You are on page 1of 11

University of Pennsylvania XR.

Bragg Reflection of X-Rays

Objectives

To use a crystal spectrometer to investigate the X-rays produced when an electron beam strikes a
metal target; to use X-rays of known wavelength to measure an unknown crystal lattice spacing.

Apparatus

TEL X-Ometer tabletop X-ray machine, NaCl crystal, LiF crystal, NaCl powder sample,
universal digital clock/counter, TEL-Atomic scaler, picoammeter.

WARNING: The apparatus is perfectly safe if used correctly. Do not attempt to open the
housing when the X-rays are on, or to disable the safety interlock in any way!

X-ray tube

Crystal post

Geiger Tube

Figure 1: TEL-X-Ometer Table Top X-Ray machine. Top view with lid closed
University of Pennsylvania XR.2

Introduction

The X-ray part of the electromagnetic spectrum is radiation with a wavelength of about 0.1 nm
(1 Angstrom or 10-10 m). That is about five thousand times smaller than the wavelength of
visible light. In frequency units, X-rays are in the 1018 Hz range. They have energy on the order
of 1-100 keV. The object of the experiment is to analyze X-rays produced by two different
physical mechanisms with the aid of a crystal spectrometer (the Tel-X-Ometer) whose
functioning depends on Bragg's Law of X-ray diffraction. We will then use the same X-rays to
measure the lattice parameter of an unknown sample.

X-rays are produced in a laboratory (or a dentist's office) by electrostatically accelerating


electrons produced by thermionic emission from a hot filament. The beam of accelerated
electrons hits a solid target. There are many choices for the target material, but in our case the
target is copper. When the electrons collide with the atoms of the target they produce a
continuous X-ray distribution spread out over a range of frequencies much like the black-body
spectrum. However, they also produce a discrete line spectrum characteristic of the target
material. The discrete spectrum is superimposed on the continuous spectrum.

Continuous Spectrum

As soon as the electrons enter the target they are rapidly decelerated. Most of their energy goes
into random vibrational motion of the target atoms (heat!). Since accelerating charged particles
produce electromagnetic radiation, a small fraction of the energy is used to produce photons. It
is clear that there is an upper limit to the energy of the photons produced, since an electron can
give up at most 100% of its energy to a single photon. On the other hand, there is no lower limit
to the energy of the photons, since some of the electrons will slow down by a series of gentle
collisions, in which they will radiate only a small fraction of their energy. Thus, the frequencies
of the radiated photons form a continuum ranging from essentially zero up to a maximum given
by the Planck relation:
h max  Ee  eV (1)

Here E e is the energy of an electron in the beam and is equal to (e)(V), where V is the potential
difference through which the electrons are accelerated and e is the charge on an electron. A
maximum frequency implies a minimum wavelength

c hc
min   (2)
 max eV

The continuous part of the X-ray spectrum contains all wavelengths greater than  min and  min
depends on the accelerating voltage V . Thus, knowledge of  min along with the speed of light
constant yields the ratio of fundamental constant, h e . You may be interested to know that the
University of Pennsylvania XR.3

modern method of measuring h e , using the “Josephson effect,” a property of superconducting


materials, was pioneered at the University of Pennsylvania by Langenberg, Parker, and Taylor.

Line Spectrum

Some of the energy of the electron beam


excites the atoms of the target, i.e., promotes
them from their ground state to some higher
quantum states. The excited atoms then
relax back to the ground state and radiate
photons of various discrete frequencies
corresponding to the energy difference
between the quantized excited states and the
ground state. The original electron beam is
purely incidental to the radiation process; its
only role is to provide the energy to excite
the atoms. Thus, the frequencies of the
emitted radiation should be independent of Figure 2: Continuous X-ray
the energy of the electron beam. However, if spectrum with superimposed lines.
the electrons do not have enough energy to
promote the atomic electrons from the ground state then no radiation should be emitted via this
process.

It turns out that the excited states most important for X-ray production are those in which an
electron is ejected from the n = 1 shell, producing a vacancy in this shell. The photons are
produced when an electron from a higher shell drops down and fills the vacancy. The radiation
process is much like that which occurs in hydrogen to produce the Balmer series, except, of
course, that a multi-electron atom is involved.

We can estimate the transition energies by assuming that the system behaves like hydrogen.
Using Balmer’s formula reflects the idea that the electron making the transition moves in the
electric field of Z (the atomic number) protons and the single electron remaining in the 1s shell,
provided that the unit nuclear charge of the proton in hydrogen is replaced by Z-1.

By Gauss’s Law, the electric field produced by the spherical charge distribution of the other
electrons in the atom, which are at larger radii, does not produce an electric field near the
nucleus and therefore does not affect the electron actually making a transition. The result is
Moseley's approximate formula

1 1 1 
 R (Z 1) 2 2  2  (3)
 1 n 

where R is the Rydberg constant (1.097  107 m-1) and n is 2 or 3 for the X-rays observed in this
experiment. We observe two characteristic lines in this experiment, conventionally denoted K 
and   , the former having the longer wavelength. The K  line is actually a doublet, but the
University of Pennsylvania XR.4

splitting is too small to be resolved with our equipment. You will find the wavelength of the K 
line is well-approximated by Moseley's formula with n = 2 but that the theoretical prediction for
the   , line (n = 3) is considerably in error!

The Crystal Spectrometer

In this experiment we will analyze the X-rays by diffracting them from the ordered array of
atoms in the single crystal of NaCl (table salt) or LiF. The apparatus is illustrated in the sketch
below, which shows a collimated beam of X-rays hitting a crystal at a glancing angle and being
reflected at the same angle. The total deviation in beam direction is 2 , as shown in Figure 3.
From the sketch it appears that the X-rays are simply being reflected by a mirror.

Figure 3: The Crystal Spectrometer

What makes the situation here different from that of visible light reflected by a looking glass is
that the wavelength of the X-rays is comparable to the distance between atoms in the crystal.
Individual atoms are ultimately the entities that affect the radiation. To understand the
interaction of the X-rays with the crystal, one must find the radiation scattered by individual
atoms and then add these up to find the total scattered radiation, all the time taking due account
of the phase differences.

Briefly, the crystal may be regarded as being made up of planes of atoms, and we can first find
the scattered radiation produced by a single plane and then combine the contributions from all
planes. The general theory shows that, independent of wavelength, the radiation scattering by
the atoms in a single plane will all be in phase at an angle of reflection equal to the angle of
incidence, exactly as in optics. This explains the geometrical arrangement of the spectrometer.
But additionally, we find that at a given angle  the radiation from different planes of atoms will
be in phase only for a wavelength satisfying the Bragg condition (see Figure 4 on the next page)

m  AB BC  2d sin  (4)

where d is the spacing between the planes of atoms and m = 1, 2, etc. as shown in Figures 4 and
5 on the next page.
University of Pennsylvania XR.5

Figure 4: Bragg Reflection

Figure 5: Model of a crystal lattice

This explains why the instrument functions as a spectrometer: at a given angle only photons
whose wavelength satisfies the Bragg condition are scattered into the detector. In the lattice
orientation chosen for our experiment, the relevant NaCl lattice spacing is d = 0.282 nm, and we
can observe up to the first three orders (m = 1, 2, 3) in the diffraction pattern. Since, for each
order, different wavelengths will satisfy the Bragg condition at different angles, the entire
spectrum is repeated at each order. Thus, the measured intensity as a function of angle 2 should
qualitatively resemble that shown in Figure 6.
University of Pennsylvania XR.6

The K spectral lines result in diffraction peaks at approximately 32, 64, and 96 degrees. The
K lines result in a second set of peaks at slightly lower angles. The continuous radiation has a
smooth distribution with a sharp low-angle cutoff corresponding to  min .

We detect the scattered photons with a Geiger-Mueller (G-M) tube. The operation of a G-M
tube is explained in the Apparatus section in experiment 92. X-rays are less effective at ionizing
the gas molecules in a G-M tube than charged particles or gamma rays. Due to this, the G-M
tube reports only a small fraction of the X-rays passing through it. Since the reported fraction is
proportional to the actual number, this does not present a problem in determining angles of
maximum reflection strength. The graph will have the right shape with all vertical (count)
values reduced by the same proportionality constant. This constant is referred to only to clarify
the apparatus. It is not necessary to determine its value or correct for it since it has no effect on
the angles at which the peaks occur or the shape of the pattern.

Note that this experiment provides a beautiful example of the Complementarity Principle. We
are using both the wave and particle properties of the X-ray radiation to make our measurement.
The Geiger-Mueller is measuring the radiation one photon at a time, and the low-angle cutoff in
the spectrum would not be present unless there was a 1:1 correspondence between the maximum
energy transferred from an electron to a photon. The characteristic radiation arises from the fact
that the atom makes a transition from one discrete quantum state to another and in the process
emits a particle --- the photon --- that carries off the energy. Conversely, the functioning of the
spectrometer requires Bragg diffraction of the X-ray radiation from the crystal. This is an
explicitly wavelike phenomenon, and would not take place unless X-ray photons of a given
energy had a well-defined wavelength.
University of Pennsylvania XR.7

Preliminary Procedure

Use of the spectrometer Crystal X-ray tube


post
1. The clear lid should not be
opened by the student(s) at any
time, and especially while
making readings. Opening the
cover could affect alignment
and shielding. If the lid is not
properly closed and centered (a
fussy operation!), no power
will be sent to the X-ray tube.
When it becomes necessary to
open the lid, for example to Geiger
change crystals or collimators, Tube
or to change the accelerating
voltage, you must ask your TA Figure 7: TEL X-Ometer Table Top X-Ray machine,
to do this. top view with the lid open

2. Learn about the mechanical functioning of the spectrometer by examining the apparatus
carefully. It is clear that the crystal and the detector must be accurately aligned at every
scattering angle. It would be impossible to collect meaningful data if we had to realign the
crystal every time the detector was moved. To keep the detector lined up with the reflected
beam, the spectrometer is constructed with a gear arrangement that changes the angle of the
detector and simultaneously changes the angle of the crystal by exactly half as much. The
detector angle is “2 ” and is marked on the aluminum scale.

(A) (B)
Figure 8: The angular scale (A) and the thumb-wheel
3. Measuring the detector angle (see Figure 8): The G-M tube detector is advanced using the
orange plastic manual control handle visible on the side of the spectrometer. As the manual
control moves the carriage arm, an indicator shows the angle of the carriage arm, and
therefore the detector (2 ). Visible on the orange manual control is a thumb wheel with
numbered graduations. The thumb-wheel is designed to measure fractional degrees, and
University of Pennsylvania XR.8

advances when the carriage arm moves. Between each number on the thumb-wheel are six
divisions permitting the detector to be advanced as little as one sixth of a degree. When you
determine the general position around which you would like more detailed data, align the
thumb-wheel and the carriage arm at the nearest whole degree on the scale marking. The
thumb-wheel moves by friction, so it can be held in place with a finger while the manual
control handle is moved to the next whole degree. The thumb-wheel should now read
fractions of a degree in step with the main scale degree reading. Note that the thumb-wheel
can slip if you do not maintain pressure against it with your finger!

Collecting the count data:

1. The timing of the counting interval is


controlled by the white plastic box with
the blue face labeled “Scaling Ratemeter”.
None of the settings on the “Scaling
Ratemeter” box should be changed in
the course of this experiment.

2. The most important setting on the


ratemeter is the slide switch on the left
above the coaxial cable connection. The
pointer of that slide knob should be at the
“1K” setting. That value is not as
revealing as the “ts” values that are hidden
under the knob. The value you can’t see is Figure 9: The Scaling Ratemeter for taking
“15” and it tells you how many seconds readings with the Geiger tube
the count will be collected.

3. While counts are being collected the green “Run” light is on. An audible beep will announce
the end of a particular count period, and the device will hold that count for 10 seconds to
permit you to record the value and reset the counter. The red “Stop” light will be on during
the “recording period”. The cycle then begins again.

Setting and Clearing the Clock/Counter

The counter we are using (labeled “CLK/CTR”) is the device that needs the most attention in this
lab. This is a small aluminum box. The toggle switch on the counter should be switched down to
the “CTR” position, and the toggle switch on the right side of the front face should be switched
down in the “CNT” position. The red button on the left side of the front of the counter (“RST”)
is the reset button which must be pushed in to clear the reading before the next count is taken. If
this is not done the counter will continue to increment.

Part I Experimental procedure


University of Pennsylvania XR.9

1. Start with the voltage set at 20 kV, no crystal, a set of two collimating slits (the first wide, the
second narrow) in front of the detector, and an attenuator in front of the detector. Measure
the intensity when the detector is close to 2  0 and determine the angle of the intensity
maximum. Check to see if this is exactly at the nominal zero that you read from the dial; if
not you may need to subtract an offset from subsequent angular measurements.

2. Have your TA remove the attenuator and insert the NaCl crystal. Perform a rough scan to
determine the angles at which strong diffraction is seen.

3. Measure, as accurately as you can, the positions of the K and K lines of the first three
orders of diffraction. Use d = 0.282 nm to determine the wavelength of the characteristic
lines.

4. Have your TA remove the narrow slit in front of the detector. This will substantially degrade
the angular resolution of the apparatus, but also greatly increase the count rate. Verify that
you still see the characteristic diffraction lines.

5. Make careful measurements of the scattered intensity every 2 degrees from 5 to 50, and plot
your data. Make the best determination you can of the low-angle cutoff. From this, use
Bragg’s law to determine  min .

6. Have your TA change the accelerating voltage to 30 kV. Repeat steps 3 to 5, starting with
the narrow slit in place.

7. Have your TA replace the NaCl crystal with an LiF crystal. Repeat steps 2-6. The d-spacing
will be quite different (it is smaller, so the angles will be larger). You will be able to
measure one or maybe two diffraction orders. Determine the d-spacing of the LiF crystal and
2 min for both the 20 kV and 30 kV settings. Use your measured d-spacing to extract  min
from 2 min .

8. Use your measured values of  min together with Equation (2) to calculate h e .

Part I Questions

1) In step 5), did you observe the low-angle cutoff? The maximum at around 20 degrees?

2) Regarding step 6), the theory says that the positions of the characteristic lines should remain
the same (although their relative intensities may change) but that 2 min should decrease in a
predictable way. Is this the case?

3) In step 7), are the cutoff wavelengths for a given accelerating voltage the same as or different
from what you measured for the NaCl crystal? Is this what you expected?

4) Does your value for h e agree with the textbook value within experimental error?
University of Pennsylvania XR.10

5) Compare your measured wavelengths of the K and K lines with Moseley’s formula. You
will need to look up the atomic number of the target material, copper. Is the theory accurate?

Part II: Powder Diffraction

In Part I we used Bragg scattering from a single crystal to explore the spectral characteristics of
an x-ray source. We considered the crystal planes to act as half-silvered mirrors for the x-rays.
The crystal could be considered in that case to be a one-dimensional object. We will now think
about Bragg scattering in a (slightly) more sophisticated way.

A real crystal is a three-dimensional object, containing many sets of planes of atoms in different
orientations. Any of these planes can diffract if they have the correct orientation with respect to
the incident beam. A “powder” sample consists of thousands of tiny crystallites, randomly
oriented, so that we measure an average over all possible crystal orientations. Under these
conditions, we will measure Bragg scattering at any angle where any crystal planes satisfy the
Bragg condition. The powder diffraction technique is commonly used when the crystal structure
is completely unknown. It is often used to detect by mineralogists or metallurgists to detect trace
amounts of a minority crystal structure, or by chemists to determine the sequence of
thermodynamic phases in some new material.

For many structures, the algebra required to calculate the angles for the Bragg condition is rather
complicated. But for crystals with cubic symmetry, it is quite simple. We define three Bragg
indices, h, k, and l, each of which can be any positive or negative integer. The Bragg condition
is satisfied when the “momentum transfer,” q , is equal to a so-called “reciprocal lattice vector”:

4 2
q sin   Ghkl  h 2  k 2  l2     
 a0

Here  is the wavelength of the incident radiation,  is half the scattering angle 2, and a 0 is the
edge length of one unit of the cubic crystal. Equation (5) can obviously be solved either to give
the scattering angle 2 in terms of a 0 , , and the indices (h,k,l), or to give the squared sum of the
indices, h2+k2+l2, in terms of 2.

Equation (5) gives the positions of the Bragg peaks. Their intensities are determined by the
arrangement of atoms within the unit cell—some peaks may be very strong and others weak.

There is one further complication. There are actually three different kinds of cubic crystal:
“simple cubic” (sc) in which the same pattern (or “basis”) of atoms is repeated in each cube,
“face-centered-cubic” (fcc) in which the same pattern is repeated on the corners and the center
of each face of the cube, and “body-centered-cubic” (bcc) in which the same pattern is repeated
at the corner and the center of the cube. For a sc crystal, all peaks are “allowed.” But fcc and
bcc crystals have “selection rules”—rules to determine which peaks are nonzero. For fcc, the
rule is that h, k, and l must be all even or all odd. So, for example, the (100) and (110) peaks are
University of Pennsylvania XR.11

not allowed, but the (200) and (111) peaks are allowed. For bcc, the rule is that h+k+l must be
even.

The sample that you will measure is a powder of NaCl, a cubic crystal with a 0 =0.564 nm.

Procedure:

1. Remeasure the “beam zero” position of your diffractometer, as in step 1 of Part I.


2. Have your TA mount the powder sample.
3. Verify that you can still observe at least the first two of the K  reflections you saw in the
single crystal NaCl samples. They should be at (approximately) the same positions as
before, but substantially weaker. Use Equation (5), the value a 0 =0.564 nm, and your
previously determined wavelength for the Cu K  line to determine the (h,k,l) indices of
these lines.
4. Measure in the vicinity of calculated angles for all the other possible Bragg peaks
between 0 and 90° (or as many as you have time for). Give some thought to what kind of
step size is required, and what angular range in the vicinity of each peak—you certainly
will not have time to do the entire angular range in (1/6) ° steps! Which peaks are strong,
weak, or unobserved?
Questions

1. Determine the indices of the peaks that you observed.


2. From your data, can you determine whether NaCl is sc, fcc, or bcc? If so, which is it, and
how do you know? If not, what further information would you need?
3. Suppose that we had measured the same sample using Mo K  radiation, for which the
photons have approximately twice the energy as Cu K  photons. Qualitatively, how
would you expect the data to change? Would you expect to see more peaks, or fewer?

PAH 12/12/08 x-rays.doc

You might also like