You are on page 1of 18

Chapter 2 Crystal binding and elastic constants

I. Chemical bonds

1. There are two major types of chemical bonds (bonding between atoms to form a molecule: (i) ionic bond, and (ii) covalent bond. 2. Inoinzation energy (first) is the energy required to move an electron from a neutral isolated atom to form an ion with one positive charge. 3. Electron affinity is the amount of energy absorbed when an electron is added to a neutral isolated atom to form an ion with one negative charge. It has a negative value if energy is released. Most elements have a negative electron affinity. This means they do not require energy to gain an electron; instead, they release energy. Atoms more attracted to extra electrons have a more negative electron affinity. Only elements that have a positive affinitive are the atoms with a complete shell (e.g. group 2 and the inert gases). 4. Electronegativity is the average of the first ionization energy and the electron affinity. It is the measure of the ability of an atom or molecule to attract electrons in the context of a chemical bond. In Pauling scale, F (the most electronegative element) is given an electronegativity value of 3.98 and Li (the least electronegative element) is given an electronegativity value of 0.98. Electronegativity 7 8 9 10 11

Group 1 2 3 4 5 6 12 13 14 15 16 17 18 Period H He 1 2.20 Li Be B C N O F Ne 2 0.98 1.57 2.04 2.55 3.04 3.44 3.98 Na Mg Al Si P S Cl Ar 3 0.93 1.31 1.61 1.90 2.19 2.58 3.16 K Ca Sc Ti V Cr Mn Fe Co Ni Cu Zn Ga Ge As Se Br Kr 4 0.82 1.00 1.36 1.54 1.63 1.66 1.55 1.83 1.88 1.91 1.90 1.65 1.81 2.01 2.18 2.55 2.96 3.00 Rb Sr Y Zr Nb Mo Tc Ru Rh Pd Ag Cd In Sn Sb Te I Xe 5 0.82 0.95 1.22 1.33 1.6 2.16 1.9 2.2 2.28 2.20 1.93 1.69 1.78 1.96 2.05 2.1 2.66 2.6 Cs Ba Lu Hf Ta W Re Os Ir Pt Au Hg Tl Pb Bi Po At Rn 6 0.79 0.89 1.27 1.3 1.5 2.36 1.9 2.2 2.20 2.28 2.54 2.00 1.62 2.33 2.02 2.0 2.2 Fr Ra Lr Rf Db Sg Bh Hs Mt Ds Rg Uub Uut Uuq Uup Uuh Uus Uuo 7 0.7 0.9

-1-

The only exception is helium. Common types of crystal bindings: (i) Ionic bonding (ii) Covalent bonding (iii) Metallic bonding (iv) Hydrogen bonding (v) Van der Waals interaction 4.44-2. -2- . atoms of the chemical elements out of which the solid is composed).16-0.24) Small difference in electronegativity ⇒ covalent bonding e. C – O (difference in electronegativeiy = 3. Helium only liquefies at low temperatures under ambient pressure. 1.g.89) O–O (difference in electronegativeiy = 0) Crystal bindings II. Cohesive energy u of a solid is the energy required to disassemble the solid into its constituent part (e. 5. electrons occupy a complete shell). 82.2 = 1.g.55 = 0.93 = 2. Na – Cl (difference in electronegativeiy = 3. but only under high pressure (higher than 27 atm. Electrons and electrostatic forces play an important role in binding atoms together to form a solid (crystal). Rev. Bonding between two atoms: Large difference in electronegativity ⇒ Ionic bonding e. 3. There must be some bindings to hold atoms together to form a solid. and a repulsive term when the inter atom distance is short (so the crystal will not collapse). All materials solidify at low enough temperatures.5.23) Moderate difference in electronegativity ⇒ polar covalent bonding e. H – O (difference in electronegativeiy = 3. 2.g. Two factors for this to occur: (i) helium atom is very light (hence high ground state energy and high zero point oscillation if it solidify).e.44-2. the cohesive energy has an attractive term when the inter atomic distance is large (so that the crystal can be formed). Helium can be solidified. and (ii) helium atom is inert (i. For a stable. Ref: Phys.g. 263–264 (1951)).

If R is the nearest neighbor distance. Uij = λ exp (-rij/ρ) ± q2/(4πε0rij) where ± q are the charges of the ions. The Coulomb attraction. 4. When the difference in electronegativity between two different types of is large. The attractive force is due to the Coulomb attraction between the ions. The repulsive force is due to the Pauli exclusion principle – this prevents the crystal from collapsing.R/ ρ 1 q2 λ − e ⎪ 4πε 0 R ⎪ U ij = ⎨ 2 ⎪± 1 q ⎪ ⎩ 4πε 0 Rp ij -3- (nearest neighbors) (otherwise ) . The equilibrium distance between two atoms is given by ∂u ∂r =0 r = r0 III. Ionic bonding 1.g. and we write rij = pij R. is long range and it will extend indefinitely. 3. The repulsive potential is really short range. then we have ⎧ . Na + Cl → Na+ + Cl-). however. The low electronegative atom will become a positive ion and the high electronegative atom will become a negative ion (e. hence it is effective only for nearest neighbors. electrons will be transferred from the low electronegative atom to the high electronegative atom. If Uij is the interacting energy between ions i and j. These ions will attract each other by electrostatic force to form a solid. and λ exp (-rij/ρ) is the repulsive potential with λ and ρ as empirical constants. 5. 2.u repulsive Equilibrium distance Distance between two atoms attractive 6. which can be determined from lattice parameter and compressibility.

.. The Madeling has to be positive for an attractive potential. the two atoms can form covalent bond by sharing a pair of electrons (one from each atom). One direct way to overcome this problem is to break up the terms and rearrange the summation so that each cell remains neutral in charge. IV.. mathematically. When the electronegativiy between two atoms is small. For example..... for a one dimensional linear lattices. C has four outer electrons (2s2p2 or more correctly 2sp3) and hence it can form 4 covalent bonds.. 2 3 8.. α for NaCl structure is 1. If the total number of ions is N and the number of nearest neighbors is z. p 0j 7. Total energy of the whole crystal Utot: U tot = N × potential energy correspondes to one single ion = N ∑ U 0j j≠ 0 ⎡ 1 q2 = N ⎢zλ e -R/ρ + ∑ ± 4πε 0 Rp 0j j≠ 0 ⎢ ⎣ ⎤ ⎥ ⎥ ⎦ ⎡ q2 1 ⎤ = N ⎢zλ e -R/ρ + ± ⎥ ∑ 4πε 0 R j≠ 0 p 0j ⎥ ⎢ ⎣ ⎦ ⎡ q 2α ⎤ = N ⎢zλ e -R/ρ − ⎥ 4πε 0 R ⎦ ⎣ where α = ∑ m j≠ 0 1 is the Madelung constant. + − + + − + + .) p 0j 3 2 1 1 2 3 α =∑ j≠ 0 = 2(1 − = 2ln2 because ln(1 + x) = x − 1 1 + + . O + + 1 1 1 1 1 1 1 m = (. It depends on the crystal structure only... 2.6. This is because the Coulomb force is a long range force and it is not decaying fast enough to ensure a convergent series. However.. Most atoms can form more than one covalent bond.) 2 3 x2 x3 + + . For example...7476. -4- . the series is divergent. Covalent bonds 1.

with atoms joined to four nearest neighbors at tetrahedral angles slowing only four nearest neighbors.e. En = kn = ⎜ ⎟ a 2m ⎝ a ⎠ 2 Energy levels are very close together.3. V. etc.74 of close-pack structure. These neutral molecules have no charge for ionic bonding with other molecules. The total Coulomb potential is negative: e -e e -i i -i U Coul = U {+ U { + U { <0 positive negative positive 3. Good example is alkali metals (Li. 1. K. and they do not have the extra electron to form covalent bonding with other molecules. A crystal can be formed with one atom forming covalent bonds with several other atoms. repulsive term) of the electrons. For example. and hydrogen bonding. Very often atoms form molecules by forming molecules. because a is large -5- . Covalent bond is highly directional and the bonds will repel from each other. carbon and silicon can have diamond structure. Na n K+ ions n free electrons 2. 4. So a crystal can be formed even the structure has a low filling factor. They need other types of crystal bonding to form a solid: Van der Waals interaction. The attractive potential is balanced by the kinetic energy (i. 5. Metallic bonding Atoms bounded by “free electrons”. One can view the metallic bond as a total delocalization of electrons from covalent bond.34 compared with 0. One can view these as electrons trapped in a box formed by the solid: h 2 ⎛ nπ ⎞ nπ . metallic bonding. Diamond structure has a filling factor of 0.

147) v v 4πε 0 | x . Van der Waals interaction 1. It is not necessary for metals to have metallic bonding. v v v v For simplicity .x 0 v v If the polarizabi lity of the seecond sphere is α . v v v Electric field at p 2 due to p 1 is E = v -p 4πε 0 r 3 v v v wher e r = x . let us assume x . Suppose there is a dipole moment of P1 at the first sphere.αp 1 v p 2 = αE = 4πε 0 r 3 -6- .x 0 is perpendicu lar to p 1 (and eventually p 2 ) as shown in above figure. VI. This will produced a dipole at the second sphere. 4.p v v Electric field at point r due to a dipole = E(x) = (Jackson 3rd edition p.x 0 |3 ) v v n is a unit vecto r along x . For example.x 0 direction.As a decreases (when the crystal is “collapsing”). 3. many transition metals show covalent properties. En (the kinetic energy of the electrons) increases and this prevents the crystal from collapsing. as long as they have some “internal charges” so that the neutral spheres can be polarized: + + + + + + P1 P2 + + Coulomb attraction 2. we have a relationsh ip between p 1 and p 2 : v . Coulomb attraction can occur between two neutral spheres. P1 r=x-x0 P2 x0 x O ) v ) v 3n(p ⋅ n) .

It is customary to use power law to describe the repulsive potential: B r 12 The total potential between two molecules can then be written as: Φ rep (r) = 5. p. Φ (r) = Φ att (r) + Φ rep (r) = - A B + r 6 r 12 ⎡ ⎛ σ ⎞ 6 ⎛ σ ⎞12 ⎤ = 4ε ⎢.⎜ ⎟ + ⎜ ⎟ ⎥ N 2v 2v r ≠0 r ≠0 ⎢ ⎣ ⎝r⎠ ⎝r⎠ ⎥ ⎦ v v ( r = r0j ) Let r = pR where R= nearest neighbor distance ∴u = ⎡ 1 1 4ε ⎢. 6.⎜ ⎟ + ⎜ ⎟ ⎥ ⎢ ⎣ ⎝r⎠ ⎝r⎠ ⎥ ⎦ 6 12 where 4εσ = A and 4εσ = B. 151) v v 4πε 0| x .6 ∑ 2v r ≠0 ⎢ ⎣ p 1 ⎛σ ⎞ ⎜ ⎟ + 12 p ⎝R⎠ 6 ⎛σ ⎞ ⎜ ⎟ ⎝R⎠ 12 ⎤ ⎥ ⎥ ⎦ ⎤ ⎡ 6 12 ⎢ 1 ⎛σ ⎞ 1 ⎛σ ⎞ ⎥ = 2ε ⎢ − ∑ 6 ⎜ ⎟ + ∑ 12 ⎜ ⎟ ⎥ v v r ≠0 p ⎝ R ⎠ r ≠0 p ⎝ R ⎠ ⎥ ⎢1 4 24 3 1 4 2 4 3 ⎥ ⎢ A6 A12 ⎦ ⎣ 6 12 ⎡ ⎛σ ⎞ ⎛σ ⎞ ⎤ = 2ε ⎢ − A 6 ⎜ ⎟ + A 12 ⎜ ⎟ ⎥ ⎝R⎠ ⎝R⎠ ⎥ ⎢ ⎣ ⎦ -7- .6 3 3 3 6 4πε 0 r 4πε 0r 4πε 0r 4πε 0 r r This is the attractive potential between two neutral molecules.x 0 |3 v v Therefore the potential energy between p1 and p 2 is : v v v v p1 ⋅ p 2 p1 .The potential energy two interacting dipoles is given as : v v ) v ) v p1 ⋅ p 2 − 3(n ⋅ p1 )(n ⋅ p 2 ) Φ= (Jackson 3rd edition. 4.αp 1 αp 1 2 1 A Φ att (r) = = ⋅ = =. The total potential energy of the solid is N N v v v U = ∑ Φ ( rij ) = ∑ Φ ( r0j ) = ∑ Φ ( r ) 2 j≠ 0 2 v i≠ j r ≠0 Potential energy per particle is: u= ⎡ ⎛ σ ⎞ 6 ⎛ σ ⎞12 ⎤ U 1 1 = ∑ Φ (r) = ∑ 4ε ⎢. Repulsive potential is due to Pauli exclusion principle.

When this happen: -8- . 10-5 times smaller than any other ions. VII. Larger molecule ⇒ stronger Van der Waals force ⇒ higher melting point. For example: He Ne Ar Kr Xe Rn Increasing melting point This is also true for many organic molecules. The nearest neighbor distance is determined by minimizing u: 6 12 ⎡ ∂u ∂ ⎛σ ⎞ ∂ ⎛σ ⎞ ⎤ = 0 ⇒ 2ε ⎢ − A 6 ⎜ ⎟ + A 12 ⎜ ⎟ ⎥=0 ∂R ∂R ⎝ R ⎠ ∂R ⎝ R ⎠ ⎦ ⎢ ⎥ ⎣ 5 11 ⎡ σ σ ⎤ ⇒ 2ε ⎢− 6A 6 6 + 12A 12 12 ⎥ = 0 R R ⎦ ⎣ ⇒ − 6A 6 + 12A 12 ⇒ R6 σ6 R6 =0 σ 6 = 2A 12 A6 1 ⎛ 2A 12 ⎞ 6 ⇒ R =⎜ ⎜ A ⎟ ⎟ σ ⎝ 6 ⎠ = 1.g. It can form only one covalent bond and it does not have sufficient bond to bind the whole crystal together with covalent bond. The complete shell of hydrogen atom is 2 electrons and a hydrogen atom has only one electron. It is highly unlikely for hydrogen to form ionic bonding.13. 9. oxygen) can often be polarized. 4. the covalent bond between hydrogen and the other atom (e.6 eV) because the outset electron is in the first shell without any shielding.7. 8. 3. Hydrogen bonding 1. 2. Hydrogen ion is just a proton (10-13 cm in radius). when the electronegativity between these two atoms has a moderate difference. A6 = 14.15 and A12 = 12. A6 and A12 depend on the crystal structure only.09017 σ for fcc. However. First ionization energy of atomic hydrogen is very high (13. For fcc (structure of inert gas).

Electron density + H X - 5. These polarized molecules will “stick” to each other by Coulomb attraction. This is possible because the hydrogen size is very small. For example. for water: H O H Hydrogen bonding O H Ice structure: H -9- .

y. i and j are not necessary the same.VIII. Strains 1. if O is fixed in the process. This point moves to point P’ at (x+ux. The displacement is a cumulative effect of other points. u= (ux.x) per unit undistored length in direction j (j=x. ∆z) from point P. z+uz).y. ∆y. z). its (Q’) displacement from point P’ after the deformation will be given by ∂u ∂u x ∂u ∆x + x ∆y + x ∆z = ε xx ∆x + ε xy ∆y + ε xz ∆z ∂x ∂y ∂z ∂u y ∂u y ∂u y ∆u y = ∆x + ∆y + ∆z = ε yx ∆x + ε yy ∆y + ε yz ∆z ∂x ∂y ∂z ∂u ∂u ∂u ∆u z = z ∆x + z ∆y + z ∆z = ε zx ∆x + ε zy ∆y + ε zz ∆z ∂x ∂y ∂z ∆u x = -10- .y. uz) is the displacement vector of P. In other words. so strain is actually a 3x3 second rank tensor.as the sample is deformed. Strain is the change in length of an object in direction i (i=x. z y O P u P’ x Let the origin be fixed as the sample is deformed. 2. y+uy. the displacement will be “proportional” to the distance from the origin: ∂u ε ij = i = constant (for homogeneous/linear system) ∂x j For another point Q with an original displacement (∆x. uy. Let the coordinates of point P before the deformation be (x.z).

with P at the origin. ∂x ∂u y ∆x = ε yx ∆x ∆u y = ∂x ∆u y ∆u y θ1 ≈ tan θ1 = ≈ = ε yx ∆x + ∆u x ∆x -11- .u+∆u Q ∆r Q’ ∆u ∆r P’ u P r=(x.y.direction.z) z O x 3. ∆y = ∆z = 0 ∴ ∆u x = ∂u x ∆x = ε xx ∆x ⇒ ε xx = extension per unit length along the x . Meaning of εij: y Let us consider a particular case when PQ1 is along (say) x-axis and PQ2 is along the y-axis. After the deformation: ∆uy Q2’ ∆u x Q2 θ2 P θ1 ∆x Q1’ ∆u y Q1 ∆ux For PQ1 .

exz. ezz. Problem of εij: Now consider the rotation of a sample along the z-axis: Q2’ Q2 θ2 P This will give a matrix of ε as: θ1 Q1’ Q1 ⎛0 −θ ⎜ ε ij = ⎜θ 1 ⎜0 0 ⎝ 0⎞ ⎟ 0⎟ 0⎟ ⎠ This has no physical sense because a rotation will not produce any deformation of the solid. Physical interpretation of the strain components: -12- . 4. ∆x = ∆z = 0 ∴ ∆u x = ∂u x ∆y = ε xy ∆x ∂y ∂u y ∆u y = ∆y = ε yy ∆x ⇒ ε yy = extension per unit length along the y . eij = eji.direction ∂y ∆u x ∆u θ 2 ≈ tan θ 2 = ≈ x = ε xy ∆y + ∆u y ∆y In summary. Obviously. they are: exx. we need to define the strain eij as: ε + ε ji e ij = ij 2 6. exy. 5. for PQ 2 . if i≠j. and eyz. εij = the angle between a line along the xi direction and the xi axis. eyy. εii = extension per unit length along the xi direction.Similarly. 7. To ensure a rotational motion will yield a zero matrix. There are only 6 independent strain components.

z’. with unit vectors x. Under strain. -13- . these will become new axis x’. 7.We can “decompose” any re-configuration (given by ε) into a rotation and a real strain (given by e): (εyx-εxy)/2 εxy = εyx + (εxy+εyx)/2 eii are the extensions per unit length parallel to the xi axis. Consider a sample of unit dimension (V=1). v ) ) ∆u x = ε xx ∆x + ε xy ∆y + ε xz ∆z ⎫ ˆ ⎧x' = (1 + ε xx )x + ε yx y + ε zx z ⎪ ⎪v ) ) ˆ ∆u y = ε yx ∆x + ε yy ∆y + ε yz ∆z ⎬ ⇒ ⎨ y' = ε xy x + (1 + ε yy )y + ε zy z v ) ) ⎪ ⎪z ˆ ∆u z = ε zx ∆x + ε zy ∆y + ε zz ∆z ⎭ ⎩ ' = ε xz x + ε yz y + (1 + ε zz )z 1 + ε xx ε yx v v v ∴ V' = x'⋅( y'×z' ) = ε xy 1 + ε yy (εxy+εyx)/2 ε xz = (1 + ε xx ) ε yz 1 + ε yy ε zx ε zy 1 + ε zz ε zy ε xy 1 + ε yy ε ε zy − ε yx xy + ε zx 1 + ε zz ε xz ε yz ε xz 1 + ε zz ε yz = (1 + ε xx )(1 + ε yy )(1 + ε zz ) − (1 + ε xx )ε zyε yz − ε yx ε xy (1 + ε zz ) + ε yx ε zyε xz + ε zx ε xy ε yz − ε zx (1 + ε yy )ε xz ≈ (1 + ε yy + ε yy + ε zz + ε xx ε yy + ε yyε zz + ε yyε zz ) − ε zyε yz − ε yx ε xy − ε zx ε xz ≈ 1 + ε yy + ε yy + ε zz = 1 + e yy + e yy + e zz In other words. the first order contribution to volume change when a material is under strain is the diagonal elements of the strain matrix. y’. z along the three side of the cube. y. the angle between them after deformation is π/2-2 eij. Dilation δ is defined as the fractional change in volume. eij (i ≠j) measure the tensor shear strain: if two line elements are drawn parallel to the xi and xj axes in the undeformed body. Note particularly that the tensor shear strain eij is one-half of the change in angle between the two elements.

1. 1 + ε xx V' = ε yx ε zx ∴ δ= IX. 2. 3. y σxy σyx x -14- .Now. there should be no torque acting on the sample. σxy here is the same as Xy in the textbook. For equilibrium. σij is written as Ij. Stress σij is the i-component of the force acting on the surface perpendicular to the j-axis: z σzz σyz σxz σzx σyy σyx σxy σxx y σzy x Note that in the textbook (Kittel 8th edition. σij is known as shear stress if i ≠ j. Stress ε xy ε xz 1 + ε yy ε yz ≈ 1 + e xx + e yy + e zz 1 + ε zz ε zy V'-V = e xx + e yy + e zz = Tr(e) V Stress is the force component acting on a unit area in the solid. 4. For example.

we have σij= σji. Hooke’s law states that for sufficient small deformations the strain is directly proportional to the stress. j Energy density U = 5. Like the case of strain. Energy stored in a spring is given by U=kx2/2. 6.σ3=σzz. e3=ezz. σzz. e2=eyy.For net torque equals to zero. C=S-1 (inverse matrix of S) is called the elastic stiffness (similar to the spring constant k in the common Hooke’s law): σ=Ce 4. C is a symmetric matrix with Cij=Cji. It has only (36-6)/2+6 =21 independent elements. we first calculate the work done (per unit volume) due to a small strain: dU = dW = ∑ σ i de i = ∑ C ij e j de i i Integrating with i. From this energy density. we require σxy= σyx. e4= eyz. σyy. Hooke’s Law – Elastic compliance and stiffness constants 1.σ6=σxy.σ4=σyz. Mathematically we require: ∂2U ∂2U = ∂e i ∂e j ∂e i ∂e j ⇒ C ij = C ji In other words. σ i: σ1=σxx. we will redefine their indices as follow: e i: e1=exx. and σyz. e5=ezx. σxz. j respect to dei: 1 ∑ C ij e j e i 2 i. e6=exy.σ2=σyy. In general. they are: σxx. X. In matrix notation: e=Sσ 3. σxy. σ5=σzx. The strain components are linear functions of the stress components: e i = ∑ S ijσ j j Sij is a six by six matrix. S is called the elastic compliance. we have 6 independent stress components. 2. For a solid. Since we have 6 independent e and 6 independent S. Hooke’s Law can be “derived”: -15- .

y.∂U 1 ∂ = ∂e i 2 ∂e i ∑ ij C ij e i e j = 1 ∑ (C ij + C ji )e i = 2 i ∑ i C ji e i 7. and z will affect the stress and strain in the following way (as an example): ex(-y) = -exy σx(-y) = -σxy 4. 3 2 4 z 1 x y Rotation along these four axes by 120o counterclockwise corresponds to transforming the x.±z) axes. y. 3. Note that there is a mistake in text (Kittel 8th edition that -x →z→y→-x and -x →y→z→-x actually represent the same axis of rotation (axis 2 in above figure. and x.) 2. one in anticlockwise direction and the other in clockwise (or 240o anticlockwise) direction. along all four axes of rotation. j -16- . and z in any way. y. XI. and add positive and negative signs to any of x.±y. The bulk modulus B is defined as: 1 dp 1 B=− ⇒ U = Bδ 2 V dV 2 and the compressibility K is defined as K=1/B.±y.±z) are surrounding the axis of rotation and then permute these (±x. Changing the sign of x. Energy density U = 1 ∑ C ij e j e i 2 i. Cubic crystals A cubic system has four three fold symmetry axes along the diagonal. 1. Considering rotation of 120o in both clockwise and anticlockwise directions. and z according to the following schemes: Axix 1: x →y→z→x Axix 2: -x →z→y→-x Axix 3: -x →z→-y→-x (or x →-z→y→x) Axix 4: x →z→-y→x (One simple way to imagine the above schemes is to ask which axes (±x. These operations mean we can permute x. y.

The longitudinal wave involves strain εii and hence the velocity is given by C11 v long = ρ Shear strain is involved in the longitudinal wave. the 21 independent Cij is reduced to only 3 under cubic symmetry.html”.lsbu. Hooke’s Law for cubic system: ⎛ σ xx ⎞ ⎛ C11 ⎜ ⎟ ⎜ ⎜ σ yy ⎟ ⎜ C12 ⎜σ ⎟ ⎜C ⎜ zz ⎟ = ⎜ 12 ⎜ σ yz ⎟ ⎜ 0 ⎜σ ⎟ ⎜ ⎜ zx ⎟ ⎜ 0 ⎜σ ⎟ ⎜ 0 ⎝ xy ⎠ ⎝ C12 C11 C12 0 0 0 C12 C12 C11 0 0 0 0 0 0 C 44 0 0 0 0 0 0 C 44 0 0 ⎞⎛ e xx ⎞ ⎟⎜ ⎟ 0 ⎟⎜ e yy ⎟ 0 ⎟⎜ e zz ⎟ ⎟⎜ ⎟ 0 ⎟⎜ e yz ⎟ ⎜e ⎟ 0 ⎟ ⎟⎜ zx ⎟ ⎜ ⎟ C 44 ⎟ ⎠⎝ e xy ⎠ 5. so the transverse wave speed is given by v trans = C 44 ρ Appendix: Ice structure See “http://www. Consider a sound wave traveling along the x-axis of a cubic sample. -17- .pdf in this folder.U has to remain invariant under above operations. and the most general form that can reserve this invariant is: U= 1 1 2 2 2 2 2 2 C11 (e xx + e yy + e zz ) + C 44 (e xy + e yz + e zx ) + C12 (e xx e yy + e yy e zz + e zz e xx ) 2 2 In other words.uk/water/ice1h. or ice_structure. The wave can be longitudinal or transverse.ac.

-18- .