Page 330 - Adsorbents fundamentals and applications

HYDROGEN STORAGE 315
◦
peak occurs between 900 and 1000 C. The desorption requires breakage of the
C–C bonds to free CO. TPD of chemisorbed hydrogen (from the edge sites) has
also been studied (Redmond and Walker, 1960), and even higher temperatures are
required. These facts add credence to the explanation of Baker/Rodriguez et al.
Yet, a number of groups have failed to reproduce the high hydrogen capac-
ities (Ahn et al., 1998; Strobel et al., 1999; Poirier et al., 2001). However, it is
interesting that, when compared with other carbon materials (activated carbon,
SWNTs, and MWNTs), GNF seems to yield the highest hydrogen capacities (e.g.,
Tibbetts et al., 2001). When expressed on a per surface area basis, GNF clearly
yielded the highest values.
In contrast, two groups have reported high capacities by following the work of
Baker/Rodriguez. Browning et al. (2002) reported 6.5 wt % hydrogen uptake at
12 MPa and 300 K. In their work, Fe/Ni/Cu catalyst was used for GNF growth.
◦
The samples were pretreated at 1000 C in flowing Ar for 36 h. They interpreted
the results by hydrogen dissociation (at room temperature) on the edge sites of
graphite to form H atoms which were subsequently adsorbed in the platelets.
Gupta and Srivastava (2001) reported ∼15 wt % uptake in GNF under similar
◦
conditions, with a pretreatment temperature of only 150 C in vacuo.
Alkali-Doped MWNTs. Alkali metals can intercalate graphite and form stable
intercalated compounds. It was reported that up to 0.137 L(STP)/g hydrogen
could be adsorbed between the layers of alkali-intercalated graphites (Wanatabe
et al., 1971; Akuzawa et al., 1990; Ichimura and Sano, 1992). Rao et al. (1997)
showed that doping by alkali metals resulted in change transfer to the SWNTs and
significantly increased the ionic character of the doped SWNTs. The charge in
the SWNTs can increase the hydrogen adsorption (Simonyan et al., 1999). With
this background, the report of Chen et al. (1999) that large amounts of hydrogen
were adsorbed in Li-doped and K-doped MWNTs at 1 atm pressure generated
considerable excitement.
The adsorption was measured gravimetrically by using a flow system at the
ambient pressure. The sample weight was monitored while the temperature was
varied. A weight gain of 20% was observed for the Li/MWNTs while the tem-
perature was lowered from 670 to 470 K. Likewise, the weight gain was 14%
for K/MWNTs when the temperature was near 300 K.
Because laboratory gas cylinders of hydrogen, even at the ultra-high purity
grades, are notoriously contaminated by moisture, Yang (2000) performed iden-
tical experiments with wet and dry hydrogen. With drying, the weight gains
were substantially lower, at near 2%. With wet hydrogen, the weight gains were
identical to those reported by Chen et al. The weight gains were caused by
the formation of Li(OH)·H 2 O and K(OH). The formation of alkali hydroxides
in these systems was further confirmed by Pinkerton et al. (2000). With strin-
gent dehydration, 1.3 wt % adsorption was measured for K/MWNTs, of which
0.2 wt % was cyclable. These amounts were measured at the ambient pressure.
Since these amounts are significant, the effects of alkali doping certainly merits
further investigation.