Introduction To The Structure and The Structural Role of Elements in Glasses

Introduction
to the structure and the

structural role of elements in glasses
Laurent Cormier
Institut de Minéralogie, Physique des Matériaux et Cosmochimie
Université Pierre et Marie Curie – CNRS
Paris, France
cormier@impmc.upmc.fr
Ecole GDR Verres Cargèse

Glass forming ability (GFA)
In principle almost any substance can be made into a glass by cooling it from
the liquid state if cooled sufficiently fast to prevent crystallization. In practice,
however, it seems useful to make some classification scheme.
No theories able to predict which materials are able to form a glass, and under
what conditions
From ’Chemistry of glasses’

Paul (2012)
Glass families and glass forming ability
Glass formation results when

• Liquids are cooled to below TM (TL) sufficiently fast to avoid crystallization
o Nucleation of crystalline seeds are avoided
o Growth of nuclei into crystallites (crystals) is avoided
Þ Kinetic theory of glass formation (Turnbull and Cohen, 1960)
From glass to crystal -Nucleation, growth and phase separation: from research to
applications (2017)
D.R. Neuville, L. Cornier, D. Caurant, L. Montagne
• Liquid is frustrated by internal structure that hinders both events

Competition between crystalline
growth and cooling of the melt.
Quantitative estimate through time- Tf liquid
temperature-transformation (TTT) Supercooled

liquid
curves
How much time does it take at any Tn

solid
one temperature for a given fraction liquid +solid

Tg
of the liquid to transform (nucleate
and grow) into a crystal?
crystal
Critical cooling rate Rc = the nose of
the TTT curve
Tf − Tn
RC =
tn
Good glass formers:

q = dT/dt ~ 10-3 K/s
€
Bad glass formers: tn td tf
q = dT/dt > 106 K/s
Competition between crystalline
growth and cooling of the melt.
Good glass formers:

q = dT/dt ~ 10-3 K/s
Bad glass formers:
q = dT/dt > 106 K/s
Quantitative estimate through time-

temperature-transformation (TTT)
curves
How much time does it take at any

one temperature for a given fraction
of the liquid to transform (nucleate
and grow) into a crystal?
Uhlmann, J. Phys. Coll., 43 (1982) C9-175-C9-190
Critical cooling rate Rc = the nose of
the TTT curve
T − Tn
RC = f
tn
Glass families and
glass forming ability
In a multicomponent material like the SiO2-
Na2O glass (c), the question of the region of
glass formation arises:
how much Na2O can be added to SiO2, while
still keeping the glass forming ability of the
system?
What is the region of glass formation ?

Glass forming region
Determination of glass forming region after extensive studies in compositions
Musgrave et al., J. Non-Cryst. Solids, 386 (2014) 61
Richet et al., Chem. Geol. 225 (2006) 388
Neuville et al., Chem. Geol. 229 (2006) 173

Glass forming region
In a phase diagram, glasses are favored
near eutectics because these freezing
depressions bring the liquid to higher
viscosities at lower temperatures
Correlation between critical cooling rate

and viscosity along the liquidus branche
Richet et al., Chem. Geol. 225 (2006) 388

Þ Daniel Neuville
Structural approach to glass-forming ability (GFA)
What internal structures promote glass formation?
How can structures be developed that increase the viscosity and

frustrate crystallization processes?
Structural approach to glass formation
Several models encompass most of the relevant aspects which are known to lead
to glass formation:
• Based on coordination number

– Goldschmidt’s radius ratio
– Zachariasen’s random network theory
• Based on bond type

– Smekal’s mixed bonding rule
– Stanworth’s electronegativity rule
• Based on bond strength

– Sun’s single bond strength criterion
• Based on field strength

– Dietzel’s field strength
• Based on Mott’s rule

– Phillips
Different structural ranges
Short range order (<3 Å):

– coordination, bond length, bond angle
– homopolar (-Se - Se- , -C - C-, -As - As) vs.
heteropolar (Si - O, B - O, Ge - S )
Medium (intermediate) range

order (3 – 10 Å typically):
– angles between structural units
– connectivity between structural units
(linkages by corner, edge, face)
– dimensionnality, rings
Almost no long range order

(no periodicity!) :
– phase separation
– inhomogeneities
Topological and chemical disorder
Salmon et al., Nature 435 (2005) 75-78

Structural model
A good structural model should
- explain structure-properties relationships

- describe atomic arrangements at short and medium range
- have general concepts applicable to a large number of systems
Initially, idea of micro-crystallites (quasi-crystals) but it was demonstrated

that the size of these crystals is 7-8 Å (size of the unit cell) !
crystallite model of Lebedev

to glass formation:





– Phillips
Goldschmidt’s radius ratio criterion
for a simple oxide of formula MmOn (M cation), glass formation is only
possible when the ratio of atomic radii, ra/rc (where c denotes cation, and a anion)
falls between 0.2 to 0.4
the cation / anion ratio determines how many anions can be packed around
a given cation, i.e. the coordination number of the compound. Most crystals
with a cation / anion radius ratio of 0.2 to 0.4 have a co-ordination number
(CN) of 4 with anions at the corners of a tetrahedron
=
tetrahedral arrangement of oxygen ions around a cation M necessary
for glass formation
assumed the oxide is purely ionic (not strictly correct as many glass-forming
oxides have covalent character, e.g. SiO2)
Victor M. Goldschmidt (1888-1947)

Goldschmidt’s radius ratio criterion
condition is fulfilled in the case of SiO2, B2O3, P2O5, GeO2 and BeF2
BeO with rBe/rO ~0.221 but does not form a glass

to glass formation:





– Phillips
Zachariasen’s random network theory
This theory, strictly applicable to oxide glasses only in its original form,
sometimes referred to as the crystallochemical theory; it led to the model
known as the 3-D continuous random network (3-D CRN) model.
Basic postulates:
- interatomic forces similar in glass and corresponding crystal
- glass is in a slightly higher energy state
- nearest neighbor coordination polyhedra similar in glass and crystal
- nature of interatomic bonds similar in glass and corresponding crystal
William Houlder Zachariasen (1906-1979)
Zachariasen, W.H., The atomic arrangement in glass. J. Am. Chem. Soc., 54 (1932) 3841
Zachariasen’s rules for glass formation
1. Each oxygen atom is linked (bonded) to no
more than two glass-forming cations (e.g. Si4+);
2. Oxygen coordination number (CN) around
glass-forming cation is small: 3 or 4;
3. Cation polyhedra share corners, not edges or
faces;
4. The polyhedral structural units form a 3-D
continuous random network in which every
polyhedron shares at least 3 corners with its
neighbors.
Network former: Si, Ge, P, B, …

Oxygen
A2O3, AO2, and A2O5 oxides met the above rules => good glass former
A2O, AO, AO3, A2O7, etc., no glasses could be formed

Zachariasen model (1932)
c-SiO2 a-SiO2
STEM images
Huang et al., Nano Lett. 12 (2012)1081

Structural model of covalent glasses
Structural units
In silicate such as SiO2: SiO4 tetrahedra
Same basic structural unit for the glass and the
crystal
Electronic structure:
Si : (Ne)3s23p2 : 4 valence electrons => ions Si4+ (can share 4 bonds)

O : (He)2s22p4 : 6 valence electron s => ions O2- (can share 2 bonds)
B : (He)2s22p1 : 3 valence electron => ions B3+ (can share 3 bonds)

O : (He)2s22p4 : 6 valence electron => ions O2-(can share 2 bonds)
Borate glass
B2O3 not composed primarily of CRN’s (continuous random network) of individual BO4
and BO3 units. These small units form structural grouping such as boroxol, diborate …
that exist in the crystalline compounds of the particular borate system. These larger
(but sill quite small) units are then connected randomly to each other to form the glass
structure.
Intermediate between the micro-crystallite and the CRN models
About ~75% of B atoms are in boroxol rings

Neutron diffraction: Hannon et al., J. Non-Cryst. Solids 177 (1994) 299
Ab initio simulations: Ferlat et al. Phys. Rev. Lett. 101 (2008) 065504
the boroxols allow one to maintain a low-

energy structure while keeping a liquidlike
density
Þ Importance of medium range order in
GFA ?
Þ Gerald Lelong
Medium range order – inhomogeneities in glasses
Introduce for the first time the notion of inhomogeneities in glasses
crystallite model of Lebedev

Multicomponent oxide glasses
Zachariasen's rules do not consider at all modified oxides or

multicomponent systems, or even non-oxide glasses
Zachariasen–Warren network theory
Bertram Eugene Warren

1902-1991
Non-network formers (alkali, alkaline-earth,
Alkali silicate glasses transition elements) decrease the network
connectivity by forming non-bridging oxygens (NBO)
(≠ bridging oxygens BO)
Þ Network modifier
Qn species
n = number of bridging oxygens by tetrahedra
Random network
Network modifiers have

- high coordination numbers
- a random distribution in the
glass interstices
No information on medium range order

Modified random network - MRN (Greaves, 1985)
Zones rich with network modifiers
Zones rich with network formers
Relationships with conductivity, alteration etc

Aluminosilicate glasses
Al substitute to Si in tetrahedral position
Al : (Ne)3s23p1 : 3 valence electrons => ions Al3+
(AlO4)- charge electroneutrality ensures by the presence of alkali or alkaline earth
Similar for (BO4)-

Role of non-network forming cations
Cations connected to NBO Cations connected to BO

associated to the negative And acting as charge compensator
charge of O- near (AlO4)-, (BO4)- …
= network modifier = charge compensator
Phosphate glasses
P in tetrahedral position
P : (Ne)3s23p3 : 5 valence electron => ions P5+
Terminal oxygen
Bridging oxygen
Oxygen position in glass structure:

- bridging oxygen
- non-bridging oxygen Þ Francisco Munoz & Lionel Montagne
- terminal oxygen
Invert glasses
Y = NBO/T = Nbre d’O pontant par tétraèdre (Y=6-200/p with p the mol% of
SiO2)
SiO2 Y=4 3D network

R2O-2SiO2 Y=3
R2O-SiO2 Y=2 (metasilicate = SiO4 chains)
After Zachariasen’s hypothesis, glasses with Y<3 are

not possible
Oxygen position in glass structure:
For Y<2, it is named invert glass
- bridging oxygen
- non-bridging oxygen
Importance of free oxygens in
- terminal oxygen
those compositions
- free oxygen
Free-oxygen : oxygen not bonded
to any network-former
Þ Louis Hennet
Þ Grant Henderson
Trap & Stevels, Phys. Chem. Glasses, 1 (1960) 181
Glasses with non glass formers
Neither R2O or TiO2 can form a glass individually

But possibility to form R2O-TiO2 glasses
(1-x)R2O-xTiO2 with x=34-75 mol% and R=K, Rb, Cs
In particular orthotitanate : 2R2O-TiO2
Which one is the glass former ?

Network former or modifier ? The case of Pb2+
xPbO (100 x)SiO2
x = 90, 67, 50, 33, 25
Netwok
former at
Modifier at high PbO
low PbO content
content
40-60 mol% PbO

Takahashi et al., J. Am. Ceram. Soc., 88 (2005) 1591
Network former or modifier ? The case of Pb2+
Low PbO content
- PbO3+3 unit is dominant
- Pb2O4 units participate in the glass network constructed
by SiO4 tetrahedra even at low
PbO content
Netwok
former at
Modifier at high PbO
low PbO content
content
High PbO content

PbO acts as a network former
consisting of PbO3 trigonal pyramids
Case of Pb
High PbO content

PbO acts as a network former consisting
of PbO3 trigonal pyramids
PbO3 trigonal pyramids are linked to each

other by edge sharing to form Pb–O–Pb
network => Pb2O4 units
6p2 Ion Pb2+
Pb atom tends to have small coordination numbers differing from other

divalent elements
Þ reason that PbO is a good glass forming material ?
Limitations of Zachariasen’s rules
- apply only to oxide glasses
- there are several exceptions to the rules, such as:

• non-oxide glasses (e.g. fluoride and metallic glasses)
• some oxides ruled out by Zachariasen, such as Al-lime can in fact form glasses
• chain-like glass structures (e.g., metaphosphate glasses)
Strong points of Zachariasen’s model
- predicts the existence of the main oxide glass formers (SiO2, GeO2, B2O3, P2O5,
...) and glass modifiers (Na2O, CaO, ...)
- suggests the existence of bridging oxygen (BO) and non-bridging oxygen
(NBO) species
Several models encompass most of the relevant aspects which are known to lead to glass
formation:





– Phillips
Smekal’s mixing bond
Pure covalent bonds incompatible with random arrangement (sharply defined

bond-lengths and bond-angles)
Purely ionic or metallic bonds lack any directional characteristics
Þ Presence of ‘mixed’ chemical bonding necessary
• inorganic compounds, e.g. SiO2, B2O3, where the A-O bonds are partly
covalent and partly ionic
• Elements (S, Se) having chain structures with covalent bonds within the chain
and van der Waal’s forces between the chains
• organic compounds containing large molecules with covalent bonds within
the molecule and van der Waal’s forces between them
Smekal, J. Soc. Glass. Technol. 35 (1951) 411

Adolf Gustaf Smekal (1895–1959)
Stanworth’s electronegativity criterion
the electronegativity of the cations, of which the oxides are glass-forming, fall
within a certain range between 1.90 and 2.20
Stanworth, J. Soc. Glass Technol. 36 (1952) 217

Mixed bond :
Si-O : ionic character of ~45%
Similar for other network formers
But
Be-F : 75% ionic character but BeF2 form a glass
Ge-Se : 12% ionic character but GeS2 form a glass
As-S : 5% ionic character but As2S3 form a glass
From ’Le verre’ Scholze (1980)

Stoch, Glass Phys. Chem. 27 (2001) 167
Mixed bond :
Si-O : ionic character of ~45%
Similar for other network formers
But
Be-F : 75% ionic character but BeF2 form a glass
Ge-Se : 12% ionic character but GeS2 form a glass
As-S : 5% ionic character but As2S3 form a glass
From ’Le verre’ Scholze (1980)

to glass formation:





– Phillips
Sun’s bond strength model
Glass formation is brought about by both

• Connectivity of bridge bonds
• Strong bonds between atoms (ions)
Sun classified oxide according to their bond strengths

• Glass formers form strong bonds to oxygen – rigid network ,high
viscosity
• Modifiers form weak bonds to oxygen - disrupt, modify network
• Intermediates form intermediate bonds to oxygen – can’t form glasses
on their own, but aid with other oxides to form glasses
Sun’s criterion: energy criterion
- Establishes a correlation between glass forming tendency and the bond
energies of elements with the anion atom in the glass
Dissociation energy of oxides into gaseous elements:
EdMOx =|1/m∆Hof (MmOn,c)–∆Hof (M,g)–n/m∆Hof (O,g)|
Ed refers to a mol of M in MOn/m.
Eb = Ed / CN is called the single bond strength
The basic idea behind this model is that, when a melt is quenched to form a
glass, the stronger the M-O bonds, the more difficult are the structural
rearrangements necessary for crystallization and, hence, the easier is glass
formation
Sun, J. Am. Ceram. Soc. 30 (1947) 277

The higher the bond strength, the better the glass former
• Single bond strength > 80kcal/mol => network former

Greater than 80 kcal/mole bond strength with oxygen
B2O3, SiO2, Geo2, P2O5, Al2O5….
• Single bond strength < 60kcal/mol => network modifier

Less than 60 kcal/mole bond strength with oxygen
Li2O, Na2O, K2O, MgO, CaO….
• Single bond strength 60 ≤ strength ≥ 80 => intermediate

Between 60 to 80 kcal/mol bond strength with oxygen
TiO2, ZnO, PbO….
From ’Fundamentals of inorganic glasses’
Varshneya (2013)
Case of Al3+
Al2O3 satisfied Zachariasen’s criteria but Al2O3 does not form a glass
According to Sun’s criteria
Ed = 317-402 kcals
CN =4 79 < ESun < 101kcal mol-1 network former

CN =5 63 < ESun < 80kcal mol-1 ?
CN =6 53 < ESun < 67kcal mol-1 modifier ?
No Al2O3 glass can be formed
But formation of 3CaO-2Al2O3 glass
Þ Louis Hennet
Limitations of Sun’s criteria
- Exclusion of the chalcogenide glasses

Bond strength ~40kcal.mol along the chains (covalent bond) and less between
the chains (van der Waals forces)
Strong points of Sun’s criteria
- Introduce the notion of intermediate

Structural role
Cations in the glass were catagorized according to their role in the glass network
Network former
- Can form a glass network alone
- Strong directional bonding
- Example: Si4+, B3+, P5+, Ge4+, As3+, Be2+, with CN of 3 or 4
Network modifier
- Break the linkages between network formers
- More ionic bonding
- Example: Na+, K+, Ca2+, Ba2+, with CN ≥ 6
Intermediates (conditional network former)

- May reinforce (CN = 4) or further loosen the network further (CN 6 to 8)
- Can substitute to a network former but cannot form a glass per se
- Example: Al2O3, TiO2, Ga2O3, As2O3, Sb2O3, Bi2O3, TeO2, V2O5, MoO3, WO3
Structural role
Structural role of constituents; in view of Coulombic interaction

An interesting intermediate class of oxide including TeO2, MoO3,
Bi2O3, Al2O3, Ga2O3 and V2O5 do not by themselves form glasses but
will do so when mixed with other (modifier) oxides
Rawson’s criteria
- Ratio of the bond strength to the energy available at the freezing point (~3/2 Tm)
- Glass formation is better correlated with Eb/Tm, where Tm is the melting

temperature, and Eb/Tm > 0.05 for glass forming systems
- The higher the value, the lower the probability for bonds to break at Tm, and
hence the higher the tendency for glass formation
- Point out the significance of the liquidus temperature

Þ eutectic favors glass formation
Vitrification if high bond strength with a liquidus temperature with the lowest
possible melting temperature
SiO2 : Eb = 106 kcals mol-1 Tm=1600°C Eb/Tm = 0.066

ZrO2 : Eb = 81 kcals mol-1 Tm=2715°C Eb/Tm = 0.030 => cannot form glass due to its
high melting point
B2O3 : Eb = 119 kcals mol-1 Tm=450°C Eb/Tm = 0.264 => never crystallizes
Rawson, Proc. IV Internat. Congr. on Glass (1956)

Extension of Rawson’s criteria
ThRGFA = ESun/(Tm.Cp)
Boubata Bull. Mater. Sci. 36(2013)457

Extension of Rawson’s criteria
ThRGFA = ESun/(Tm.Cp)
Modifier Intermediate Former
Boubata Bull. Mater. Sci. 36(2013)457

to glass formation:





– Phillips
Dietzeld and field strength criteria
Sun classifies Al as both a network former and an intermediate

• Al2O3 does not form glass at normal quenching rates
• More factors are important than just bond strength
o Small cations with high charge – network formers
o Large cations with small charges – modifiers
o Medium size cations with medium charge - intermediates
By extending Goldschmidt’s original consideration of glass formation to radius
and charge of the constituent atoms / ions, Dietzel classified elements
according to their field strength, Fs
This considers the forces (attraction / repulsion) between cations and anions
in the glass
Fs = ZC / (rc + ra)2 Fs = ZC / a2
r = ionic radius of the cation (c) ZC = valence of the cation

a is the distance between cation and oxygen
or anion (a)
Dietzel, Glastechn. Ber. 22, (1948) 41

Fs = ZC / a2
ZC = valence of the cation

a is the distance between cation and oxygen
High field strength (for C it is 2.4) => covalent bonds, difficulty to form a glass
Intermediate field strength (1-2) => mixed bonds, can form glasses
Low field strength => ionic bonds, do not form glasses

Mixing two network forming elements
On cooling a binary melt with cations of approximately the same field
strength, phase separation or crystallization of the pure oxide phases is
normally seen (e.g. SiO2-P2O5, SiO2-B2O3, B2O3-P2O5).
Field strength of two cations approximately equal => demixing occurs
To form a single stable crystalline compound normally requires ∆Fs > 0.3
As ∆Fs increases, so does the number of possible stable compounds, and the
tendency to form a glass. For a binary system, glass formation is likely for ∆Fs
> 1.33
this theory can usefully categorize glass forming ability in conventional

systems, but not universally
Vogel “Glass chemistry” (1994)

SiO2, B2O3, and P2O5 the best glass-formers (Zachariasen–Warren conditions
for glass formation also met for any of their combinations)
But
SiO2–B2O3 melts solidify as glasses at any proportion (with some region of
immiscibility)
SiO2–P2O5 and B2O3–P2O5 melts solidify to crystalline solids for most
proportions, or upon rapid quenching, form phase-separated glasses
differences in field strength ∆Fs => crystalline solidification of the B2O3–P2O5

and SiO2–P2O5 systems

SiO2–B2O3
Si4+ (in SiO4) Fs = 1.57

B3+ (in B2O3) Fs = 1.63
Small difference in field strength : tendency for the division of the O2- ions
between the two competing cations => formation of immiscible glasses

Þ Thibault Charpentier
SiO2–P2O5
Si4+ (in SiO4) Fs = 1.57

P5+ (in PO4) Fs = 2.1
P higher field strength => favor the formation of PO4 tetrahedra

Si cannot compete with P to maintain SiO4 tetrahedra => SiO6 octahedra
P often promote phase separation due to its high field strength

Zachariasen’s classification of ions as network-formers, network-modifiers, or

intermediates correlates well with Dietzel’s field strength values
to glass formation:





– Phillips
Topological constraints hypothesis
Glass forming ability discussed by Phillips (1979) in term of a constraint

model.
Most inorganic covalently bonded glasses have low values of atomic

coordination number. An atom which has all covalent bonds satisfied, obeys
the (8-N) rule (Mott’s rule) i.e. Se has Nc=2, Ar has Nc=3, Si has Nc=4, etc
For a binary alloy AxB1-x, the average coordination (m):
m = x Nc(A) + (1-x) Nc(B)
Ex :
<m>As2Se3 = 2.4
<m>SiO2 = 2.67
Topological constraints hypothesis
Phillips theory: the glass-forming tendency is maximized when the

number of constraints is equal to the number of degrees of freedom, Nd.
(usually Nd =3, 3D) => overconstrained (rigid) materials (a-Si,…) ≠
underconstrained (floppy) materials
Þ Matthieu Micoulaut
Structure of chalcogenide glasses
a-S 3s23p4
a-Se 4s24p4
a-Te 5s25p4
Electronic structure of a-Se

Energy band gap 1.7eV
Valence : lone pair electrons + bonding orbitals
lone pair electrons ➪ specific « defects »
Chains or non-connected rings
Modèle structural de a-Se

Electronic configuration of a-Se and end-of-chain defect Se Se
Se Se
Rebuilding of end-of-chain defect
a-S 3s23p4
a-Se 4s24p4
crystal
a-Te 5s25p4
System Ge-S or Ge-Se
➪ GeS4 units
GeS4 tetrahedra linking S chains
S Ge
GexSe1-x
chains, All Ge are in 4-fold
➪ GeS4 are ➪
rings coordination (iso- Tetrahedra can shares
connecting chains
structural to a-Si) corners and edges≠silicates
Glass GeSe 2
System As-S or As-Se

cristal
•Modifiers ?
Þ Annie Pradel
Þ Bruno Bureau
Þ Eugene Bychkov
Notion of network former applicable
to all kind of glasses ?
Possible to define network former or modifier in

- metallic glasses ?
- organic glasses ?
Structure of metallic glasses
Consider packing of dense spheres
Most compact configuration for spheres which are all identical in a 3D-space
(cfc, hc)
Þ compacity 0.74
Proposed by J. Kepler in 1611, this results was demonstrated only in 1998 by T.

Hales
Compacity is the ratio between the volume of particules within the cell and the
volume of the cell. It is the degree of filling of the space
cfc, hc C=0.74
cc C=0.68
And for a disordered stacking ?
Colloidal
Granular medium
Sand pile or peas
Principle : hard sphere dense random packing
Packing ratio: 0.636
Þ random packing of spheres

Þ Importance of free volume
Þ metallic glasses: minimization of free volume
Þ dense random packing of spheres
Canonical cavities of Bernal

Archimedes anti- Tetragonal
prism dodecaedron
Cavités canoniques de Bernal
Tetrahedron Octahedron Icosaedron
Local arrangements for a model

John
of dense random packing
Desmond
Bernal
(1901-
1971)
Zr41.2Ti13.8Cu12.5Ni10Be22.5
Mixing various elements with different sizes
Metal (Fe,Ni,Al,Cr,Pd)
Metalloids (P,Si,B,Ge)
Principle of confusion Tang et al., Nature 402, 160 (1999)
A more complicated chemical composition translate into a greater number of

compounds that could nucleate and, thus, in mutual competition such that crystal
nucleation and growth is frustrated and does not take place on sufficiently rapid
cooling
From ’Silicate glasses and melts: properties and structure’
Mysen & Richet (2005)
+ medium range order (2006)
Stacking of blocks with 5-fold symmetry
Sheng et al., Nature 439, 419(2006)
icosaedron
Miracle, Nature Materials 2004 Þ Lindsey Greer
Yavari, Nature Materials 2005
Sheng et al, Nature 2006
Yavari, Nature 2006
Energetic Stability: Clusters

clusters lead to energetic stability
prevalence of efficiently-packed clusters (low free volume and energy) that do
not have symmetry suitable for crystal formation explain the stability of
supercooled liquids
icosahedrally-coordinated atomic clusters are the probable clusters for pure
liquids
Þ The Zachariasen’s rule do not applied

Þ Possible to talk of network former ?
Structures of polymers
chains packed in a regular way => crystal
Both amorphous and crystalline areas can

exist in the same polymer
Chain entanglement: Long polymer

chains get entangled with each other
= random coil model

Modèle de pelote aléatoire
Structures des polymères
(a) Linear structure : van der Waals
bonding between chains
Direction of increasing strength
(b) Branched structure : Chain packing

efficiency is reduced compared to linear
polymers - lower density
(c) Cross-linked structure : Chains are

connected by covalent bonds. Often
achieved by adding atoms or molecules
that form covalent links between chains
(d) Polymerized structure : Three active

covalent bonds - Highly cross-linked
3 main structural atomic models for non-crystalline
solids
•Random Network
Continuous random network ® ioni-covalent glasses SiO2
Modified random network ® ioni-covalent glasses Na2O-SiO2
• Random Close Packing

Empilement aléatoire compacte ® metallic glasses
• Random Coil Model

Modèle de pelote aléatoire ® polymeric organic glasses
References
http://www.lehigh.edu/imi/teched/OPG/lecture2.html
https://www.lehigh.edu/imi/teched/GlassCSC/Lecture_2_Martin.pdf
https://www.lehigh.edu/imi/teched/AtModel/Lecture_2_Micoulaut_Atomistics_Glass
_Course.pdf
http://web.mst.edu/~brow/pdf_structure1.pdf
www.eng.uokufa.edu.iq/staff/sawsandh/Polymers.ppt
www.phy.bme.hu/Kristalyos_es_amorf.../Introduction.ppt

Introduction To The Structure and The Structural Role of Elements in Glasses

Uploaded by

Copyright:

Available Formats

Introduction To The Structure and The Structural Role of Elements in Glasses

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Introduction To The Structure and The Structural Role of Elements in Glasses

Uploaded by

Copyright:

Available Formats

Introduction

to the structure and the

Ecole GDR Verres Cargèse

From ’Chemistry of glasses’

Glass formation results when

• Liquid is frustrated by internal structure that hinders both events

Quantitative estimate through time- Tf liquid

temperature-transformation (TTT) Supercooled

How much time does it take at any Tn

one temperature for a given fraction liquid +solid

Good glass formers:

Good glass formers:

Quantitative estimate through time-

How much time does it take at any

What is the region of glass formation ?

Determination of glass forming region after extensive studies in compositions

Musgrave et al., J. Non-Cryst. Solids, 386 (2014) 61

Richet et al., Chem. Geol. 225 (2006) 388

Neuville et al., Chem. Geol. 229 (2006) 173

Correlation between critical cooling rate

Richet et al., Chem. Geol. 225 (2006) 388

What internal structures promote glass formation?

How can structures be developed that increase the viscosity and

• Based on coordination number

• Based on bond type

• Based on bond strength

• Based on field strength

• Based on Mott’s rule

Short range order (<3 Å):

Medium (intermediate) range

Almost no long range order

Salmon et al., Nature 435 (2005) 75-78

A good structural model should

- explain structure-properties relationships

Initially, idea of micro-crystallites (quasi-crystals) but it was demonstrated

crystallite model of Lebedev

• Based on coordination number

• Based on bond type

• Based on bond strength

• Based on field strength

• Based on Mott’s rule

Victor M. Goldschmidt (1888-1947)

BeO with rBe/rO ~0.221 but does not form a glass

• Based on coordination number

• Based on bond type

• Based on bond strength

• Based on field strength

• Based on Mott’s rule

William Houlder Zachariasen (1906-1979)

Network former: Si, Ge, P, B, …

A2O, AO, AO3, A2O7, etc., no glasses could be formed

Huang et al., Nano Lett. 12 (2012)1081

Si : (Ne)3s23p2 : 4 valence electrons => ions Si4+ (can share 4 bonds)

B : (He)2s22p1 : 3 valence electron => ions B3+ (can share 3 bonds)

About ~75% of B atoms are in boroxol rings

the boroxols allow one to maintain a low-

Introduce for the first time the notion of inhomogeneities in glasses

crystallite model of Lebedev

Zachariasen's rules do not consider at all modified oxides or