletal metals.

Use of boron- or nitrogen-containing analogs of the noted materials results, however, in cage- substituted fullerenes.

The reaction is believed to take place in or shortly in front of the tip of the plasma in that the carbon is preferably substantially completely vaporized. However, I have found that by maintaining the plasma arc in a sleeve or insert which is in contact with the plasma and which acts as thermal insulation to the plasma, e.g., a refractory mass such as zirconia, alumina, thoria, graphite, pyrolytic carbon, coke or other carbonaceous mass, and so allows the reactants to remain at an elevated temperature, the yield of fullerene is significantly higher. Specifically, the insert should be of a length at least as long as five diameters of the diameter of the plasma gun nozzle. At that downstraem five diameter point, the temperature of the interior of the insert should be ot least 500° C, preferably at least 1000° C, and most preferably 1500° C. Preferably the inside diameter of the insert is less than about twice the size of the diameter of the plasma gun nozzle. The insert need not adhere to the listed desired physical dimensions if it is heated by soe other means, e.g., by direct resistive heating or by R.F. induction heating of the insert.

Cooling or quenching the plasma shortly after exiting the refractory insert is most desirable, but is not absolutely necessary from the aspect of producing fullerene-containing materials. The gases may further be cooled prior to collecting the fullerene-containing product and any by-product soot. After removal of the fullerenes by extraction with, e.g., ethyl ether, or sublimation and removal of any other extraneous materials, the soot may be recycled to the plasma reactor. Obviously, the plasma gases may be recycled as well.

The pressure of the reaction is desirably at atmospheric pressure or below.

Figure 1 shows a suitable reactor for the operation of the inventive process. The plasma generator is made up of a cathode (102) and an anode (104) which also serves as a nozzle to accelerate the plasma gas supplied to the plenum (106) between the cathode (102) and anode (104) through gas supply line (108) . Cooling water may be introduced through supply line (110) and withdrawn through line (112) . Carbon powder may be supplied into the plasma flame found in plenum (106) through line (114) . The plasma gases leave the plenum (106) and proceed down the open pathway (116) downstream of the plenum (106) . This pathway acts as a quench zone both because of the expansion of the plasma gas and because of the coolant circulating about the outside of the spoolpiece (118) through entrance and exit coolant lines (120) and (122) . The first zone also allows the fullerene to condense typically on the particulate carbon flowing through the zone.

A second cooling zone within spoolpiece (124) may be added if desired. I have found that air cooling is sufficient for this stage. This stage is typically used only to lower the included gas temperature to a level sufficiently low so that temperature sensitive filter materials may be used. Finally, a filter assembly (126) containing one or more filters (128) of appropriate porosity is also included to remove the particulates and fullerenes before the plasma gas is disposed of via vent (130) . The plasma gas may obviously be recycled if so desired. Figure 2 shows a preferred variation of the previously described plasma reactor for the operation of the inventive process. The plasma generator is made up of a cathode (202) and an anode (204) which also serves as a nozzle to accelerate the plasma gas supplied to the plenum (206) between the cathode (202) and anode (204) through gas supply line (208) . Cooling water may be introduced through supply line (210) and withdrawn through line (212) . Carbon powder may be supplied into the plasma flame found in plenum (206) through line (214) . The plasma gases leave the plenum (206) and proceed down the open pathway (216) downstream of the plenum

(206) . In this preferred variation, a refractory insert (217) is included within the pathway (216) . The refractory insert (217) is highly insulated and maintains the temperature at a high level and is thought to allow the fullerene precursors to

"anneal" into the complex fullerene molecule as the plasma gas proceeds through the reactor. The remainder of the device may otherwise be quite similar to the right hand apparatus as is depicted in Figure 1.

EXAMPLE 1 A plasma jet reactor was constructed. A commercial high energy plasma powder spray torch manufactured by Abrasive Industries, Westborough MA. using argon as the plasma gas was bolted directly to a water cooled quench tube of 1.25" I.D. The quench tube was 66" long and was constructed of copper because of its high thermal conductivity. A second stage of cooling was attached to the exit of the quench tube. The second stage was air cooled and was 37" long. The reactor was held under a vacuum (3 in-Hg) and the reactor output was collected using a paper filter backed up by a 5 micron nylon mesh filter bag. Particulate carbon (-300 mesh crystalline graphite powder having a BET specific surface area of approximately 12.0 M²/gm sold by Johnson Matthey) was loaded into the powder feeder for the plasma gun. The gun was started with 690-700 amperes at 34-39.8 volts. The arc gas flow rate was about 70

SCFH and the powder gas flow rate was 8 SCFH. During the following one hour run, the voltage on the gun varied between 34 and 37 volts; the amperage varied between 711 and 727 amperes.

About 345 grams of carbon were processed in the reaction run. Approximately 90.80 grams of a gritty, particulate, carbonaceous-appearing material was collected in the filters. Additional carbon particles were found in the entrance to the quench portion of the reactor.

The particulate material taken from the filters was extracted with 250 ml. of toluene using magnetic stirring. The extraction took about two hours; during the final one and one-half hours, the extractant was heated using a hot water bath at near boiling temperature. After about five minutes of extraction, the solution was seen to be pink. A few minutes later the solution turned orange.

The mixture was cooled and filtered to remove carbon particles. About 145 ml of the total liquid, corresponding to about 52.66 gm of soot from the filter, was distilled to a volume of about 20ml.

The pot liquid was evaporated on a watch glass. About 10 mg. of a dark blue-black solid product was secured. This amount corresponded to a production rate of about 72 mg. per hour. The product was dissolved in benzene-d₆ and subjected to a ¹³C NMR for determination of fullerene presence according to the method found in Ajie et al, J. Phys. Chem., 94, 8630-8633, 1990.

The resulting NMR graph is shown in Figure 3. The presence of a peak at 143.2 ppm established the presence of C₆₀-fullerene (buckminsterfullerene) and the presence of peaks at 145.4 and 148.1 ppm confirmed the presence of C₇₀-fullerene. Integration of those characteristic peaks showed that the sample contained approximately 60% C₆₀-fullerene and the remaining 40% C₇₀-fullerene. EXAMPLE 2 This Example compares operation of the process with a thermal insert in the reactor to operation without such an insert. A similar reactor to that described was constructed using a cooling tube having a 1.3125" ID and with a length of 53.375".

A carbon insert as shown in Figure 2 was installed at the plasma discharge. The insert was of 99% carbon and of a generally tubular shape of 5.6875" in length, 1" OD, and 0.4375" ID. The insert was insulated from the cooling tube by two layers (0.20") of zirconia felt which were cemented in place with a commercial zirconia cement. The cement was dried in warm air for a few hours, dried in an oven at 120°C overnight, and calcined at about 425° C for three hours prior to use.

The soot collector was a TYVEK filter having a nominal rating of 0.1 micron. A weighed amount of feed (crystalline graphite - minus 300 mesh) was introduced into the feed hopper in each instance and the hopper purged with helium as used to transport the powder to the gun. A vacuum (29 in-Hg) was then applied to the hopper and the inert gas, at 30 psig, introduced to the hopper. This cycle was repeated three times and the hopper was finally bled to its working pressure of 1-2 psig. Helium was added to the reaction downstream of the arc to prevent excessive arc voltage. Typical helium content in the reaction environment was 20- 25% (vol) .

Argon flow was initiated to the arc and the arc ignited. Operating arc voltage was 30-35 volts and arc current was adjusted to 700-950 amperes. The carbon powder flow was started, with 21.6 SCFH of helium, and the voltage and current were monitored for the .length of the run.

At run end, the remaining feed powder and the product soot were separately collected. The fullerene product was isolated using the following procedure:

45-90 grams of well mixed sample were soxlet extracted with toluene for 2-4 hours until the filtrate in the upper chamber was colorless. The solution was concentrated by distillation to about 30-50 ml and the liquid blown down with air without further heating. The resulting dry extract was washed with 3-5 portions of 2 ml ethyl ether to remove hydrocarbon contaminants. I have found that some fullerene oxide is removed with the ether extract. The ether was blown down with air at room temperature to produce an oily solid. - The ether soluble material was washed with hexanes to remove hydrocarbon contaminants and leave fullerene oxides.

The washed extract of fullerene and fullerene oxides was dissolved in toluene and filtered down a column of silica gel -60 (230-400 mesh) to trap insoluble particulates (particularly minor carbon particles and fullerene oxides) . After washing the fullerenes from the column with toluene, the silica column was washed with 2-3 portions of 2 ml ether to remove the remaining amounts of fullerene oxides. The various washings were combined and blown down with air in a tared flask. The resulting fullerene solid product was dried in air at about 120°C for 30 minutes. The following Table depicts the results of the comparative runs.

The yield of fullerenes is clearly much higher with the insert than without.

It should be clear that one having ordinary skill in this art would envision equivalents to the processes found in the claims that follow and that those equivalents would be within the scope and spirit of the claimed invention.

Claims

I CLAIM AS MY INVENTION:

1. A process for the production of fullerenes comprising the steps of: a.) introducing particulate carbon into the plasma jet of a plasma jet reactor passing through a thermal insert under conditions suitable for the production of a fullerene- containing stream, b.) recovering the fullerene-containing stream.

2. The process of claim 1 where the plasma jet comprises a noble gas.

3. The process of claim 2 where the noble gas is selected from neon, argon, helium, or mixtures of thereof.

4. The process of claim 3 where the noble gas comprises helium or argon or their mixtures.

5. The process of claim 1 where the particulate carbon is selected from graphite powder, carbon black, amorphous carbon, soot, carbon fibers, coal powder, active carbon powder, charcoal powder, and coke powder.

6. The process of claim 5 where the particulate carbon is powdered crystalline graphite.

7. The process of claim 6 where the powdered crystalline graphite has a particle size of minus 300 mesh.

The process of claim 6 where the plasma jet comprises helium or argon or their mixtures.

9. The process of claim 1 where the insert comprises a refractory material.

10. The process of claim 9 where the insert comprises carbon.

11. The process of claim 9 where the insert comprises pyrolytic carbon.

12. The process of claim 9 where the insert comprises zirconia, alumina, or thoria.

13. the process of claim 1 including the step of cooling the fullerene-containing stream before recovering the fullerenes from the fullerene- containing stream.

14. The process of claim 6 including the step of cooling the fullerene-containing stream before recovering the fullerenes from the fullerene- containing stream.