All A P C V D films in this study were exposed to air prior to their analysis by X P S . The film s collected carbon dioxide and other contaminants on the surface. B eca u se X P S is a
purely surface technique (top ten atomic layers interrogated), these contaminants show up and affect the com position analysis o f the film s. An X P S spectrum o f the surface o f a representative vanadium nitride film formed in this study is show n in figure 3.22a with a com position analysis in table 3.22b.
Figure 3.22a X P S elem ental scan from surface (prior to ion gun use).
SURVEY C:\V G \G A R ETH \S207D P\SU RV EY \1 .D TS
CAE - 6 0 eV S T E P - 6 0 0 .1 0 9 m eV S C A N S - 1 TIME - 3 m 3 .4 0 s Al K -alpha L arg e A rea XL D epth Profile - O s etching
V 2s AM 11-02-1998 X P S ANALYSER S O U R C E LABEL 1 0 0 — O 1s 80 V2s1 60 40 20 1 0 0 0 8 0 0 6 0 0 4 0 0 2 0 0 0
Table 3.22b C om position calculated from figure 3.22a X PS scan.
Elem ent Chem ical Shift Elem ental % (eV ) present in bulk Vanadium 5 1 3 .7 1 + 5 1 5 .8 24.6
O xygen 529.57 + 531.77 25.5 N itrogen 396.59 + 398.54 26.4 Carbon 284.25 + 286.27 24.5
Chlorine - 0
The com positional analysis from X PS enabled a determination o f the percentage elem ental abundance in the film. This was obtained by integrating the area under each o f the XPS peaks and applying a significance factor for each elem ent. The XPS was calibrated against internal machine standards and specific m etal nitrides (TiNi.o and VNi.o, Alpha Chem icals).
The chem ical shift values in table 3.22b are consistent with the values expected from a highly contam inated metal oxynitride surface with absorbed carbon dioxide and elem ental carbon (from the high vacuum pumps). For all the elem ents described two peaks were observed, one due to elem ents in the bulk material and the other due to oxidised elem ents on the top m ost surface layers.
T o overcom e this problem with XPS com positional information, an argon sputtering gun w as used to effectively strip o ff the film s’ surface, layer by layer, until the com position measured was self-consistent and the com position o f the bulk revealed. The argon sputtering gun that was used w asn’t particularly pow erful, rem oving only a few nanom eters on each etching. A depth profile, show ing how the com position o f a representative sam ple prepared from VCI4 and N H3 changed after around 45 minutes o f etching is show n in figure 3.22c.
Figure 3.22c Depth profile showing composition change after 45 minutes o f etching. 60 K e y □ O 1 s L'_: v 2p3 C ] n i s c 1 s 5 5 5 0 4 5 4 0 3 5 3 0 2 5 20 200 4 0 0 6 0 0 8 0 0 1000 Time (Seconds) 1200 1 4 0 0 1 6 0 0 1 8 0 0
Another feature o f XPS is that the binding energies o f electrons change depending on w hich atom they com e from, and the local chemical environment o f that atom. For exam ple, an electron ejected from surface bound oxygen w ill have a different environm ent and hence binding energy to an electron emitted from an o x y g en atom in the bulk. The largest shifts in binding energy are due to changes in oxidation state o f the elem ent in question. A 2p3 electron from a vanadium atom w ill be slightly harder to rem ove if the vanadium has a +5 oxidation state (V O N ) than if it has a +3 (V N ) or zero (vanadium m etal) oxidation state, due to the electrostatic force. Photoelectrons from atoms in a more electropositive state are emitted with less kinetic energy (and therefore have greater binding energy) and vice-versa for atoms w hich are more electronegative. B ecause the kinetic energy (Vimv2) and frequency (v) o f an excited electron can be m easured, binding energies can be calculated by E instein’s equation (equ. 3.2a)
Oxidation states, and even specific environments (eg whether the carbon is bound to oxygen ), may be determined by comparing the binding energy o f the peak with known binding energies, obtained from standards. An exam ple o f how the binding energy can change is shown in figure 3.22d.
Figure 3.22d X PS chem ical shift show ing change in vanadium peak position and intensity at various stages o f etching
<n a. o JC 5 0 4 5 2 2 5 1 6 5 1 4 5 1 0
In the exam ple given the peak in the background is due to a surface bound vanadium. A s the surface is etched away by the argon ion gun the vanadium peak shifts to the right as its binding energy changes. In the foreground, the peaks are due to vanadium atoms in the bulk which are bound to nitrogen and oxygen.
Once the com position o f the film has levelled o ff in a consistent manner, a final scan o f the w hole XPS range can take place. The final scan for a film prepared from VCI4 and NH3 (fig. 3.2e) and the resultant com position (fig. 3.2f) are shown below . It can be seen from this plot that there are no elem ents present other than vanadium, nitrogen, small amounts o f oxygen and trace amounts o f carbon and chlorine.
k
C
P
S
Figure 3.22e XPS scan of a VN film, following 45 minutes of ion gun use.
200 V2p3+V2p1 141 12( 101 (8 __ 8 0 60 0
_
4 0 20 200 4 0 0 600Binding Energy (eV) 8 0 0
1000
Figure 3 .2 f C om position calculated from figure 3.22e scan.
Elem ent Chemical Shift (eV ) Elemental % present in bulk Vanadium 513.06 4 9 .6 Nitrogen 397.08 4 0 .4 0 O xygen 530.78 9.99 Carbon - 0 Chlorine - 0
An X P S chem ical shift o f 513 eV is half way between the expected result for vanadium