SUMMARY

Fullerenes are a recently discovered allotrope of carbon that have been found to possess unusual properties, some of which may be ideal for methane activation. This project is designed to evaluate these carbon-based materials for conversion of methane into higher hydrocarbons. The project is divided into three technical tasks. Task 1 deals with synthesis and characterization of the fullerenes and fullerene soots, Task 2 with testing of the catalysts, and Task 3 with evaluation of the results and technical reporting. The results and accomplishments for this quarter are summarized below.

Task 1. Synthesis and Characterization of Fullerenes

Reconstituted fullerene soot was prepared by adding Qjo to a toluene-extracted Terrasimco Nlerene soot.

Kdoped fullerene soots of different potassium concentrations were prepared.

Fullerene soot doped with cesium was prepared by addition of cesium carbonate.

Task 2. Testing of Catalysts

The reconstituted fbllerene soot was tested for methane activation. Negligible Merences in activities or selectivities were found between this modified soot and other modified soots (toluene-extracted soot, C@-activated soot, and fullerene soot).

K-doped soots at different K concentrations were tested for methane activation. The K-soots with higher K concentration (12 and 24 wt%) were found to have higher activities and selectivities to C2-Q hydrocarbons than the 6% K-soot.

K-doped soot was exarnined for methane activation in the presence of co-feeds of either ethane or ethylene. A higher selectivity to C3 and Q hydrocarbons was observed for the R-soot than when these reactions were conducted using soot as a catalyst, or when thermally induced.

The Csdoped soot was tested for methane activation. The selectivity for hydrocarbons from the methane conversion catalyzed by this soot was found to be higher than catalyzed by fullerene soot but lower than by K-soot.

1

PROJECT OBJECTIVES

GENERAL OBJECTIVES

Methane is one of the most abundant sources of energy and is found naturally in underground reservoirs and as a by-product of indirect liquefaction processes. Although methane is useful as a fuel, it is not easily stored or transported, and for that reason, the efficient direct conversion of methane to higher hydmcarbons is essential to p v i d e an economical alternative energy source. However, because the C-H bond of methane is stronger than that of the higher hydrocarbons, high pyrolytic temperatures are needed, and the products tend to rapidly polymerize to coke and unwanted hydrocarbons. 12

This project is designed to evaluate the feasibility of using fdlerene materials as methane activation catalysts. Fullerenes are a new allotrope of carbon consisting of closed shells of sixty or more atom.3 The full scope of the reactivity of these novel materials is not yet known. However, SRI and others have demonstrated that fullerenes have unique properties, including the ability to stabilize methyl radicals, shuttle H atoms, and act as electrophilesp5 Fullerenes have been found to act as "radical sponges" that readily accommodate organic radicals. Thus we expect that fdlerenes or fullerene-based catalysts may be ideal for methane activation, and since these catalysts are easily produced in soot, they can potentially be inexpensive catalysts and make the direct conversion of methane into higher hydrocarbons inexpensive and environmentally sound.

In this project, novel fullerene-based catalysts are being synthesized and examined for their ability to convert methane into olefins and other higher hydrocarbons. They will be examined using a short-contact-time reactor to minimize any by-product formation due to free-radical polymerization reactions. The primary objectives of this project are to synthesize and examine the reactivities of fullerene-based catalysts and to develop an understanding of these catalysts in terms of hydrogen activation, polymerization of methane into higher hydrocarbons, and minimizing of coke formation.

PROJECT OBJECTIVES FOR THIS QUARTER

We had three objectives for this quarter. The first objective was to compare modified soots prepared from a single source in order to determine the effect of variations in surface area and

2

soluble fullerene content. Last quarter we compared soots that were modified by solvent extraction and Ca activation, and observed negligible differences in their activities and selectivities. We completed this study this quarter by testing the catalytic activity of reconstituted soot (where Qjo was added to the soot).

The second objective for this quarter was to determine the effect of K doping concentration on the activity of K-doped soot. In attempts to study this effect, we synthesized soots doped with different K concentrations and compared their effects on the activities and selectivities for methane activation. We are also in the process of comparing the effect of K-soot with other alkali promoted soots.

The third objective for this quarter was to study the effect of adding co-feeds of ethane and ethylene during the methane activation experiment, and comparing the effects of soot and K-soot on the activities and selectivities of the products. The interest on the addition of co-feeds of ethane and ethylene is two-fold. The first interest is to determine if the promoter can reduce the concurrent decomposition of ethylene during the methane activation experiment, as suggested in the literature, and how to optimize the reaction conditions. The second interest is in the use of co- feeds of ethane or ethylene as sensitizers for the methane activation reaction. Thus we want to investigate these co-feeds to improve the economics of methane upgrading process.

3

TASK 1: PREPARATION OF CATALYSTS

The objectives of this task are to synthesize and characterize fullerenes, fullerene soots, and catalysts based on these materials. One of the novelties of these fullerene preparations is that other atoms can be incorporated into the structure: inside the cage, outside the cage, or within the framework itself. During this quarter, we continued to study the effect of modifying a fullerene soot (from Terrashco, Inc.) to compare differences in fullerene content and surface area on a soot prepared from it single batch. We also continued to examine the unusual effect of alkali metal incorporation on soot. We prepared a fullerene soot (from Terrasimco) with various concentrations of potassium carbonate, and for comparison, we also prepared a soot doped with cesium carbonate.

MODIFICATIONS OF SOOT

In previous work we had observed an apparent enhancement in reactivity when Qo was added to an extracted soot. However, because this soot was prepared from a commerciai source different from that for the other modified soots, a valid comparison could not be made. During this quarter, we prepared and compared a fullerene soot, a high surface area soot, an extracted soot, and a reconstituted soot (where (&o was added to an extracted soot), where dl materials were prepared from a single commercial fullerene soot.

Extracted soot was prepared as follows. Under nitrogen, 3.00 g of the fullerene soot was stirred with 500 mL of toluene at room temperature for 24 hours. The mixture was then filtered and dried at 60°C for 24 hours to yield 2.5 1 g of extracted soot, or a loss of 16.3% of extractable material. Then 0.3 g of the & dissolved in 200 mL toluene was added slowly to 2 g of extracted soot with proper stirring. After overnight stirring, toluene was removed under reduced pressure, and 2.25 g of reconstituted soot was obtained.

PREPARATION OF DOPED FULLERENE CATALYSTS

We continued our efforts to prepare catalysts containing fullerene doped with metal. As reported in Quarterly Report 8, we observed a significant enhancement of selectivity and yields of C2 and higher hydrocarbons when we doped a soot (obtained from MER corporation) with 12% K2CO3. Since in previous work we found that soots from different sources sometimes showed differences in reactivity, we prepared the new doped catalysts using soot obtained from

4

Terrasimco, Inc., which we are now using as the standard soot. We prepared K-soot with concentrations of 6%, 12%, and 24% K2CO3. We also prepared a Cs-doped soot to see whether otha alkali metals have a similar effect on the catalytic activity of the soot (the concentration of the Cs2CO3 was adjusted to be equivalent to the 12% K2CO3 loading on a molar basis).

As described in Quarterly Report 8, the desired amount of QCO3 or Cs2CO3 was dissolved in 100 mL of a water:methanol9: 1 mixture, and this solution was slowly added to 4 g of Nlerene mot with proper stining. The mixture was stirred, the solvents were evaporated under flowing nitrogen, and the product was dried under vacuum to remove the residual solvent to give the respective alkalidoped soot.

5

TASK 2: METHANE ACTIVATION

The objectives of this task are to evaluate and compare the soots and promoted soots listed in Table 1 for their reactivities, selectivities, and coking propensities for methane activation. As described in Quarterly Report 8, the extracted soot, Co2;activated soot, and fullerene soot do not show significant differences in activities and selectivities. The extracted soot was prepared by removing the soluble fullerenes (i.e., &) from the fullerene soot. Although this process removes the soluble fullerenes, it also may change the physical properties of the fullerene soot. For instance, the surface area of the material increases from approximately 125 m2/g to 185 m2/g, thus it does not necessaxily only compare the effects of the soluble fullerenes on the methane activation experiment. Thus to complete this study, a reconstituted soot (Qjo added to an extracted soot) was tested for methane activation. The results of this test, compared with previous results in Figure 1, suggest that the differences in activity and selectivity between these pretreated soots and fidlerene soots are negligible within experimental error and that the difference between pretreated soots that we observed previously can be attributed to the properties of soots produced from the different commercial sources.

Table 1

SOOTS INVESTIGATED DURING THIS QUARTER

Catalyst Description

Reconstituted soot 15% c60 added to toluene-extracted Terrasimco fullerene soot

K-SOOt

cs-soot

Terrasimco fullerene soot doped with different amounts of K2C03

Terrasimco fullerene soot doped with Cs$Q

The flow rate studies and ethylene cu-feed experiments described in the previous reports show that the low selectivity for methane activation catalyzed by fullerene soot is due to secondary reactions such as the decomposition of the C2 products. Thus these fullerene-based reactions are not "reactivitytt limited, but rather limited by subsequent reactions. Efforts to reduce these secondary reactions, which lead to coke, are essential. One approach to limiting these secondary reactions is to modify the catalysts by some agent that can block or limit the C2 decomposition reactions.

6

40 I I I I

A

$? v

5 30 UJ W > 0 0 UJ

a

20

10 E %!

0

a Fullerene Soot

CO, Activated Soot

Extracted soot

0 C60 + Extracted Soot

1 Fullerene Soot >-

i- 2 I - 8 0 El Extracted soot w a 6 (13 Z g 4 U 0 2 2

C02 Activated Soot

0 c60 + Extracted Soot

a

n >. I N O

800 900 1000 TEMPERATURE (“C)

(b) Selectivity

600 700 800 900 TEMPERATURE (“C)

(a) Methane conversion

1000

h

Y $? 10 I I i 1

Fullerene Soot >- i- 2 I - 8 0 El Extracted soot w a 6 (13 Z g 4 U 0 2 2

C02 Activated Soot

0 c60 + Extracted Soot

a

n >. I N O

800 900 1000 TEMPERATURE (“C)

(b) Selectivity CAM-4069-20

7

Figure 1. Methane conversion and selectivity as a function of pretreatment for Terrasimco soot.

A recent paper showed that the presence of potassium on platinum can decrease the ethylene decomposition reaction and thus may be useful in enhancing the selectivity in our methane activation reactions.6 In a related studies, workers found that enhanced catalytic behavior was observed toward gasification reactions when the alkali metals K2CO3 or Cs2Ca and carbon were heated together at 5OOOC and above; an effect which was attributed to the alkali metal-carbon interactions.7 Thus we hope that the doping of fullerene soot with these alkali metals may provide a path to reducing the secondary decomposition reactions.

Last quarter we reported an unusual reactivity and selectivity for K-soot when the methane activation experiment was conducted for an extended period at 950°C at methane flow rates of 150 mI.imin. Under these reaction conditions, both fullerene soot and potassium-doped acetylene black showed a slow decrease in the methane conversion until after approximately 15 hours the reaction had stabilized. However, in the case of the potassium soot, the decrease in methane conversion was accompanied by a concurrent increase in C2 hydrocarbons. Thus the K-soot shows an increase in selectivity to C2 hydrocarbons from approximately 5% to 55%, while the fullerene soot and K-acetylene black show only a maximum selectivity of 20%. Furthermore, in addition to the C2 production, the K-soot also produces significant quantities of C3-C4 hydrocarbons, so that its overall selectivity to useful hydrocarbons under these conditions approaches 70%. The soot for these reactions was obtained from MER Corporation, and since we are now trying to conduct all our experiments using the same soot, during this quarter we investigated these reactions using K- doped soot prepared using soot obtained from Terrasimco, Inc.

The K-doped soots (using Temasimco soot) tested during this quarter were doped With 6%, 12%, and 24% K2CO3. Their activity for the methane activation reactions was determined by GC analysis, using both a TCD detector and FID detector. The TCD detector allowed us to determine a more careful mass balance using argon as an internal standard During the 12% K2C03 run, however, the TCD detector could not be used and the selectivity could not be accurately . determined. According to the data using the FID detector, the activity and selectivity of the 12% K2CQdoped soot was identical to that of the 24% K2COydoped soot, although they are not shown in the following figures.

The results are shown in Figures 2 and 3 for the methane conversion and selectivity, respectively. At 950°C, the methane conversion stabilizes at 3% for the 24% doped K-soot and at approximately 6% for the 6% doped K-soot. For comparison, the fullerene soot stabilized in the range of 6% conversion under these conditions. As shown in Figure 3, the selectivity of the 6%

8

Flow Rate = 150 mUmin 20

A

6% K-Soot s z ' 6 0 0 24% K-Soot iz

Y

E 12 > z 0 O 8 u z a f : 4 3

0 0 5 10 15

TIME (hours) HELD AT 950°C CAM-4069-21

Figure 2. Effect of K doping concentration on methane activation. Methane conversion as a function of time.

9

C2 Hydrocarbons C, Hydrocarbons

Y n Total Selectivity

h > A A A g - 3 20

i

(a) 6% K-soot

-60 s Y

w' 20

0

0 C2 Hydrocarbons a C3 Hydrocarbons

(b) 24%K-S00t

80 1 I I

I I I

5 10 15 0 ' 0

TIME (hours) HELD AT 950°C (c) Fullerene soot

CAM-4069-22

Figure 3. Effect of K doping concentration on methane activation. Selectivity as a function of time.

10

K-soot stabilized at 45%, while the 24% K-soot showed selectivities to C2-C4 hydrocarbons of about 70%, which is similar to the result described last quarter for the 12% K-soot prepared from the MER soot. In contrast, the fullerene soot exhibited a selectivity of only 20%. Thus after stabilization, for a given conversion, the K-soot clearly gives a higher selectivity. As we obtain more data we will plot the conversion vs. selectivity information for these various catalysts.

As previously mentioned, a possible effect of the potassium loading would be to decrease the decomposition of ethylene. Thus we tested the effect of introducing ethylene and ethane during the methane activation catalyzed by the K-soot. This experiment serves two purposes. The first purpose is to determine if the decomposition of ethylene or ethane is indeed reduced, and the second purpose is to test if the addition of these cefeeds will increase the conversion of methane into useful hydrocarbons as shown in previous work.7~8 The results of this study are shown in Figures 4 and 5. As shown in Figure 4, for the K-soot catalyzed reaction, the ethylene concentration remains unchanged until above 700°C. In contrast, in the presence of fullerene soot, the ethylene begins to convert at about 500°C. At 800°C, 12% of ethylene has reacted in the presence of K-soot compared to 37% of ethylene when in contact with the fullerene soot. As shown in Figure 5 , the yield of C3 and Q hydrocarbons for the methane/ethylene activation reaction catalyzed by K-soot is significant, with the maximum amount of higher hydrocarbons observed around 85OOC. These higher hydrccarabons are not observed for the fullerene soot catalyzed reaction. This observation is consistent with the suggestion that K doping can decrease the ethylene decomposition (or secondary decomposition reactions) and thereby enhance the selectivity to C3 and C4 hydrocarbons. As seen in these figures, the presence of ethylene appears to have accelerated the methane conversion more for the fullerene soot case than for the K-soot case. However, at the same time, more useful hydrocarbons (C2-C4) were produced for the K-soot catalyzed reaction than for the soot catalyzed reaction.

Figure 5 illustrates the effect of K-doping in methandethane conversion. At 800°C, 69% ethane conversion is observed for K-soot catalyzed reaction, and a nearly quantitative conversion (> 90%) of this reacted ethane can be accounted for as ethylene. About 80% ethane conversion was observed for the reaction of methane/ethane catalyzed by fullerene soot, but only 58% of the expected ethylene yield was obtained (assuming all the reacted ethane was converted to ethylene). Thus this increase in selectivity caused by K-doping might provide us a way to apply the high activity of fullerene soot toward hydrocarbon dehydrogenation with less coking than obtained by current methods. In future work we plan to combine Kdoped soots with transition metals of this promoter to further modify the catalytic activity of fullerene soot.

11

2

(a) Change in gas composition

I I I I

@) Formation of C3 and C4 hydrocarbons

L

-41 I 1 1 1

500 600 700 800 900 TEMPERATURE ("C)

(c) Conversion catalyzed by fullerene soot CAM-4069-23

Figure 4. Effect of ethylene on methane conversion catalyzed by 24% K-soot.

12

UI' 0' 0

n c) Y

(0 0 0

CHANGE IN GAS COMPOSITION ("3) A 0

FORMATION OF C3-C4 HYDROCARBONS ( " 3 ) CHANGE IN GAS COMPOSITION ("A)

0 0 0 I 4

The interactions between carbons and alkali metals have been the subject of some interesting work in the subject of coal gasification. For instance, Wood and Sancierg reported the presence of h e radicals when a mixture of carbon black and certain alkali metal salts were heated to high temperams in helium. Among the alkali metal carbonates, the magnitude of the effect decreased in the following ordec Cs2Ca > Li2Ca > K2CO3 > Na2C@ > no additive. Prompted by this order of reactivity, we also prepared and tested Cs-doped soot to compare the activity with that of K-soot. The atomic metal content of this Cs-soot is the same as that of 12 wt% K2C03doped soot. The result shown in Figure 6 suggests that the selectivity for hydrocarbons is lower than that of the corresponding K-soot (40% versus 70%). To further understand the trend in alkali metal soot, we will test Li2COydoped soot in the next quarter.

Conclusions

During this quarter we focused on three topics to study methane activation on fullerene based catalysts. The first topic was to determine if there were any reactivity differences observed from fullerene soots obtained by different methods or modified with physical properties. We found that although there were some differences between soots obtained from different sources, they were not significant, and the reactivity was not a function of either surface area or extent of extractable fullerenes. The second topic was that of exploring the apparent alkali metal effect on the activity and selectivity of the fullerene soot. We have not yet completed our study, but it is apparent that for a given level of conversion, the K-soot is more selective for C2-C4 hydrocarbons than is the fullerene soot itself. However, there appears to be a periodic trend for the alkali promotion, and we have not yet found if the potassium promotion is the optimum, and if it can be used in combination with transition metals to further enhance the catalyst performance. The third topic was to explore the use of sensitizers such as ethane and ethylene to further enhance the activities and selectivities of these catalysts. These sensitizers have different effects on the promoted and non-promoted soots with activity and selectivity and it is not obvious at the moment if they will provide any advantages for methane activation at this time. However, we do see a significant advantage for dehydrogenation reactions using the potassium promoted soots. During the next quarters we will continue to study the effects of promotion of these fullerene soots with the goal of developing a facile methane-to-higher-hydrocarbons process.

14

1 . 1 . 1 1 0 2 4 6 a 10 12

TIME (hours) HELD AT 950°C (a) Methane conversion as a function of time

0 Total Selectivity

0

20

0 0 5 10 15

TIME (hours) HELD AT 950°C (b) Hydrocarbons selectivity as a function of time

CAM-4069-25

Figure 6. Methane activation catalyzed by Cs-soot.

1.

2.

3.

4.

5.

6.

7.

8.

9.

REFERENCES

D. McMillen and D. Golden, Ann. Rev. Phys. Chem. 1982,33,493-532.

J. Cox and G. Pilchem, Thennochemistry of Organic and Organometallic Compounds (Academic Press, 1970).

R. E. Haufler, J. Conceicao, L.P.F. Chibante, Y. Chai, N. E. Byrne, S. Flanagan, M. M. Haley, S. C. O'Brien, C. Pan, 2. Xiao, W. E. Billups, M. A. Ciufolini, R. H. Hague, J. L. Margrave, L. J. Wilson, R. F. Curl, and R. E. Smalley, J. Phys. Chem. 1990,94, 8634.

E. Gretz, T. Oliver, and A. Sen, J. Am. Chem. SOC. 1987,109, 8109-81 11.

F. Wudl, A. Hirsch, K. C. Khemani, and T. Suzuki, "Survey of Chemical Reactivity of QJ, Electrophile and Dieno-polarophile Par Excellence," in Fullerenes, G. S. Hammond, and V. J. Kuck, Eds. (ACS Symposium Series 481; Atlanta, 1991, American Chemical Society, Washington, DC, 1992), p. 161.

R. C. Windham and B. E. Koel, J. Phys. Chem. 1990,94, 1489-1496.

A. Bossard and M. H. Back, Can. J. Chem. 1990,68, 1401-1407.

A. Bossard and M. H. Back, Can. J. Chem. 1990,68, 37-42.

B. Wood and K. Sancier, Catal. Rev.-Sci. Eng., 1984,26,233-279.

16