Doping Silicon Wafers

Posted on 18 December 2012

Doping silicon wafers


Turning silicon into semiconductor chips used to create integrated circuits is a complex procedure that must be carried out meticulously. A single silicon wafer may undergo many successive processes to achieve the desired layers of conductor, semiconductor, and insulating material needed to produce the required circuitry. The wafer undergoes various steps to become properly doped for its intended use.


Doping silicon wafers, however, is not a one way fits all procedure. Wafers are therefore classified to ensure each particular wafer receives the exact treatment that is ideal for its purpose to eliminate unsatisfactory wafer materials from the process stream. Wafers are sorted into batches of uniform thickness for doping.

Characterization of Silicon Wafers

Some of the factors considered when characterising silicon wafers include:

  • Production method (CZ, FZ), doping type (n-type or p-type)
  • Crystal orientation (<1 0 0>, <1 1 0>, <1 1 1>, off-orientation)
  • Type and density of crystal defects
  • Type of Dopant used (phosphorus, arsenic, antimony, boron)
  • Electrical resistivity ρ (in Ω cm)
  • Diameter and thickness of the wafer
  • Flatness tolerances (thickness variation)
  • Information on flats or notches, edge rounding
  • Sawed, etched, lapped or polished surface
  • Surface quality, hydrophobic (water repellent) or hydrophilic (wettable by water) surface
  • Packaging


Alloying is the combination of two or more chemical elements one of which is always a metal.

Diffusion refers to the movement of molecules from an area of high concentration to an area of low concentration. Diffusion can be explained using Fick's laws of diffusion.

Fick's first law of diffusion

Fick's first law of diffusion states that the diffusive flux density is proportional to the existing concentration gradient

J = -D \frac{dN}{dx}

where J is the diffusive flux density (1/cm²s), N is the concentration of diffused particles (1/cm³), and D is the diffusion constant (cm²/s)

The negative sign indicates that the the flux flows from areas of high concentration to areas of low concentration.

Fick's second law of diffusion

Fick's second law of diffusion states that the rate of change in concentration is proportional to the local rate of change of the particle flux density.

\frac{dN}{dt} = \frac{\delta J}{\delta x}

Particle diffusion equation

The particle diffusion equation is acquired by combining both of Fick's laws giving

\frac{\delta N}{\delta t} = D \frac{\delta^2 N}{\delta x^2}

Note: This equation is commonly regarded as Fick's second law in place of the form given above.

Solving the equation for a non-depleted source gives

{N(x,t) = N_0 \cdot erfc (\frac{x}{2\sqrt{Dt}}})

where N(x,t) is the concentration at point x at time t, N0 is the surface concentration, and erfc(z) is the complementary error function.

The diffusion constant D is determined by the exponential of -1/T :

D = D_0 e^{- \Delta E/kT}

The higher the temperature the higher the rate of diffusion.

The diffusion profile takes the form of a Gauss bell curve. The shape of the concentration distribution is affected by crystal damage on the surface or other impurities in high concentration.

Diffusion mechanisms

The diffusion speed of a given doping agent and the its ability to dissolve in silicon are of importance in the doping process. The diffusion of a molecule does not occur in a straightforward movement but in zigzag motion towards the point of least concentration.

There are different mechanisms in which an atom moves within the crystal:

a) Swapping places with a neighbouring atom on the grid

b) Swapping rings with neighbouring atoms on the grid

c) Interstitial diffusion (fast diffusing atoms such as Cu, Fe)

d) Empty space diffusion (for slow diffusing atoms such as P, B)

Concentration distribution

Figure 1. Concentration Distribution

Carrier gas diffusion doping

The doping material used can either be solid, liquis or gaseous. In carrier gas diffusion doping, a carrier gas carries the doping atoms to the silicon wafers which been brought by radiant heat to temperatures of 1000°C to 1285°C.

Carrier gas diffusion

Figure 2. Carrier gas diffusion doping

Liquid diffusion doping

In liquid diffusion doping, the dopant is dissolved, for example, in a glass forming agent such as TEOS (Tetraethyl Orthosilicate) and applied to the silicon wafer by either spraying or by letting the solvent simply run over it. Upon heating, this liquid changes into glass upon which the doping material for (Phosphorous, Boron, or Aluminium) diffuses into the silicon. After the diffusion process is complete, the glass is then removed. A simultaneous diffusion of phosphorous on one side and Boron on the other side can also be done. After the diffusion process, the glass layers are etched away.

The main advantage of employing liquid diffusion doping is the fact that many wafers can be doped simultaneously using phosphorus and boron. A disadvantage, however, is that removal of the boron glass is quite complicated. There is an inherent risk of redoping due to the overflow of the dopant on the edges of the wafer.

Diffusion using solid dopants

One method doping using solid dopants is ampoule tube diffusion. After loading of a quartz tube (vial) with silicon wafers and the solid dopant (AlSi, Ga2O3), the vial is purged with argon or evacuated. Then a quartz cap is tightly fastened at the ends of the vial. The vial is then passed through the diffusion furnace.

Ampoule tube diffusion

Figure 3. Ampoule tube diffusion

The doping wafers used in solid dopant diffusion are made of a boron-nitride ceramic and part silicon oxide. The pores of the wafers are filled with boron oxide. When the wafers are brought to a high enough temperature, the doping process procedes.

Diffferent levels of concentration can be reached depending on the chosen process parameters and the composition of the boron nitride slices. A regular doping spread on the silicon wafer can be achieved. Doping papers soaked with  boron or phosphorus solvents can be used to replace doping slices.

An advantage of ampoule tube doping is that it allows a large number of wafers to be doped at a time and also provides high doping uniformity. A major disadvantage is that the quartz ampoules are expensive and can only be used a few times before they must be disgarded.

Ion Implantation into Silicon

For the process of implantation, the doping material is vaporized in order to remove electrons from the atoms. The ionised atoms are accelerated using voltages ranging from  50 000V to 6 million volts and then shot into the silicon. This process is carried out within a hard vacuum.

Any type of atom (boron, phosphor, helium, oxygen) can be implanted into silicon. The ions' energy is reduced by elastic collisions with the atoms of the silicon until the ions are stuck within the crystal  grid. The mass, charge and acceleration voltage determine just how deep the ions penetrate into the silicon. The crystal orientation of the silicon in question is also an influential factor since the ions experience less resistance at certain points on the crystal and can therefore be implanted even further into the silicon (channeling).

depth of penetration

Figure 4. Silicon crystal lattice

The energized ions break into the silicon crystal (damage) and remain stuck in at a random spot in the crystal grid. The ions must, however, land in the right position in the crystal grid in order to work as doping atoms. Applied temperatures ranging from 800°C to 1000°C (annealing) ensures that the atoms go into the right positions within the crystal grid and eliminate crystal damages. Annealing is done in an oven or through Rapid Thermal Annealing (RTA) seconds using halogen lamps.

Rapid Thermal Annealing

Figure 5. Rapid Thermal Annealing

Through implantation, doping can be done within the silicon without changing the doping concentration on the surface (buried layer). Oxide layers can be used within the silicon to provide insulation (e.g. SOI drivers).


Epitaxy refers to growing of a crystal layer on a substrate in the same crystal lattice and with the same crystal orientation as the substrate. In order to do this, the surface of the substrate must be thoroughly polished and be extremely clean. The epitaxy procedure takes place in a quartz-Epi-Reactor. There are various methods of carrying out the epitaxy procedure such as gas phase epitaxy and liquid phase epitaxy. Doping through epitaxy generally involves the addition of Phosphine (PH3) or Diborean (B2H2) during growth. High quality epitaxy layers with few crystalline defects can only be acquired through slow growth of the layer.


For more information, please read:

Silicon Production

Neutron Transmutation Doping of Silicon Rods

What is a Semiconductor?

Semiconductor Doping


This post was written by:

- who has written 30 posts on PowerGuru – Power Electronics Information Portal.

Contact the author

Leave a Response

You must be logged in to post a comment.