NbN is a strong pinning superconductor with a relatively high Tc.

NbN can be reactively sputtered by adding N₂ to the sputtering gas.

• Links to papers about the hysteresis loop of reactive sputtering come here
• Links to papers about reactive NbN sputtering go here
• Links to papers about physical properties of NbN go here

In the Z-400 2 recipes are in use for sputtering NbN: Gertjan uses 1.7 kV and blend, Olaf uses 1.0 kV and no blend. See the Z-400 page for details.

Details about sputtering NbN in the UHV system go here.

There is a systematic difference in X-ray and RBS thickness. Fitted composition makes sense and scatter yield make sense, it seems that (nuclear) collision crosssection and (electronic) stopping exponent are calculated OK. RUMP uses the wrong density, so the following procedure should be used:

1. Fit the thickness t in /cm2
2. Area nuclear density a = t * 1e15
3. Thickness = a/d where d = atomic density

The density of Nb is 8.57 g/cm², the density of Nb₁N₁ is 8.47 so the density of Nb_xN_1-x can be assumed to be 8.52 with an accuracy < 1% for x between 1 and say 0.45. Therefore, the following algorithm can be used for calculating the thickness (gnuplot):

atpercm2=471e15 # whatever RUMP fit gives
rho=8.52
n_a=6.022e23
m_nb=92.906
m_n=14.007

molmass(x)=x*m_nb+(1-x)*m_n
atpercm3(x)=n_a*rho/molmass(x)
cm(x)=atpercm2/atpercm3(x)
nm(x)=cm(x)*1e7

set term postscript color
set xrange [0.45 to 1]
set title "NbxN1-x thickness sample 'Gertjan2'"
set xlabel "composition"
set ylabel "thickmess (nm)"
plot nm(x)

The result is shown below:

So for the fitted composition Nb_0.6N_0.4 the thickness is 55 nm (57 X-ray)