The first reflection is probably the (100), so
.
We then manipulate the equation for the interplane
distances:
h2 + k2 + l2 = (a/dhkl)2.
If we were right about the first reflection being the
(100) interplane distance, we should be able to
determine the Miller indices corresponding to each of
the other reflections by computing the
approximate sum of
their squares by the above formula and then finding a
consistent set of indices by trial-and-error:
dhkl (pm) |
221 |
179 |
156 |
140 |
h2+k2+l2 |
2 |
3 |
4 |
5 |
(hkl) |
(110) |
(111) |
(200) |
(210) |
Now that we know the Miller indices, we can calculate
a = dhkl(h2+k2+l2)1/2 from each reflection
and average the results.
(hkl) |
(100) |
(110) |
(111) |
(200) |
(210) |
dhkl (pm) |
308 |
221 |
179 |
156 |
140 |
a (pm) |
308 |
313 |
310 |
312 |
313 |
The average is therefore
.
The better method is to combine the equation for the
interplane distances with the Bragg condition. We get
Thus, a plot of
vs
2(h2+k2+l2)-1/2 should have a slope of a and an intercept near zero.
Note that the standard linear regression method
implemented in (e.g.) calculators assumes that the
abscissas are error-free while the ordinates contain
random errors so that the above-described plot is, of
all the possible ways of plotting the data, a
statistically correct choice.
(degrees) |
6.6 |
9.2 |
11.4 |
13.1 |
14.7 |
(pm) |
616 |
443 |
358 |
312 |
279 |
(hkl) |
(100) |
(110) |
(111) |
(200) |
(210) |
2(h2+k2+l2)-1/2 |
2 |
1.41 |
1.15 |
1 |
0.89 |
The plot is reasonably linear:
The slope is
.
The intercept of
represents small systematic errors in the measurements
which slightly inflate our original estimate of a.