X-ray diffraction

K. Wolski
February 2004

Diffraction diagram

The figure above represents the diffraction diagram (or diffractogram) of a eutectic Pb- alloy.
Sn obtained by slow cooling from the liquid phase followed by cold rolling. The length
used for the analysis isλ = 1.78897 Å

NAME AND NATURE OF THE PHASES

Intensity
Sn

Sn

Pb

10 µm
Energy of photons [k eV]

Phase diagram microstructure (SEM) local chemical analysis (EDX)

We have the Pb-Sn phase diagram, the microstructure of the alloy, and the chemical analysis.
a locale of its phases.
• Justify the existence of two phases and discuss their chemical composition.

Corrected
Eutectic alloy 26.1 at%Pb and 73.9 at%Sn
During cooling, there is precipitation of the two phases:
almost pure tin
from a solid solution of tin (3%at) in lead (97%at)
The heavier the element is (mPb=207.2 g/mol, mSn=118.7 g/mol), the clearer it stands out in the micrograph.
MEB. The dark phase is dominant, it is tin.
PHASE Pb

It is known that the structure of Pb is cubic. Let a be the lattice parameter (edge length).
of the cube) and (h,k,l) the Miller indices of a crystallographic plane.
• Show that the interplanar distances dhklbetween two successive crystallographic planes can
to be expressed in the form:
d hkl= a 2/(h2+ k 2+ l 2 ) .
Corrected
CC
C b,
We consider a system of orthonormal axes, defined by three vectors. a, c.
To represent the cubic system, we can immediately assume that these three vectors are identical.
a= a . C
and of length
In an orthonormal axis system, a direction [hkl] is perpendicular to the plane (hkl).

The equation of a plane (hkl) passing through the origin is: h*x + k*y + l*z = 0 (1)
The equation of an immediately adjacent plane (hkl) is: h*x + k*y + l*z = a (2)
(we recall that this plane intersects the x, y, z axes at the points
(a/h,0,0), (0,a/k,0) and (0,0,a/l).
A line perpendicular to these two planes and passing through the origin is defined by the equation:
x= y = z (3)
h k l
This line intersects the two planes at the points (0,0,0) and (p,q,r).
the coordinates of this last point being to be defined.

The interreticular distance d will be obtained from the equation:

d 2 =(p−0)2 +(q−0)2 +(r−0)2 (4)
To find the coordinates of the point (p,q,r) belonging to both the plane (2) and the line (3),
we have three equations:
h*p + k*q + l*r = a (5)
p q
= (6)
h k
q r
= 7
k l
From (5), (6), and (7) we deduce:

p= a⋅h , q= a⋅k , r= a⋅l (8)

h2 +k 2 +l 2 h2 +k 2 +l 2 h2 +k 2 +l 2
By substituting (8) into (4) we get
a 2h2 a2k 2 a 2l 2
d 2= + + (9)
h2 +k 2 +l 2)2(h2 +k 2 +l 2)2(h2 +k 2 +l 2)2
from where:

d= a (10)
(h2 +k 2 +l 2)

• calculate the distances ofhklcorresponding to the diffraction peaks of Pb. The angles 2θ measured on the
The diffractogram peaks are at 36.517°, 42.373°, 61.479°, 73.614°, 77.504°, 92.527°, 103.899°, and
107.821°.

Corrected
We use Bragg's law, assuming that all peaks of Pb
corresponding to first-order diffractions (n=1, in Bragg's law 2*dhklsineθ = n*λ).
The wavelength used isλ 1.78897 Å

2θ di = λ [Å]
2⋅sin(θi )
36,517 2,855
42,373 2,475
61,479 1,750
73,614 1,493
77,504 1,429
92,527 1,238
103,899 1,136
107,821 1,107

• Determine the 'hkl' indices of all the peaks.

Corrected
It is recalled that in a cubic structured hkl= a , with a - lattice parameter (11)
(h2 +k 2 +l 2)
Preliminary question: what are the indices of the peaks corresponding to the angles of diffraction 2?θ weak?
Bragg's law tells us that forθ weak ones we obtain from high ones,
The formula (11) indicates that d is all the higher as hkl are low.
Should we conclude that the first peak corresponds to the (100) plane???

No, because the structural factor (see lessons 94-96) can lead to the extinction of this peak.
So, maybe the first peak corresponds to the (110) plane???
We have to see...

but first think and not forget to make the students think...

Oh yes, we have eight peaks with a single equation (11) with three unknowns (h, k, l) for each of these peaks.
so the problem is mathematically 'hard'.
You need to find a trick.

This trick involves reasoning about the ratios of interreticular distances,

on one hand the experimental reports,
On the other hand, the theoretical reports,
But to talk about relationships, one must favor a peak.

We will therefore make an assumption that the first peak on the left is the (001) peak.
and calculate the two series of reports for the entire set of crystal planes that we will take in
the order of decreasing interplanar distances:

plan
001
011
111
and so on.

The series of experimental reports related to the interreticular distance (d1=2,855 Å) corresponding to
the first pic (of which we do not know the clues) will be calculated once and for all:

2θ di = λ dI
[Å]
2⋅sin(θi) d1
36,517 2,855 1,000
42,373 2,475 0.866 (=2.475/2.855)
61,479 1,750 0.613
73,614 1,493 0.525
77,504 1,429 0.500
92,527 1,238 0,433
103,899 1,136 etc.
107,821 1,107 etc.
a
d hkl d hkl (h2 +k 2 +l 2)
1
The series of reports will be calculated from the formula = =
d001 d001 a (h2 +k 2 +l 2)
(02 +02 +12)
d hkl
We obtain: Plan
d001
001 1
011 0.707, and it's already getting stuck!!!
111
Indeed, the second experimental peak is relatively close to the first (exp. ratio = 0.866),
while our hypothesis forbids it (theoretical ratio = 0.707),
so the first experimental peak is not (001).

We will therefore make the hypothesis that the first peak on the left is the (011) peak.
and recalculate the theoretical ratios from the formula
a
d hkl (h2 +k 2 +l 2)
2
= =
d011 a (h +k 2 +l 2)
2

(0 +1 +1 )
2 2 2

d hkl
We obtain: Plan
d011
001
011 1,000
111 0.816, and it ’s not getting better!!!

We are going to make a third (and final) hypothesis that the first peak on the left is the (111) peak.
and recalculate the theoretical ratios from the formula
a
d hkl (h2 +k 2 +l 2)
3
= =
d111 a (h +k 2 +l 2)
2

(12 +12 +12)

d hkl
We obtain: Plan
d111
001
011
111 1,000
002 0.866, phew, it's getting better (see the comparison with the series of reports)
experimental), but it's not over.

It is now that we will have to respond to the 'question': 'Your plan 002 is ridiculous,' although we will have
asked the students to accept Bragg's formula with n=1.

The diffraction of order n on the planes (h k l), whose interplanar distance is equal to [Link], is equivalent to
the first-order diffraction on the planes (nh nk nl), whose interplanar distance is equal to dhkl/n"

dhkl dhkl
⋅sineθ =λ and it can easily be shown that
From Bragg's formula, it is inferred that2⋅ =dnhnknl
n n
Indeed, dnhnknl
= a = a =
dhkl

((nh) +(nk) +(nl) )n(h +k +l )

2 2 2 2 2 2 n
We can therefore use Bragg's law in the form2⋅dnhnknl⋅sinθ =λ , that is to say with n=1 and accepting the
style indices 002 associated with fictitious plans (002).

We return to the calculation of the ratios related to the distance between the (111) planes:
d hkl di
We obtain: Plan
d111 d1
111 1,000 1,000
002 0.866 0.866
012 0.775
112 0.707
022 0.612 0.613
003 0.577
122 0.577
013 0.548
113 0.522 0.523
222 0.500 0.500
 123 0.463
004 0.433 0.433

The two series of theoretical and experimental reports stick "not too badly", but some plans do not
did not lead to diffraction. Why?
Because of the structural factors, of course!

And now the scoop of this TD:

The diffracting planes are (111) (002) (022) (113) (222) (004)…
Is there a particularity on the side of the indices???

The expected answer is YES!, they are of the same parity,

but in general she does not arrive right away.

In any case, we index the peaks from left to right: (111) (002) (022) (113) (222) (004).

• Calculate the lattice parameter 'a'

We take, for example, the first peak for which we know both the indices h=1, k=1, l=1, as well as the
interplanar distance d111=2,855 and we apply the formula:
a=d hkl ⋅ (h2 +k 2 +l 2)=d111⋅ (12 +12 +12)=2,855⋅ 3=4,945 [Å]

• Is it a body-centered (bcc) or face-centered (fcc) structure? To answer this question, one can
either use the cc and cfc structure factors, or calculate the density of Pb knowing its mass
atomic is 207.19 and that Avogadro's number is NAequals 6.023 times 1023at/mole.

For structural factors see the course

For the density, we will use the formula:

4⋅207.19⋅1.6604⋅10−24 g
ρ = m= =11.38 ]
V (4,945⋅10−8) 3 cm3

with 4since there are 4 atoms per cell in a fcc and

1.6604 x 10^-24 g is the atomic mass unit.
which led to a reasonably good density for Pb,
its structure is therefore of cfc type.

PHASE Sn

This second exercise should be considered optional, depending on the available time.
After indexing the Pb peaks, we will assume that the others belong to the Sn-rich phase.
assimilated to pure tin.
You will note that the first two peaks have already been indexed.

• Show that the structure of Sn is not cubic (the distances dhklcorresponding to the Sn peaks are
2.9155, 2.793, 2.062, 2.017, 1.659)

We return to the calculation of the ratios related to the distance between the planes corresponding to the first peak, namely (002):
a
dhkl (h2 +k 2 +l 2)
2
= =
d002 a (h2 +k 2 +l 2)
(02 +02 +22)
dhkl di
We obtain: Plan
d002 d1
002 1,000 1,000
0.958 (=2.793/2.916)
012 0.894
The second experimental peak gives a ratio of 0.958, while the nearest theoretical peak (in
The hypothesis of a cubic structure corresponds to the ratio 0.894.
The structure of SN cannot, therefore, be cubic.

• The structure of Sn is actually tetragonal. Knowing the hkl indices of the first two
pics (cf. Fig. 1), determine the parameters a and c of the tetragonal structure, where:
d h2kl= 1 (12)
h 2 +k 2l 2
+ 2
a2 c

Since Santa Claus has indexed the first two peaks and we know the interplanar distances
associée, on déduit les valeurs de a et de c de la formule (12) ci-dessous :
d 2200 = 1 =2,9162 etd1201= 1 =2,7932
22 +0202 12 +0212
+ 2 +
a2 c a 2c 2
where: a = 5.831 Å and c = 3.182 Å

Attention in this structure of200is not equivalent to d002 .

• Index the following peaks

For each peak, we have an interreticular distance from the spectrum and an equation (12).
The three indices are to be determined from this one and only equation.
Impossible, say the mathematicians!

For the third peak:d h2kl= 1 =2,0622from: 0.125*(h2+k2) + 0.42*l2= 1

h2 +k 2 l2
+
5,83123,1822
and we guess: h=2, k=2 and l=0.

For the fourth peak:d h2kl= 1 =2,0172from: 0.11965*(h2+k2) + 0.40180*l2= 1

h2 +k 2 l2
+
5,83123,1822
and we guess: h=2, k=1 and l=1.

• Do the indices of the diffracting planes correspond to a well-defined rule? Which one? The structure,
Is it centered or face-centered?

Yes.
h+k+l is even.
It is a centered structure.

• Calculate the density of Sn knowing that its atomic mass is 118.69 and compare it to the value.
theoreticalρ 7.287 [g/cm]3Discuss the results obtained.

2⋅118.69⋅1.6604⋅10−24 g
ρ = m= =3,643 ] , instead of the theoretical value equal to 7.287 g/cm3.
V(5,831⋅10− ) 8⋅3,182⋅10
2 − 8 cm3

In fact, the structure of tin is more complicated than you thought.

It is a tetragonal centered structure of type A5, i.e. the motif consists of four atoms positioned in
(0,0,0), (1/2, 1/2, 1/2) as in all centered structures and two 'intruders'
["(0, 1/2, 1/4), (1/2, 0, 3/4)."]
To realize this, it would have been necessary to calculate the structural factors…
With the 4 atoms, the theoretical density is found to be equal to 7.287 g/cm.3.