Skip to content

Latest commit

 

History

History
364 lines (242 loc) · 11.2 KB

lecture3.md

File metadata and controls

364 lines (242 loc) · 11.2 KB

class: title, no-number time: 0

Lecture 3 - Ionic Conductivity

.footer[- Return to course contents ]


class: roomy

Lecture summary

$\require{mediawiki-texvc}$

  • Recap of defect types
  • Ionic conductivity
  • Conduction mechanisms
  • Ionic migration paths
  • Energetics of conduction

Defect recap

:vote


Defect recap results

:results


class: compact

NaCl Schottky

  • Intrinsic defect (composition doesn't change) so formula mass change is 0
  • Vacancies essentially create a hole in the structure, so density decreases

AgCl Frenkel

  • Intrinsic defect
  • Vacancies and interstitials created at the same rate, so density doesn't change significantly

TiO2 substitution with Zr

  • Zr is heavier than Ti (91.224 vs 47.867) so formula mass will increase
  • Zr4+ is slightly larger than Ti4+ so we might expect volume to increase slightly, but overall the density will increase
    • NOTE: I wrongly said Zr4+ was smaller than Ti4+ in the lecture - sorry!

Shear phase in WO3

  • $\ce{WO_{3} -> WO_{3-x}}$, so the formula mass will decrease
  • Shear phase formation is a substantial rearrangement which is driven by reducing the overall volume. The result is that decrease in volume far outweighs the decrease in formula mass, so density will increase.

class: compact

** Replacing O by F in CeO2**

  • $\ce{CeO_2 -> CeO_{2-x}F_{x}}$
  • F is slightly heavier than O (18.998 vs 15.999) so formula mass will increase slightly
  • Charge balance requires us to reduce some Ce4+ to Ce3+. There are a few ways to imagine this with KV notation:
    1. Replacing $O^{2-}$ by $F^{1-}$ occurs directly during synthesis:
      • $\ce{O_{O} + Ce_{Ce} + H_{2} + \frac{1}{2}F_{2} -> F_{O}^{\bullet} + Ce_{Ce}^{'} + H2O}$
      • (H2 and F2 are used to balance the equation here, but a real experiment would use a different fluorinating reagent such as CeF3)
    2. As-synthesised CeO2 is fluorinated by filling existing oxygen vacancies:
      • $\ce{V_{O}^{\bullet\bullet} + Ce_{Ce}^{'} + \frac{1}{2}F_{2} -> F_{O}^{\bullet} + Ce_{Ce}^{x}}$
      • Note that the fluorite structure is often prone to anion vacancies due to the mismatch between ionic radii and the geometry (this is particularly prevalent when the cation can be reduced). We'll see why this again in lecture 6.

class: compact

Conductivity

  • Many ionic solids conduct electricity; due to ionic and/or electronic motion.

  • Most ionic solids are electrically insulating/semiconducting (localised electrons) --

  • Ionic conductors are important!

.pull-center[

![Electric vehicle](./images/electric_vehicle.jpg# db maxh-4) ![Gas sensor](./images/gas_sensor_H2.jpg# maxh-4 )
Batteries (Lecture 4) Sensors
![Li separation from seawater membrane](./images/separation_membrane.jpg# maxh-4 ) ![Fuel cell stack](./images/bloom_energy_fuel_cell.jpg# maxh-4 )
Separation Membranes Fuel Cells (Lecture 6)
]

class: roomy

Origin of ionic conduction

  • Ionic conductivity is dominated by defects

    • In an ideal crystal, ions can't easily move
    • vacancies and/or interstitials are the main charge carriers --
  • Conductivity, $\sigma = nq\mu$, where

    • $n$ is number of charge carriers
    • $q$ is charge
    • $\mu$ is the mobility of charge carriers --
  • In ionic solids, conductivity covers $\ce{10^{-16}\ S\ m^{-1}\ \bond{-} 10^3\ S\ m^{-1}}$

    • most solids are limited to around $\ce{10^{-2}\ S\ m^{-1}}$
    • Liquid electrolytes typically $\ce{10^{-1}\bond{-} 10^3\ S\ m^{-1}}$

???

The Siemen $(S)$ is the unit of conductance and is the inverse of resistance, measured in Ohms $(\Omega)$.


class: roomy

Measuring Conductivity

  • For electronic conductors, this is simple:

    • Apply a voltage $(V)$ and measure the resulting current $(I)$
    • Resistance (in $\Omega$) is found through Ohm's law; $ V = IR $
    • Resistivity (in $\Omega\ cm$) of the material calculated from geometry
  • Resistivity $\rho$ (in $\Omega\ cm$) = $\frac{1}{\text{Conductivity}\ \sigma\ \mathrm{(in\ S\ cm^{-1})}}$

.pull-center[ ]

.footer[

  • In reality, we normally use two wires for V and two for I ]

???

Resistance (in $\Omega$) depends on how much material there is; twice the length of material will give twice the resistance. Resistivity is a material property and therefore allows us to compare materials.


class: compact

Measuring Ionic Conductivity

  • Ions cannot flow round a circuit, so current drops with a constant applied voltage

.center[ ]

--

  • Instead, we use an alternating voltage - this is called Impedance spectroscopy (see lecture 5)

.center[ ]

???

For comparison, during cyclic voltammetry the voltage is continuously varied but over a much slower timescale than impedance spectroscopy.

At each measured point in a CV the current due to ionic motion is 0 A, but current still arises due to the chemistry occurring at the electrodes.


class: compact

Ion migration mechanisms

Three 'main' mechanisms of ionic migration

1. Vacancy mechanism

Vacancies move throughout the lattice (atoms move into vacancy)

.center[ ]

???

These are the most common suggested mechanisms, but others have been proposed in various materials (in particular cooperative mechanisms involving lattice vibrations [phonons] and multiple ion types).


2. Interstitial mechanism

Ions hop between interstitial sites

.center[ ]


3. Interstitialcy (knock-on) mechanism

Interstitial ions 'push' into a neighbouring site

.center[ ]

Vacancy, Interstitial or Interstitialcy?

:vote


Suggestions

:results


name: migration_paths

Migration paths

Ion paths are rarely .red[direct], but will take the .gold[lowest energy route].

.pull-left[ :jmol 350, 350, 1, 1, 1, rotate x 100; rotate y 10; select (Na1)<20>; (sodium).radius=0.5; (chlorine).radius=1.0; color atoms TRANSLUCENT 0.8 gray; draw ID "observed" ARROW (2.81 2.81 5.62) (4.2 4.2 4.2) (5.3 5.3 5.2) WIDTH 0.3 COLOR orange; draw ID "direct" ARROW (2.81 2.81 5.62) (5.2 5.2 5.62) WIDTH 0.3 COLOR red ]

.pull-right[ ![Close-packed migration path](./images/migration_path_cp.svg# w-80pct) ]


class: compact

Pathways can be complex

  • Migration pathways can be calculated and/or experimentally determined

e.g. NASICON $\ce{Na+}$ conductor, $\ce{Na3Zr2(SiO4)2(PO4)}$:

![Nasicon migration paths](./images/ionic_conduction_pathway_nasicon.gif# w-75pct absolute l-10pct)

.footer[

???

Three approaches to determining diffusion are shown here:

Molecular Dynamics (MD)

  • Computational model of ionic motion over a time period (using either electronic structure calculations or atomic potentials methods)
  • Resulting trajectory is integrated to determine location of conducting ions

** Bond Valence Energy Landscape (BVEL) **

  • Maps the bond valence sum (a measure of local electron density estimated from atomic positions) throughout the structure
  • Surface is the region where the probe ion (e.g. Na+) would experience optimum coordination

** Maximum Entropy method (MEM) / Rietveld refinement**

  • Fits a model to experimental data to determine ionic positions
  • If Na+ is diffusing throughout the measurement, the positional information is present within the diffraction data. MEM is one method to fit this.

class:compact

Migration energetics

  • Defect mobility $(\mu)$ is a thermally-activated process: $$ \mu = \mu_0 \exp \left( -\frac{\mathrm{E_a}}{\mathrm{RT}} \right) $$
  • interstitial sites are higher energy than vacancies, so will be more mobile.

![Migration energy](./images/migration_energy.svg# w-50pct relative l-3-12th)

???

Here, $\mu_0$ is the mobility at 0 K, $E_a$ is the activation energy for migration, $R$ is the gas constant and $T$ is the temperature.

Inserting an atom into an interstitial site creates a high density region which is energetically unfavourable. Creating a vacancy is also unfavourable, but atoms can normally "relax" around the vacancy, reducing the enthalpic cost. This (usually) means creating an interstitial is more energetically costly than creating a vacancy. (Assuming the migration pathway goes through the same high energy state).


class: no-number

Variation with temperature

As $\sigma = nq\mu$ and $\mu$ is thermally-activated, $$ \begin{align} \sigma &= nq\mu_0 \exp \left(-\mathrm{\frac{E_a}{RT}} \right) \\ &= A \exp \left(-\mathrm{\frac{E_a}{RT}} \right) \\ \end{align} $$

???

Here, $n$ is the number of charge carriers, and $q$ is the charge on the charge carrier. It is difficult to measure the carrier mobility $(\mu)$ directly, so these terms are combined into a single pre-exponential factor $A$.

--

![Nasicon conductivity vs temperature](./images/nasicon_arrhenius.jpg# w-third fr relative b-3) Plotting $\ln \sigma$ vs. $\frac{1}{\mathrm{T}}$ should give a straight line

  • more commonly we plot $\log_{10} \sigma$ vs. $\frac{1000}{\mathrm{T}}$ for high temperature measurements
  • gradient is $\frac{-E_a}{R}$ (or $\frac{-E_a}{2303R}$ using base 10).

.footer[

  • Nasicon conductivity ]

???

In reality, the number of defects increases with temperature, so both $\mu$ and $n$ have a T-dependence. Because defects have an associated formation energy $(\Delta \mathrm{H_{f}})$, it turns out that their formation is also thermally-activated (it is possible to derive this from the defect formation equations, see e.g. West). In practice, it is common to plot $\log(\sigma T)$ vs $\frac{1000}{T} to account for the T-dependence of A.


class: roomy

Lecture recap

  • Defects can give rise to ionic conduction
    • Occurs by three main mechanisms:
      • Vacancy hopping
      • Interstitial hopping
      • interstitialcy (knock-on) cooperation
  • Ionic conductivity is thermally-activated
    • shows Arrhenius-like behaviour
  • Different defects have different conduction energetics
    • Pathways can sometimes be determined experimentally

Feedback

:vote

.footer[- Return to course contents ]