H

Answers to exercises and hints

2. Sources 2γ/[γ(1 − β)] = 2/(1 − β) = 4γ2 . The X-ray

wavelength for the parameters given is 1 Å.
2.1: Start from Eq. (1.1) and use the fact that 2.8: u = 4 × 107 J/m3 =ǫ0 c2 hB2 iav
1 eV is equivalent to 1.602×10−19 J. =ǫ0 c2 B2L /2=ǫ0 c2 B2u /2/4. ∴ Bu ≈ 20 T. With
2.2: With ρ =24.8 m, I=0.2 A, and Ee =6 GeV, K ≈ 10−3 , and N = 104 , the flux is of order 1011
P =1.2 MW. photons/s/0.1%BW.
2.3: For protons, γ = 7.4 × 103 , ρ = 2.8 km, and
P=3.9 kW. For electrons, γ = 1.4 × 107 , and 3. Reflection and refraction from
P=44 PW. interfaces
2.4: No. The characteristic energy is ~ωc =
3~cγ3 /(2ρ) = 44 eV corresponding to a 3.1: Inside the material, mλ′ = 2d sin θ′ , and with
wavelength of λ=280 Å. λ′ = λ/n = λ/(1 − δ) this becomes mλ= 2d (1 − δ)
√ θ . Since n cos θ = cos θ, we have sin θ =
′ ′ ′
sin
2.5: The formulae given in the question can be
1 − cos2 θ′ ≈ sin θ (1 − δ/ tan2 θ). Therefore
rearranged to read ν = c/(λU (1 − β)), and with
mλ= 2d sin θ (1 −δ/ sin2 θ), and the formula
λ = c/ν and 1 − β= 1/(2γ2 ) the desired result is
stated in the question follows on making the
obtained.
substitution 1/ sin θ = 2d/(mλ).
2.6: Use Eq. (2.14) as your starting point, and
3.2: αc = 3 mrad, and ∴ α = 2.4 mrad. The
note that with the electron energy in GeV
minimum penetration depth is
1/(2γ2 )= (0.511 ×10−3 )2 / (2E2e ) =13.056×10−8/
1/Qc =1/(2kαc) = 32 Å.
E2e . An additional factor of 108 is introduced
when converting from cm to Å. 3.3: A plot of the square of the angular position
of the intensity maxima against the square of the
2.7: (a) The formula for the fractional energy
fringe order should produce a straight line with
change can be considered as a function G(ψ) of
an intercept of α2c and a gradient of (λ/(2t))2.
the scattering angle ψ. As the initial photon
energy χi is much less than the electron energy 3.4: 1.5 × 10−2 , 2.3 × 10−3 and 6.6 × 10−4 for
γi , and that βi is close to unity, one obtains that m = 2, 3 and 4, respectively.
G(ψ)= (1 − cos ψ)/ (1 + (χi /γi ) cos ψ) which 3.5: α = 2.5 mrad, and Ec =24.8 keV.
clearly has a maximum when ψ = π. (b) With
the same approximations as in part (a), and 3.6: t(r) = r2 /D, c.f. the equation immediately
R d/2
setting ψ = π, the fractional energy change is below Eq. (3.42), and tav = 0 2πr t(r)dr/

Elements of Modern X-ray Physics, Second Edition. Jens Als-Nielsen and Des McMorrow
© 2011 John Wiley & Sons, Ltd. Published 2011 by John Wiley & Sons, Ltd.
398 Answers to exercises and hints

R d/2
0
2πrdr = (d/8)α with α = d/D. 5. Kinematical scattering II: crystalline
3.7: The number of lenses required is N=235.
materials
The average thickness of a lens is tav =25 µm √
within the 400 µm aperture. The average 5.1: (a) d10 = 3/2, d11 = 1/2 (b) Let a∗1 = (α,β)
transmission is 55%. and a∗2 = (γ,δ), then use ai · a∗j = 2πδi j to
3.8: Footprint; 200 mm. ρtangential =2 km,
determine the √ coefficients α, β, γ and
√ unknown
δ. (e) dhk = 3a/(2 h2 + hk + k2 ).
ρ sagittal =5 cm. Alignment accuracy required:
0.025 mrad. 5.2: First show that sin2 θ is proportional to h2 +
k2 + l2 . Then calculate the eight values of the
ratio (sin θi / sin θ1 )2 for i= 1· · · 8, noting that θ
4. Kinematical scattering I:
is half of the given scattering angles, and
non-crystalline materials compare with h2 + k2 + l2 . This allows the peaks
4.1: The intrinsic width in energy (FWHM) of to be indexed. Lastly compare with the selection
the inelastic line can be estimated by subtracting rules for simple cubic, bcc and f cc, from which
in quadrature the resolution estimated from the you should conclude that the sample has a
elastic line. This assumes that both the elastic simple cubic lattice.
and inelastic lines may be approximated by 5.3: Use Eq. (5.34) which requires evaluating
Gaussians. The FWHM of the momentum the multiplicities of all reflections. Note that the
distribution is then estimated to be ∼ 0.3 Å−1 . (300) and (221) have the same value of h2 + k2 +
4.2: Assume that both electrons in the He atom l2 .
have the same form factor given by Eq. (4.9). 5.4: Fh,k,l = 4( fNa − fH ) for (h, k, l) all odd, and
The desired result follows immediately, which Fh,k,l = 4( fNa + fH ) for (h, k, l) all even. For
yields essentially perfect agreement when X-rays, f ∝ Z for Q=0, and since ZNa ≫ ZH ,
compared with Fig. 4.5(a). both types of peaks are present. In the case of
4.3: I(Q)= 2Z 2 [1+ sin (Qa)/(Qa)]/ [1+ neutrons, the fact that the (h, k, l) even peaks are
(Qa/20)2]4 . Qa=0 for the first maximum. absent implies that the scattering lengths of Na
Numerical plot reveals the second maximum at and H are of equal magnitude but of opposite
Qa ≈ 6.75. sign.
4.4: 2 f12 + f22 + 4 f1 f2 sin (Qa) / (Qa)+ 5.5: Use dhkl = 2π/|G|.
2 f12 sin (2Qa) / (2Qa). 5.6: The (1,0,0), |F100 |2 =1; and the (0,0,2),
4.6: Integration of ρ(r) over all space must yield |F100 |2 =4, are the first two allowed reflections
the total charge of 6 × 60 = 360 electrons for the that are closet to the origin.
C60 molecule, and hence A=360. 5.7: By definition, M = BT (sin θ/λ)2
4.7: Guinier approximation, yG = e−x /5 = 0.7020
2
=BT (G/4π)2 =BT (h2 + k2 + l2 )/(4a2). Θ/T =2.36
for x= 1.33. Exact, y = 3[sin x −x cos x ]/ x3 = =⇒ φ=0.557 from Fig. 5.16, BT = 0.287 Å2 and
0.6954 for x= 1.33. B0 = 0.148 Å2 . For kinematical diffraction we
R require that Ihkl (293)/Ihkl(0) = e−2M293 /e−2M0
4.10: F (Q) ≈ [V pR −(1/2) (q · r)2 dV p +...] /V p <0.95. By trial and error the (422) is calculated
= 1− (q2 /(6V p )) r2 dV p = 1− q2 R2g /6, and the to be the lowest order, non-forbidden Bragg
result follows when substituting the form factor reflection that loses 5% of its intensity on
into Eq. (4.22). warming to room temperature.
5.8: It follows from Fig. 5.15(c) that the density
of states g(ω) is proportional ωd−1 , where d =1,
2 or 3 for one, two and thee-dimensional lattices.
 399

The integrand in the equation for hu2 i on 6.3: Consider Fig. 6.5 and Eq. (6.20) from which
page 176 is so strongly divergent for d =1 that it should be clear that the correct factor√is
the integral is infinite. obtained from the condition that (x − √ x2 − 1)2
= 1/2 which has the solution x= 3/(2 2).
5.9: The contribution to the scattering amplitude
from the layers above z = 0 form a geometric 6.4: g is the amplitude reflectivity per layer. The
series ηe−i2πl + η2 e−i4πl + · · · which can be amplitude reduction at the n’th layer is ∆A =
summed to yield ηei2πl /(1 − ηe−i2πl ). Once this −gAn , ⇒ A(n) = A0 exp(−gn). After N = 1/g
amplitude has been added to the standard layers the amplitude is reduced by a factor of
expression for the crystal truncation rod 1/e. The amplitude extinction depth is therefore
scattering amplitude, the scattered intensity can Nd =d/g. The intensity extinction depth is found
be calculated. For small η, the intensity at the from I(n) = I0 exp(−2gn) to be d/(2g).
anti-Bragg point is ∼ 1/4 − η, from which it is 6.5: ∆θ [rad]=(ζDtotal/2) (|F0 |/|F|) tan θ = (g /mπ)
clear that roughness decreases the intensity in (|F0 |/|F|) tan θ =(g0 /mπ) tan θ where we have
the specular CTR rod. used Eq. (6.28), (6.25) and (6.2), respectively.
5.10: The scattering amplitude is: 1/(1 − ei2πl ) With g0 = mπδ/ sin2 θ we obtain ∆θ [rad]=δ
+ei2πl /2, and the intensity at the anti-Bragg point tan θ/ sin2 θ = 2δ/ sin 2θ as required.
is zero. Thus at the anti-Bragg point l = 0.5 the 6.6: (a) ∆E/E= cot θ∆θ. (b) θ = 90◦ ⇒ λ = 2d =
intensity will oscillate between 0 and |F(Q)|2 /4 0.5226 Å, or E= 23.72 keV. (c) µ−1 (23.72 keV)=
during layer-by-layer growth allowing the 1720 µm, from Eq. (6.33) Λext = 174 µm, which
process to be studied. is much less than the absorption length so
dynamical diffraction applies. From Eq. (6.26)
6. Diffraction by perfect crystals the relative bandwidth (FWHM) is 5.07 ×10−8
yielding a resolution of ∆E =1.2 meV.
6.1: (a) The (400) is the first allowed symmetric 6.7: We have√from Eq. (6.1) g ≈ 2(a2 /3)r0 /(a3 )
Bragg peak; see Section 5.1.7. (b) The key is to 4|1 + i|Z =8( 2/3) (r0 /a)
√ Z and an extinction
√
use Eq. (6.1). From Tables 4.1 and 6.1 the depth given by Λ = (a/ 3)/(2g) = a 3/2
atomic form factor is f (G400 )=7.51. The (1/16) 1/((r0 /a)Z). Hence the extinction depths
Debye-Waller factor DW = exp(−BT (sin θ/λ)2 ) are estimated to be ΛC = 0.57 µm, ΛS i = 0.57 µm,
at room temperature is evaluated by taking and ΛGe = 0.27 µm, within a factor two of the
BT = 0.33 (Table 5.1) and sin θ/λ= values obtained from a full calculation.
G400 /4π=0.3683, yielding DW=0.9562. Thus F
6.8: The tabulated value of the absorption length
=8 f (G400 ) × DW=57.45, and g = 2.6 × 10−6. (c)
at 12.4 keV is µ−1 = 11 µm. The absorption
The required amplitude reflectivity√ is 1/10. From depth is (µ−1 /2) sin θ ≈ 1 µm. The extinction
Eq. (6.20) this implies that x + x2 − 1 = 10, or
length (given in the text) is 8 µm. The scattering
x ≈ 5. The width of total reflection
√ (∆x = 2) is
is therefore kinematic, because absorption
26.3 × 10−6 multiplied by 2 2/3 or 24.8 × 10−6
prevents significant multiple scattering, and the
(Table 6.1). The offset angle is then calculated as
integrated intensity is proportional to |F|2 .
5/2 times 24.8 × 10−6 multiplied by tan θ.
6.9: w333 /w111 = (1/9) (|F333 |/|F√
111 |)
6.2: x ≤ −1, r is always real and negative and
(tan θ333 / tan θ111 ) =(1/9) (8/(4 2)) ( f333 / f111 )
therefore the phase is equal to π or −π, where in
(tan θ333 / tan θ111 ).
our exposition we have made the latter choice;
|x| ≤ 1, the√phase is equal to −acos 6.10: The asymmetry parameter is
(x/(x2 + ( 1 − x2 )2 )1/2 ) = −acos(x); x ≥ 1, r is sin(45◦ )/ sin(15◦ ) ∼ 3, or its reciprocal, 1/3 . In
always real and positive and hence the phase case (a) the beam incident on the second crystal
angle is zero. is spatially extended, so it can only accept a
400 Answers to exercises and hints

√
narrow angular range equal to 1/ 3 of the 8. Resonant scattering
symmetric Darwin width. I(θ) is therefore the
convolution of a top hat, say of height 1 and of 8.1: Equation (8.13) can be rearranged to read
width 1 from
√ the first crystal, with a top hat of | f ′′ | = E[keV] σa [Å2 ]/ 6.993 × 10−6 , which
width 1/ 3 from the second crystal. The when substituted into Eq. (1.18) yields the
convolution
√ is a parallelogram with
√ baseline of desired result.
1+1/ 3 and top width of 1−1/ 3. With a
8.2: (a) Conventional bcc unit cell has two atoms
height of 1, the integrated intensity (area) is 1. In
with fractional coordinates of (0,0,0) and
case (b) the convolution
√ is also a parallelogram
√ (1/2,1/2,1/2). Thus Fhkl = fav. (1+exp(
with baseline 3+1 and top width of 3−1.
iπ(h + k + l))) with fav. = ( fCu + fZn )/2. The
Requiring the same area means that √ this condition on the Miller indices is that (h + k + l)
parallelogram has a height of 1/ 3.
must be even. (b) Fhkl = ( fCu ± fZn ) where the +
and − refer to (h + k + l) even, and (h + k + l)
7. Photoelectron absorption odd, respectively. (c) Ratio is approximated as
[(30-29) /(30+29)]2 ≈ 2.9×10−4.
7.1: Use Eq. (1.19) with σa = (µ/ρm ) (M/NA )
from Eq. (1.18). 8.3: (a) We have σa = (4πro /k) | f ′′ | =2× 12.398
7.2: From Eq. (7.2), 1/µ= 50 µm. r0 [Å] | f ′′ | / E [keV]. For a composite material,
P
µ = j ρat, j σa, j = (σCu, j +σZn, j )/a2 . The values
7.3: 1/µ= 20 µm, which is shorter than the value of aµ at the four energies are 9.53, 9.24, 39.7
at 50 µm due to the presence of the Ga and As K and 38.5 all multiplied by 10−6 . (b) |F100 |2 : 6.95,
edges at 10.367 and 11.867 keV, respectively. 57.3, 68.4, 16.7; from Eq. (5.31), I sc /I0 : 7.58,
7.4: 1.7 m. 62.0, 17.5, 4.29 all multiplied by 10−6 . (c) 100
eV below the edge elastic scattering only and
7.5: In the absence of any absorption edges, the photons of 8878.9 keV will be observed. 100 eV
absorption length decreases as the cube of the above the edge in addition to the elastic
X-ray energy. The absorption length in air at 1 scattering there is also fluorescence at Kα =
keV is approximately 1.7 mm. 8.048 keV and Kβ = 8.910 keV with an intensity
7.7: The result follows from making the ratio of about 10 to 1.
substitution κ = 2/a in Eq. (4.9). 8.4: (a) The beam divergence of the scattered
7.8: The total elastic scattering from an atom is beam is 2 mrad, and fills the detector aperture
approximately Z(8πr02 /3), and equating this to when the detector is placed a distance of L= 500
Eq. (7.18) leads to the required result. The mm from the sample. (b) ∆Ω = πr2 /L2 = 3.14×
absorption and scattering cross-section are 10−6 . (c) From 8.3(b), I sc /I0 = 4.29 × 10−6 . All
estimated to be equal at ∼ 1.87 keV, which is incoming photons will be absorbed in the thick
much greater than the energy of the K edge, sample in symmetric Bragg reflection geometry.
justifying setting the Stobbe correction factor 30% will be re-emitted as fluorescence, and the
f (ξ) equal to unity. fraction ∆Ω /4π will reach the detector. Thus
I f /I0 ≈ 0.075 × 10−6 , and the ratio of scattering
7.9: 0.02 atm.
to fluorescence is ∼66.
7.10: √
The second √ neighbour shell is at a distance
8.5: The two possibilities are F111 = fZn + fS
of a/ 2 = 6.48/ 2 = 4.58 Å in good accord
e2π(1+1+1)/4 or F1̄1̄1̄ = fS + fZn e2π(1+1+1)/4 .
with the data in Fig. 7.9. ′ ′′
Inserting fZn = fZn + i fZn , etc., and rearranging
7.11: (a) ǫ(Z)=0.0026Z−0.23. (b) EK (Ga)= leads to the desired result.
10.367 keV, EK (Ga)= 11.868 keV, both within 1
8.6: The result follows directly from the answer
eV of their tabulated values.
to Exercise 8.5.
 401

8.8: The calculated ratios are 1.58, 1, and 0.51, the number of electrons in the cubic f cc unit
in good agreement with the measured ratios. cell, 4Z, divided by its volume, or 4.687 el. per
Å3 , yielding λ= 2.38 Å. The focal length is equal
9. Imaging to 21 mm. The depth of focus is approximately
the resolution 1.22 ∆r M divided by the aperture
9.1: C = 1− e−(µ2 −µ1 )z2 if µ1 < µ2 , and C = 1− angle D/ f , or 12.8 µm.
e(µ2 −µ1 )z2 if µ1 > µ2 .
9.2: C= 0.98, 0.64, 0.32, 0.24, ordered in terms
of increasing photon energy.
9.3: Taking the over all thickness of the chest to
be 30 cm, with two chest walls each of 2.5 cm,
and the rib bones to have a thickness of 1 cm
each, Cribs = 0.64, Ctumour = 0.30.
9.4: (a) 267, 385 and 526 eV for C, N and O,
respectively. (b) 0.52 and 7.4 µm for protein and
water, respectively.
9.6: The wedge angle is determined
√ √by tan ω=
2, so the wedge thickness is y = 2x, with x
denoting denoting the beam position. The
transmission T (x) = exp(-y/y0) with y0 = 184
µm. The refracted
√ angle is α= δ tan ω = λ2
[ρr0 /(2π)] 2 = 5.5 × 10−6 radians.
9.7: The maximal intensity < y > is the overlap
integral of the triangular pattern with the
Gaussian centred at a triangle peak position.
When (σ/p) ≪ 1 the overlap with neighbouring
triangles
√ R ∞be neglected, and one finds < y > =
can
( 2πσ)−1 −∞ exp(−x2/(2σ2 ) T (x) dx with the
triangular function given by T (x) = 1 − (2/p)x
when |x| < p/2 or 0 otherwise. The integral is
readily evaluated to obtain the desired result.
9.8: For a crystal of around (500)3 unit cells, the
integrated intensity can be calculated from Eq.
(5.30), yielding ∼ 300 counts per second. If we
assume a detection limit of around 1 count per
second, then we can estimate by considering the
properties of | sin(x)/x|2 that around 4 − 5 fringes
should be visible.
9.9: From D = 4M∆r M one finds the number of
zones M=500. The outer √ zone widthp ∆r M is
related to λ and f by 2 M ∆r M = λ f . The
wavelength is determined by which depth of the
zones, t, provides a phase shift of π, i.e. tδk = π
or λ = π /(tρr0 ). The electron density in Au is