Cubic Lattices and Close Packing

1 Close-packing of identical spheres

Crystals are of course three-dimensional objects, but we will begin by exploring the properties of arrays in twodimensional space. his will make it easier to de!elop some of the basic ideas without the added complication of getting you to !isuali"e in #-$ % something that often re&uires a bit of practice. 'uppose you ha!e a do"en or so marbles. (ow can you arrange them in a single compact layer on a table top) *b!iously, they must be in contact with each other in order to minimi"e the area they co!er. +t turns out that there are two efficient ways of achie!ing this,

he essential difference here is that any marble within the interior of the s&uare-packed array is in contact with four other marbles, while this number rises to six in the hexagonal-packed arrangement. +t should also be apparent that the latter scheme co!ers a smaller area -contains less empty space. and is therefore a more efficient packing arrangement. +f you are good at geometry, you can show that s&uare packing co!ers /0 percent of the area, while hexagonal packing yields 11 percent co!erage. +f we go from the world of marbles to that of atoms, which kind of packing would the atoms of a gi!en element prefer) +f the atoms are identical and are bound together mainly by dispersion forces which are completely nondirectional, they will fa!or a structure in which as many atoms can be in direct contact as possible. his will, of course, be the hexagonal arrangement. $irected chemical bonds between atoms ha!e a major effect on the packing. he !ersion of hexagonal packing shown at the right occurs in the form of carbon known as graphite which forms 2-dimensional sheets. 3ach carbon atom within a sheet is bonded to three other carbon atoms. he result is just the basic hexagonal structure with some atoms missing. he coordination number of # reflects the sp2-hybridi"ation of carbon in graphite, resulting in plane-trigonal bonding and thus the sheet structure. 4djacent sheets are bound by weak dispersion forces, allowing the sheets to slip o!er one another and gi!ing rise to the lubricating and flaking properties of graphite.

2 Lattices
he underlying order of a crystalline solid can be represented by an array of regularly spaced points that indicate the locations of the crystal5s basic structural units. his array is called a crystal lattice.

Crystal lattices can be thought of as being built up from repeating units containing just a few atoms. hese repeating units act much as a rubber stamp, press it on the paper, mo!e -6translate6. it by an amount e&ual to the lattice spacing, and stamp the paper again.

he gray circles represent a s&uare array of lattice points.

he orange s&uare is the simplest unit cell that can be used to define the 2- 7uilding out the lattice by mo!ing dimensional lattice. -6translating6. the unit cell in a series of steps,

4lthough real crystals do not actually grow in this manner, this process is conceptually important because it allows us to classify a lattice type in terms of the simple repeating unit that is used to 6build6 it. 8e call this shape the unit cell. 4ny number of primiti!e shapes can be used to define the unit cell of a gi!en crystal lattice. he one that is actually used is largely a matter of con!enience, and it may contain a lattice point in its center, as you see in two of the unit cells shown here. +n general, the best unit cell is the simplest one that is capable of building out the lattice.

'hown abo!e are unit cells for the close-packed s&uare and hexagonal lattices we discussed near the start of this lesson. 4lthough we could use a hexagon for the second of these lattices, the rhombus is preferred because it is simpler. 9otice that in both of these lattices, the corners of the unit cells are centered on a lattice point. his means that an atom or molecule located on this point in a real crystal lattice is shared with its neighboring cells. 4s is shown more clearly here for a two-dimensional s&uare-packed lattice, a single unit cell can claim 6ownership6 of only one&uarter of each molecule, and thus 6contains6 : ; < = 1 molecule.

he unit cell of the graphite form of carbon is also a rhombus, in keeping with the hexagonal symmetry of this arrangement.

9otice that to generate this structure from the unit cell, we need to shift the cell in both the x- and y- directions in order to lea!e empty spaces at the correct spots. 8e could alternati!ely use regular hexagons as the unit cells, but the x>y shifts would still be re&uired, so the simpler rhombus is usually preferred. 4s you will see in the next section, the empty spaces within these unit cells play an important role when we mo!e from two- to three-dimensional lattices.

# Cubic crystals
+n order to keep this lesson within reasonable bounds, we are limiting it mostly to crystals belonging to the socalled cubic system. +n doing so, we can de!elop the major concepts that are useful for understanding more complicated structures -as if there are not enough complications in cubics alone?. 7ut in addition, it happens that cubic crystals are !ery commonly encountered@ most metallic elements ha!e cubic structures, and so does ordinary salt, sodium chloride. 8e usually think of a cubic shape in terms of the e&uality of its edge lengths and the 1AB angles between its sides, but there is another way of classifying shapes that chemists find !ery useful. his is to look at what geometric transformations -such as rotations around an axis. we can perform that lea!e the appearance unchanged. Cor example, you can rotate a cube 1AB around an axis perpendicular to any pair of its six faces without making any apparent change to it. 8e say that the cube possesses three mutually perpendicular four-fold rotational axes, abbre!iated C: axes. 7ut if you think about it, a cube can also be rotated around the axes that extend between opposite corners@ in this case, it takes three 12AB rotations to go through a complete circle, so these axes -also four in number. are three-fold or C# axes. Cubic crystals belong to one of the se!en crystal systems whose lattice points can be extended indefinitely to fill three-dimensional space and which can be constructed by successi!e translations -mo!ements. of a primiti!e unit cell in three dimensions. 4s we will see below, the cubic system, as well as some of the others, can ha!e !ariants in which additional lattice points can be placed at the center of the unit or at the center of each face.

he three types of cubic lattices

he three 7ra!ais lattices which form the cubic crystal system are shown here.

'tructural examples of all three are known, with body- and face-centered -7CC and CCC. being much more common@ most metallic elements crystalli"e in one of these latter forms. 7ut although the simple cubic structure is uncommon by itself, it turns out that many 7CC and CCC structures composed of ions can be regarded as interpenetrating combinations of two simple cubic lattices, one made up of positi!e ions and the other of negati!e ions. 9otice that only the CCC structure, which we will describe below, is a close-packed lattice within the cubic system.

: Close-packed lattices in three dimensions

Close-packed lattices allow the maximum amount of interaction between atoms. +f these interactions are mainly attracti!e, then close-packing usually leads to more energetically stable structures. hese lattice geometries are widely seen in metallic, atomic, and simple ionic crystals. 4s we pointed out abo!e, hexagonal packing of a single layer is more efficient than s&uare-packing, so this is where we begin. +magine that we start with the single layer of green atoms shown below. 8e will call this the 4 layer. +f we

place a second layer of atoms -orange. on top of the 4-layer, we would expect the atoms of the new layer to nestle in the hollows in the first layer. 7ut if all the atoms are identical, only some of these !oid spaces will be accessible.

+n the diagram on the left, notice that there are two classes of !oid spaces between the 4 atoms@ one set -colored blue. has a !ertex pointing up, while the other set -not colored. has down-pointing !ertices. 3ach !oid space constitutes a depression in which atoms of a second layer -the 7-layer. can nest. he two sets of !oid spaces are completely e&ui!alent, but only one of these sets can be occupied by a second layer of atoms whose si"e is similar to those in the bottom layer. +n the illustration on the right abo!e we ha!e arbitrarily placed the 7-layer atoms in the blue !oids, but could just as well ha!e selected the white ones.

wo choices for the third layer lead to two different close-packed lattice types
9ow consider what happens when we lay down a third layer of atoms. hese will fit into the !oid spaces within the 7-layer. 4s before, there are two sets of these positions, but unlike the case described abo!e, they are not e&ui!alent.

he atoms in the third layer are represented by open blue circles in order to a!oid obscuring the layers underneath. +n the illustration on the left, this third layer is placed on the 7-layer at locations that are directly abo!e the atoms of the 4-layer, so our third layer is just a another 4 layer. +f we add still more layers, the !ertical se&uence 4-7-4-7-4-74... repeats indefinitely. +n the diagram on the right abo!e, the blue atoms ha!e been placed abo!e the white -unoccupied. !oid spaces in layer 4. 7ecause this third layer is displaced hori"ontally -in our !iew. from layer 4, we will call it layer C. 4s we add more layers of atoms, the se&uence of layers is 4-7-C-4-7-C-4-7-C..., so we call it 47C packing.

Cor the purposes of clarity, only three atoms of the 4 and C layers are shown in these diagrams. 7ut in reality, each layer consists of an extended hexagonal array@ the two layers are simply displaced from one another. hese two diagrams that show exploded !iews of the !ertical stacking further illustrate the rather small fundamental difference between these two arrangements% but, as you will see below, they ha!e widely di!ergent structural conse&uences. 9ote the opposite orientations of the 4 and C layers

he (exagonal closed-packed structure

he (CP stacking shown on the left just abo!e takes us out of the cubic crystal system into the hexagonal system, so we will not say much more about it here except to point out each atom has 12 nearest neighbors, six in its own layer, and three in each layer abo!e and below it.

he cubic close-packed structure

(ere on the left we reproduce the CCC structure that was shown abo!e. Dou will notice that the 7-layer atoms form a hexagon, but this is a cubic structure. (ow can this be) he answer is that the CCC stack is inclined with respect to the faces of the cube, and is in fact coincident with one of the three-fold axes that passes through opposite corners. +t re&uires a bit of study to see the relationship, and we ha!e pro!ided two !iews to help you. he one on the left shows the cube in the normal isometric projection@ the one on the right looks down upon the top of the cube at a slightly inclined angle.

7oth the CCP and (CP structures fill /: percent of the a!ailable space when the atoms ha!e the same si"e. Dou should see that the two shaded planes cutting along diagonals within the interior of the cube contain atoms of different colors, meaning that they belong to different layers of the CCP stack. 3ach plane contains three atoms from the 7 layer and three from the C layer, thus reducing the symmetry to C#, which a cubic lattice must ha!e.

he CCC unit cell

he figure at the right shows the the face-centered cubic unit cell of a cubic-close packed lattice. (ow many atoms are contained in a unit cell) 3ach corner atom is shared with eight adjacent unit cells and so a single unit cell can claim only 1E0 of each of the eight corner atoms. 'imilarly, each of the six atoms centered on a face is only half-owned by the cell. he grand total is then -0 ; 1E0. > -F ; G. = : atoms per unit cell.

H +nterstitial !oid spaces

he atoms in each layer in these close-packing stacks sit in a depression in the layer below it. 4s we explained abo!e, these !oid spaces are not completely filled. -+t is geometrically impossible for more than two identical spheres to be in contact at a single point.. 8e will see later that these interstitial !oid spaces can sometimes accommodate additional -but generally smaller. atoms or ions. +f we look down on top of two layers of close-packed spheres, we can pick out two classes of !oid spaces which we call tetrahedral and octahedral holes.

etrahedral holes
+f we direct our attention to a region in the abo!e diagram where a single atom is in contact with the three atoms in the layers directly below it, the !oid space is known as a tetrahedral hole. 4 similar space will be be found between this single atom and the three atoms -not shown. that would lie on top of it in an extended lattice. 4ny interstitial atom that might occupy this site will interact with the four atoms surrounding it, so this is also called a fourcoordinate interstitial space.

$on5t be misled by this name@ the boundaries of the !oid space are spherical sections, not tetrahedra. he tetrahedron is just an imaginary construction whose four corners point to the centers of the four atoms that are in contact.

*ctahedral holes
'imilarly, when two sets of three trigonally-oriented spheres are in close-packed contact, they will be oriented FAB apart and the centers of the spheres will define the six corners of an imaginary octahedron centered in the !oid space between the two layers, so we call these octahedral holes or six-coordinate interstitial sites. *ctahedral sites are larger than tetrahedral sites. 4n octahedron has six corners and eight sides. 8e usually draw octahedra as a double s&uare pyramid standing on one corner -left., but in order to !isuali"e the octahedral shape in a close-packed lattice, it is better to think of the octahedron as lying on one of its faces -right..

3ach sphere in a close-packed lattice is associated with one octahedral site, whereas there are only half as many tetrahedral sites. his can be seen in this diagram that shows the central atom in the 7 layer in alignment with the hollows in the C and 4 layers abo!e and below.

he face-centered cubic unit cell contains a single octahedral hole within itself, but octahedral holes shared with adjacent cells exist at the centers of each edge. 3ach of these twel!e edge-located sites is shared with four adjacent cells, and thus contributes -12 ; <. = # atoms to the cell. 4dded to the single hole contained in the middle of the cell, this makes a total of : octahedral sites per unit cell. his is the same as the number we calculated abo!e for the number of atoms in the cell.

F 'ome common cubic close-packed structures

+t can be shown from elementary trigonometry that an atom will fit exactly into an octahedral site if its radius is A.:1: as great as that of the host atoms. he corresponding figure for the smaller tetrahedral holes is A.22H. Iany pure metals and compounds form face-centered cubic -cubic close- packed. structures. he existence of tetrahedral and octahedral holes in these lattices presents an opportunity for 6foreign6 atoms to occupy some or all of these interstitial sites. +n order to retain close-packing, the interstitial atoms must be small enough to fit into these holes without disrupting the host CCP lattice. 8hen these atoms are too large, which is commonly the case in ionic compounds, the atoms in the interstitial sites will push the host atoms apart so that the face-centered cubic lattice is somewhat opened up and loses its close-packing character.

he rock-salt structure
4lkali halides that crystalli"e with the 6rock-salt6 structure exemplified by sodium chloride can be regarded either as a CCC structure of one kind of ion in which the octahedral holes are occupied by ions of opposite charge, or as two interpenetrating CCC lattices made up of the two kinds of ions. he two shaded octahedra illustrate the identical coordination of the two kinds of ions@ each atom or ion of a gi!en kind is surrounded by six of the opposite kind, resulting in a coordination expressed as -F,F..

(ow many 9aCl units are contained in the unit cell) +f we ignore the atoms that were placed outside the cell in order to construct the octahedra, you should be able to count fourteen 6orange6 atoms and thirteen 6blue6 ones. 7ut many of these are shared with adjacent unit cells.

4n atom at the corner of the cube is shared by eight adjacent cubes, and thus makes a 1E0 contribution to any one cell. 'imilarly, the center of an edge is common to four other cells, and an atom centered in a face is shared with two cells. aking all this into consideration, you should be able to confirm the following tally showing that there are four 47 units in a unit cell of this kind. *range 0 at corners, 0 x 1E0 = 1 F at face centers, F x G = # total, : 1 at body center = 1 total, : 7lue 12 at edge centers, 12 x < = #

+f we take into consideration the actual si"es of the ions -9a> = 11F pm, ClJ = 1F/ pm., it is apparent that neither ion will fit into the octahedral holes with a CCP lattice composed of the other ion, so the actual structure of 9aCl is somewhat expanded beyond the close-packed model. he space-filling model on the right depicts a face-centered cubic unit cell of chloride ions -purple., with the sodium ions -green. occupying the octahedral sites.

he "inc-blende structure, using some tetrahedral holes

'ince there are two tetrahedral sites for e!ery atom in a close-packed lattice, we can ha!e binary compounds of 1,1 or 1,2 stoichiometry depending on whether half or all of the tetrahedral holes are occupied. Kinc-blende is the mineralogical name for "inc sulfide, Kn'. 4n impure form known as sphalerite is the major ore from which "inc is obtained. his structure consists essentially of a CCC -CCP. lattice of sulfur atoms -orange. -e&ui!alent to the lattice of chloride ions in 9aCl. in which "inc ions -green. occupy half of the tetrahedral sites. 4s with any CCC lattice, there are four atoms of sulfur per unit cell, and the the four "inc atoms are totally contained in the unit cell. 3ach atom in this structure has four nearest neighbors, and is thus tetrahedrally coordinated. +t is interesting to note that if all the atoms are replaced with carbon, this would correspond to the diamond structure.

he fluorite structure, all tetrahedral sites occupied

Cluorite, CaC2, ha!ing twice as many ions of fluoride as of calcium, makes use of all eight tetrahedral holes in the CPP lattice of calcium ions -orange. depicted here. o help you understand this structure, we ha!e shown some of the octahedral sites in the next cell on the right@ you can see that the calcium ion at 4 is surrounded by eight fluoride ions, and this is of course the case for all of the calcium sites. 'ince each fluoride ion has four nearestneighbor calcium ions, the coordination in this structure is described as -0,:..

4lthough the radii of the two ions -CJ= 11/ pm, Ca2> = 12F pm does not allow true close packing, they are similar enough that one could just as well describe the structure as a CCC lattice of fluoride ions with calcium ions in the octahedral holes.

/ 'imple- and body-centered cubic structures

+n 'ection : we saw that the only cubic lattice that can allow close packing is the face-centered cubic structure. he simplest of the three cubic lattice types, the simple cubic lattice, lacks the hexagonally-arranged layers that are re&uired for close packing. 7ut as shown in this exploded !iew, the !oid space between the two s&uare-packed layers of this cell constitutes an octahedral hole that can accommodate another atom, yielding a packing arrangement that in fa!orable cases can approximate true close-packing. 3ach second-layer 7 atom -blue. resides within the unit cell defined the 4 layers abo!e and below it. he 4 and 7 atoms can be of the same kind or they can be different. +f they are the same, we ha!e a body-centered cubic lattice. +f they are different, and especially if they are oppositely-charged ions -as in the CsCl structure., there are si"e restrictions, if the 7 atom is too large to fit into the interstitial space, or if it is so small that the 4 layers -which all carry the same electric charge. come into contact without sufficient 4-7 coulombic attractions, this structural arrangement may not be stable.

he cesium chloride structure

CsCl is the common model for the 7CC structure. 4s with so many other structures in!ol!ing two different atoms or ions, we can regard the same basic structure in different ways. hus if we look beyond a single unit cell, we see that CsCl can be represented as two interpenetrating simple cubic lattices in which each atom occupies an octahedral hole within the cubes of the other lattice.

7ack to 'tates of Iatter 7ack to chem1 -'tephen LowerE'imon Crasier Lni!ersity.