The Journal of
Physical Chemistry VOLUME 97, NUMBER 27, JULY 8,1993
0 Copyright 1993 by the American Chemical Society
LETTERS Thermogravimetric Analysis of Carbon Nanotubes and Nanoparticles Louis S. K. Pang,' John D. Saxby, and S. Peter Chatfield CSIRO Division of Coal and Energy Technology, PO Box 136, North Ryde, NSW 2113, Australia Received: March 26, 1993; In Final Form: May 10, I993
The oxidation of carbon nanotubes and nanoparticles has been studied by thermogravimetric analysis (TGA) in air. The maximum rateof weight loss took place a t 695 OC at a heating rate of 1 OC/min. This result shows that the nanotubes and nanoparticles are more resistant to oxidation than other forms of carbon (diamond, soot, graphite, and c60) studied previously under identical conditions.' The TGA of the nanotubes/nanoparticles in argon showed no weight change or detectable thermal transformation up to 1000 OC.
Introduction Carbon nanotubes have been identified by IijimaZand isolated in largequantities by Ebbesen and Ajayam3 Carbon nanoparticles and onion structures4Jwere also reported later. Recently, Raman studies were reported on these nanoscale carbon materiah6The nanotubes can be filled with lead by capillary suction' and can beusedasacatalyst support.8 It isnowknownthatthesematerials can also be prepared from coal.g The oxidation of fullerenes c 6 0 has been reported previ0 u s 1 y ' J ~and ~ ~compared with the oxidation of diamond, graphite, and soot by the method of thermogravimetric analysis (TGA).' We now report the TGA of carbon nanotubes and nanoparticles in air and in argon. As the structures of these nanometer-sized carbon materiaPs are uniquely different from any other form of carbon, a comparison of oxidation behavior is most informative.
Experimental Section Nanotubes and nanoparticles were prepared by the carbon arc method described by Ebbesen and Ajayan3 A dc arc plasma was formed between a 12" graphite negative electrodeand a 8-mm positive electrode in 900 Torr of helium. The hard deposit rod formed on the negativeelectrode was mounted in epoxy resin and dissected. As reported earlier, the rod consisted of a hard outer shell and a soft inner core.3 The inner core consisted of alternate sectionsof a dull blacksoft matrix and shiny sections of featherlike materials. Thedull blacksoft matrix wascollected and~onicated~ in 2-propanol. The colloidal suspension was pipetted off, leaving 0022-3654/93/2097-6941$04.00/0
a residualprecipitate. Transmission electron microscopy showed that both fractions were predominantly nanotubes and nanoparticles. Therefore, the fractions were combined and used in the following studies. Thermal analyses were carried out in a thermogravimetric analyzer controlled by a Rigaku TAS 100 data station. A g proximately 5 mg of powdered sample was heated in an open platinum pan in excess air or argon (100 mL/min).
Results and Discussion Results are shown in Figure 1 for purified C a and a typical graphite, as well as for the mixtureof nanotubes and nanoparticles. Since our earlier work,' we have carried out TGA measurements on several batches of Car together with a range of partially oxidized products and insoluble residues. Because of possible decomposition and oxidation at low temperatures,lZJ3 it is clearly difficult to obtain and measure properties of C60 freeof extraneous material. Figure 1 shows the c60 result in which we have most confidence, but we emphasize that the temperatures of maximum rate of weight loss (Le., oxidation) can vary by up to f20 O C depending on sample preparation and treatment. It is clear that the nanotube/nanoparticle sample is more resistant to oxidation than either CSOor graphite. The temperature (f5 OC)ofthemaximum rateofoxidation (peakin thedifferential curve) is 695 OC for the nanotube/nanoparticle sample, 420 OC for C a , and 645 OC for graphite (Figure 1). The other forms of carbon studied earlier' also exhibit significantly lower tem0 1993 American Chemical Society
Letters
6942 The Journal of Physical Chemistry, Vol. 97, No. 27, 1993
0.02
E
0.01
E"
w
.- 60
c
-9
o
.-g
-0.01
'5 0
Nano-Particles
2o0
1 -0.02
-0.03
100
z5
0.25 g .E
.E"
80
'1 .-
"
60
0
-
:.-
.-F
40
Graphite
-0.25
-0.5
0
- 0.02
100
2
30 min in the presence of lead.' After this Letter was submitted, an article by Ajayan et al. appeared reporting similar oxidation results.15 It is apparent from Figure 1 that any initial weight increase due to addition of oxygen atoms to the carbon structure is small (