Добавил:
Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:
Скачиваний:
19
Добавлен:
08.01.2014
Размер:
988.22 Кб
Скачать

 

 

B.2 Characteristic Temperatures of the Elements

597

 

Atomic

Name of

Chemical

Melting

ΘD

Ordered

Transition

number

element

symbol

point (C)

(K)

phase

temperature

20

calcium

Ca

842

230

 

 

 

 

21

scandium

Sc

1541

360

 

Tc = 0.40 K

22

titanium

Ti

1668

420

S

23

vanadium

V

1910

380

S

Tc = 5.46 K

24

chromium

Cr

1907

630

AF

TN = 311 K

25

manganese

Mn

1246

410

AF

TN = 100 K

26

iron

Fe

1538

467

F

TC = 1043 K

27

cobalt

Co

1495

445

F

TC = 1388 K

28

nickel

Ni

1455

450

F

TC = 627 K

29

copper

Cu

1085

343

 

Tc = 0.86 K

30

zinc

Zn

420

327

S

31

gallium

Ga

30

320

S

Tc = 1.08 K

32

germanium

Ge

938

370

 

 

 

 

33

arsenic

As

817

282

 

 

 

 

34

selenium

Se

221

90

 

 

 

 

35

bromine

Br

7

 

 

 

 

 

36

krypton

Kr

157

72

 

 

 

 

37

rubidium

Rb

39

56

 

 

 

 

38

strontium

Sr

777

147

 

 

 

 

39

yttrium

Y

1522

280

 

Tc = 0.63 K

40

zirconium

Zr

1855

291

S

41

niobium

Nb

2477

275

S

Tc = 9.25 K

42

molybdenum

Mo

2623

450

S

Tc = 0.92 K

43

technetium

Tc

2157

351

S

Tc = 7.8 K

44

ruthenium

Ru

2334

600

S

Tc = 0.49 K

45

rhodium

Rh

1964

480

S

Tc = 0.035 mK

46

palladium

Pd

1555

274

 

 

 

 

47

silver

Ag

962

225

 

Tc = 0.52 K

48

cadmium

Cd

321

209

S

49

indium

In

157

108

S

Tc = 3.41 K

50

tin

Sn

232

199

S

Tc = 3.72 K

51

antimony

Sb

631

211

 

 

 

 

52

tellurium

Te

450

153

 

 

 

 

53

iodine

I

114

106

 

 

 

 

54

xenon

Xe

112

64

 

 

 

 

55

cesium

Cs

28

38

 

 

 

 

56

barium

Ba

727

110

 

Tc = 5 K

 

 

57

lanthanum

La

920

142

S

 

 

58

cerium

Ce

798

146

AF

TN = 12.5 K

 

 

 

 

 

Continued on the next page

598 B The Periodic Table of Elements

 

 

 

 

Atomic

Name of

Chemical

Melting

ΘD

Ordered

Transition

number

element

symbol

point (C)

(K)

phase

temperature

 

 

 

 

 

 

 

 

59

praseodymium

Pr

931

85

AF

TN = 0.03 K

60

neodymium

Nd

1016

159

AF

TN = 6 K

61

promethium

Pm

1042

158

 

TN = 14.0 K

62

samarium

Sm

1074

116

AF

63

europium

Eu

822

127

AF

TN = 90.4 K

64

gadolinium

Gd

1313

195

F

TC = 293 K

65

terbium

Tb

1356

150

F

TC = 220 K

66

dysprosium

Dy

1412

210

F

TC = 90 K

67

holmium

Ho

1474

114

F

TC = 20 K

68

erbium

Er

1529

134

F

TC = 18 K

69

thulium

Tm

1545

127

F

TC = 32 K

70

ytterbium

Yb

824

118

 

Tc = 0.1 K

71

lutetium

Lu

1663

210

S

72

hafnium

Hf

2233

252

S

Tc = 0.13 K

73

tantalum

Ta

3017

240

S

Tc = 4.47 K

74

tungsten

W

3422

400

S

Tc = 0.02 K

75

rhenium

Re

3186

430

S

Tc = 1.70 K

76

osmium

Os

3033

500

S

Tc = 0.66 K

77

iridium

Ir

2446

420

S

Tc = 0.11 K

78

platinum

Pt

1768

240

 

 

 

79

gold

Au

1064

165

 

Tc = 4.15 K

80

mercury

Hg

39

72

S

81

thallium

Tl

304

78

S

Tc = 2.38 K

82

lead

Pb

327

105

S

Tc = 7.20 K

83

bismuth

Bi

271

119

 

 

 

84

polonium

Po

254

81

 

 

 

85

astatine

At

302

 

 

 

 

86

radon

Rn

71

 

 

 

 

87

francium

Fr

27

 

 

 

 

88

radium

Ra

696

89

 

 

 

89

actinium

Ac

1051

124

 

Tc = 1.37 K

90

thorium

Th

1750

170

S

91

protactinium

Pa

1572

159

S

Tc = 1.4 K

92

uranium

U

1135

207

S

Tc = 0.68 K

93

neptunium

Np

644

121

 

 

 

94

plutonium

Pu

640

171

 

Tc = 0.60 K

95

americium

Am

1176

 

S

96

curium

Cm

1345

 

 

 

 

 

 

 

 

 

 

 

 

B.2 Characteristic Temperatures of the Elements

599

References

1.CRC Handbook of Chemistry and Physics, Editor-in-Chief D. R. Lide, 85th Edition, CRC Press, Boca Raton (2004).

2.http://www.webelements.com

3.Springer Handbook of Condensed Matter and Materials Data, Editors: W. Martienssen and H. Warlimont, Springer-Verlag, Berlin (2005).

C

Mathematical Formulas

C.1 Fourier Transforms

When surface e ects are neglected and only bulk properties are examined in macroscopic crystalline samples, periodic boundary conditions are frequently applied, since the Fourier components that appear in the Fourier series or Fourier integral representation of the position-dependent quantities are often easier to determine. Nevertheless, owing to the invariance of crystals under discrete translations, functions that show the same periodicity as the crystal lattice and functions defined at the vertices of the crystal lattice are frequently encountered, too. It is often more convenient to specify them using the Fourier components associated with the vectors defined in the reciprocal lattice. The most important formulas of such functions are listed in the present section.

C.1.1 Fourier Transform of Continuous Functions

A periodic function of period L (f (x + L) = f (x)) can be expanded into a Fourier series as

f (x) = 21 a0

an cos

 

2πn

x + bn sin

 

2πn

x .

(C.1.1)

+ n=0

L

L

 

 

 

 

 

 

 

 

 

 

Making use of the orthogonality relation of trigonometric functions, it can be shown that the coe cients are given by the integrals

an = L

L/2

f (x) cos

L x dx,

bn = L

L/2

f (x) sin

L

x dx .

 

 

2

 

 

 

2πn

2

 

 

 

2πn

 

 

−L/2

 

 

 

 

 

 

−L/2

 

 

 

(C.1.2)

It is often more convenient to use exponential functions:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

f (x) =

fne2πinx/L ,

 

 

 

(C.1.3)

n=−∞

602 C Mathematical Formulas

where

fn = L

L/2

f (x)e2πinx/L dx .

(C.1.4)

 

1

 

 

 

−L/2

This can be demonstrated directly by exploiting the completeness relation

N

e2πinx/L = lim

N →∞

n=−∞

= lim

N →∞

 

e2πinx/L

(C.1.5)

n=−N

 

sin[π(2N + 1)x/L]

= (x)

 

sin(πx/L)

 

 

 

and the orthogonality relation

L/2

sin[π(n − n )]

 

 

 

 

e2πi(n−n )x/L dx =

=

n,n

.

(C.1.6)

π(n − n )/L

 

 

 

−L/2

The coe cients fn of the Fourier series make up the Fourier spectrum of function f .

The Fourier series representation is straightforward to generalize to functions defined in d-dimensional space, provided they satisfy periodic boundary conditions on hypercubes (or even more generally, hyperparallelepipeds) of volume Ld – and are repeated periodically outside it. For simplicity, consider a function f defined inside a three-dimensional general parallelepiped of edges N1a1, N2a2, N3a3, and volume V that satisfies the periodic boundary conditions

f (r + Niai) = f (r) , i = 1, 2, 3

(C.1.7)

on and beyond the boundaries. The Fourier series can then be written as

 

1

 

 

f (r) =

V

fˆ(k) eik·r .

(C.1.8)

 

k

 

 

 

 

On account of the periodic boundary conditions, the allowed vectors k are most easily expressed in terms of the primitive vectors bi of the reciprocal

lattice:

3

k = mi bi , (C.1.9)

i=1 Ni

where the mi are arbitrary integers. Recall that the primitive vectors of the direct and reciprocal lattices are related by (5.2.13).

ˆ

The explicit form of the Fourier coe cient f (k) can be derived either using the generalization

ei(r−r ) = V δ(r − r )

(C.1.10)

k

 

 

C.1

Fourier Transforms

603

of the completeness relation (C.1.5), or the orthogonality relation

 

V

ei(k−k )·r dr = V δk,k ,

(C.1.11)

leading to

 

 

 

 

fˆ(k) = V

f (r) eik·r dr .

(C.1.12)

It is easily seen that the convention used in the one-dimensional case is recovered if instead of (C.1.8) the Fourier series is defined as

f (r) =

 

(C.1.13)

 

fˆ(k) eik·r ,

 

k

 

 

and consequently the Fourier coe cients are given by

 

fˆ(k) = V

V

f (r) eik·r dr

(C.1.14)

1

 

 

 

instead of (C.1.12). The rationale behind choosing a di erent convention is that in su ciently large samples, where discrete sums are replaced by continuous integrals, the obtained formulas are independent of the sample volume in the V → ∞ limit. Since each vector k in the primitive cell of the reciprocal lattice is associated with a volume (2π)3/V , the sum over the k vectors can be replaced by an integral, using the formal substitution

 

V

 

dk .

(C.1.15)

(2π)3

k

Then the Fourier integral representation of an arbitrary function f (r) defined on the whole space is, by definition,

 

f (r) = (2π)3 fˆ(k) eik·r dk ,

 

 

 

1

 

 

where

 

fˆ(k) =

 

 

as in the V → ∞ limit

 

f (r) eik·r dr ,

 

 

(2π)3 ei(r−r ) = δ(r − r ) ,

 

 

 

 

 

dk

 

 

 

and

ei(k−k )·r dr = (2π)3δ(k − k ) .

 

(C.1.16)

(C.1.17)

(C.1.18)

(C.1.19)

604 C Mathematical Formulas

ˆ

The function f (k) is the Fourier transform of f (r) and (C.1.16) defines the inverse Fourier transform.

More generally, the Fourier transform of a function defined in d-dimensional space is given in the space of the d-dimensional k vectors as

ˆ ik·x d

f (k) = f (r) e d x ,

and the inverse Fourier transform is defined by

f (x) = (2π)d

fˆ(k) eik·x ddk .

1

 

 

The completeness and orthogonality relations then take the form

 

ddk

(2π)d ei(x−x ) = δ(d)(x − x )

and

ei(k−k )·xddx = (2π)dδ(d)(k − k ) .

(C.1.20)

(C.1.21)

(C.1.22)

(C.1.23)

In quantum mechanics, a common choice for the Fourier transform of the function f (r) is

fˆ(k) = (2π)d/2

f (x)eik·x ddx ,

 

1

 

 

 

 

−∞

 

and the inverse Fourier transform is then

 

f (x) = (2π)d/2

 

 

 

 

 

 

1

 

 

 

 

 

 

−∞

With this choice

 

f (x)

2ddx =

 

 

 

 

 

 

 

or more generally

 

 

 

 

 

 

f (x)g(x)ddx =

fˆ(k)eik·x ddk .

fˆ(k) 2ddk ,

 

 

ˆ d

f (kg(k)d k ,

(C.1.24)

(C.1.25)

(C.1.26)

(C.1.27)

which indicates that the Fourier transform is a unitary transformation in the space of square integrable functions that preserves lengths and inner products.

Another convention is used for the time variable. The Fourier transform of an arbitrary time-dependent function f (t) is defined as

 

 

 

fˆ(ω) =

 

f (t) eiωt dt ,

(C.1.28)

−∞

 

 

 

C.1 Fourier Transforms

605

and the inverse transform as

 

 

 

 

 

 

1

 

 

 

f (t) =

 

fˆ(ω) eiωt dω .

(C.1.29)

 

2π

−∞

Therefore the following formula is used for spaceand time-dependent functions that satisfy periodic boundary conditions at the boundaries of a sample

of volume V :

fˆ(k, ω) =

dr dt f (r, t) ei(k·r−ωt) ,

(C.1.30)

V

−∞

 

and

f (r, t) =

1 1

V k 2π

while for samples of infinite extent

dω fˆ(k, ω) ei(k·r−ωt) ,

(C.1.31)

−∞

fˆ(k, ω) =

dr

dt f (r, t) ei(k·r−ωt) ,

(C.1.32)

 

 

−∞

 

and

 

 

 

f (r, t) = (2π)4

 

dω fˆ(k, ω) ei(k·r−ωt) .

(C.1.33)

dk

1

 

 

 

 

−∞

These formulas can be applied to lattice-periodic functions, whose values inside the primitive cell of volume v spanned by the vectors a1, a2, a3 are repeated with the periodicity of the lattice – in other words, for each translation vector tn that can be written in the form (5.1.1),

f (r + tn) = f (r) .

(C.1.34)

Since condition (C.1.7) is now met by the choice N1 = N2 = N3 = 1, the vectors k appearing in the Fourier representation are the same as the vectors G of the reciprocal lattice, hence the Fourier transform of f (r) is

fˆ(G) = v

f (r) eiG·r dr ,

(C.1.35)

while the inverse transform is

 

 

 

 

1

 

f (r) =

fˆ(G) eiG·r .

(C.1.36)

v

 

G

 

 

 

 

606 C Mathematical Formulas

C.1.2 Fourier Transform of Functions Defined at Lattice Points

Functions defined at the vertices of a discrete lattice are frequently used in solid-state physics. Consider a discrete lattice of volume V , with N lattice points. When the function f (Ri) is subject to periodic boundary conditions, it can be represented as

 

 

 

1

 

 

 

 

 

f (Ri) =

 

N

fˆ(k) eik·Ri

,

(C.1.37)

 

 

 

k

 

 

 

 

 

 

 

 

 

 

ˆ

 

 

 

 

 

 

and the Fourier transform f (k) can be written as

 

 

 

fˆ(k) =

 

f (Ri) eik·Ri .

 

(C.1.38)

 

 

 

 

 

 

 

Ri

 

 

 

ˆ

ˆ

 

 

 

 

 

 

Since f (k) = f (k + G) for any vector G of the reciprocal lattice, it is su cient

to consider one vector k in each set of equivalent vectors (which di er by a reciprocal-lattice vector) – in other words, the vectors k given in (C.1.9) are defined within the primitive cell or Brillouin zone of the reciprocal lattice. The number of allowed vectors k is just N .

To justify the previous formulas, we shall demonstrate that in the limit

where the number of lattice points is large,

 

 

1

 

ei(k−k )·Ri = δk k ,G .

(C.1.39)

 

 

 

 

 

N

Ri

G

 

 

 

 

 

The equality obviously holds when k reciprocal lattice, since each of the N

− k is the same as a vector G of the terms in the sum

 

(C.1.40)

ei(k−k )·Ri

Ri

is then unity. Otherwise the phase factors cancel out to a good approximation. This cancellation can be most easily demonstrated in the case where the crystal contains odd numbers of lattice points along the direction of each primitive vector (a1, a2, a3). Expressed in terms of the primitive vectors b1, b2, b3 of the reciprocal lattice, k − k is

 

 

 

 

k − k

 

 

 

 

 

 

= (ki − ki)bi ,

(C.1.41)

 

 

 

 

 

i

 

and so the sum in question reads

 

 

1

N1

N2

N3

 

 

 

 

 

 

 

 

 

 

 

exp{−2πi[(k1 −k1)n1 + (k2 −k2)n2 + (k3 −k3)n3]} .

 

N

 

 

 

 

n1=−N1 n2=−N2 n3=−N3

 

 

(C.1.42)

C.1 Fourier Transforms

607

Performing the sum separately along the three directions,

1

 

Ni

 

 

 

 

 

 

 

 

 

exp[2πi(ki − ki)ni]

 

 

 

 

 

 

 

 

 

 

 

 

2Ni + 1

 

 

 

 

 

 

ni =−Ni

 

 

 

 

 

 

 

 

 

 

 

 

2Ni

 

 

 

 

 

 

=

exp[2πi(ki − ki)Ni]

 

exp[

 

2πi(k

k )n

]

 

 

 

 

2Ni + 1

 

 

 

 

 

 

 

ni =0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

=

exp[2πi(ki − ki)Ni]

 

exp[2πi(ki − ki)(2Ni + 1)] 1

 

 

 

 

 

 

 

 

 

 

2Ni + 1

 

exp[2πi(ki − ki)] 1

=

sin[π(ki − ki)(2Ni + 1)]

.

 

 

 

 

(C.1.43)

(2Ni + 1) sin[π(ki − ki)]

 

 

 

 

 

 

 

 

 

 

 

 

Wherever the denominator is zero, i.e., when ki −ki = 0, ±1, ±2, . . . (in other words, k − k is a vector of the reciprocal lattice), the expression tends to 1 in the limit N → ∞. Everywhere else it vanishes in the same limit.

Since the allowed vectors k are distributed continuously in the N → ∞

limit, the relation (C.1.39) can be recast in the equivalent form

 

1

 

 

)

 

 

(2π)3

 

δ(k − k − G) .

 

 

 

 

 

 

 

 

 

 

 

 

N

 

ei(k−k

 

·Ri

=

 

V

 

 

 

(C.1.44)

 

 

Ri

 

 

 

 

 

 

 

G

 

 

In one dimension this can be rewritten as

 

 

 

 

1

 

 

 

) R

 

2π

 

 

 

 

 

 

ei(k−k

·

 

n =

 

 

 

δ(k − k − Gh) ,

(C.1.45)

 

 

 

N

 

 

L

 

 

 

 

 

Rn

 

 

 

 

 

 

 

 

h

 

where Rn = na and Gh = (2π/a)h. Owing to the properties of the delta function, this is equivalent to

 

 

 

h) .

 

e2πinx =

δ(x

(C.1.46)

n=−∞

h=−∞

 

 

 

 

It can be shown along the same lines that for the sum over the discrete vectors k in the Brillouin zone

1

 

 

 

ei(Ri −Rj ) = δRi ,Rj .

(C.1.47)

 

N k BZ

 

In the N → ∞ limit, where the sum over the reciprocal lattice can be replaced by an integral,

N (2π)3

 

dk ei(Ri −Rj ) = δRi ,Rj .

(C.1.48)

V

1

 

 

 

k BZ

Naturally, integration is once again over the primitive cell or Brillouin zone of the reciprocal lattice.