ASTR 3130 (Majewski) Lecture Notes

DETECTORS: DOPING OF A SEMICONDUCTOR

• One can dramatically alter the conductivity of a semiconductor by "preloading" it with an excess of conduction electrons or holes.
• Add to molten valence 4 elements impurities of valence 3 or valence 5 elements. When crystallizes, the impurities are incorporated into lattice.
• Can also add valence 3 or 5 impurities by high speed injection.

• The net effect is to generate some allowed energy levels in the normally forbidden bandgap zone into which electrons can jump from the valence band (P-type doping) or in which electrons already exist to jump into the conduction band (N-type doping). In either case, the normal bandgap energy is "short circuited".

• N-Type Doping: Column Va element has 5 valence electrons --> surplus 1 electron in the lattice that is relatively easily detached to conduction band. Can tune lattice with a density of impurity that exceed the number of thermal and photo-excited (E_gap = hf ) electrons if desired.
.
"Donor", or N-type impurity added
• From an energy level standpoint, the effect of N-type doping is to create an energy level close to the conduction band with an electron only needing to overcome a small effective energy bandgap to E_i.
• At normal temperatures, this allows there to be an excess of electrons available for electrical conduction.


• P-Type Doping: Column IIIa element has 3 valence electrons --> shortage of electrons, surplus of holes. Atoms want to capture electrons from valence 4 atoms - holes migrate and conductivity again increases.

"Acceptor", or P-type impurity added
• From an energy level standpoint, the effect of P-type doping is to create an energy level close to the valence band into which it is easy for an electron to jump from the valence band.

• E_i is the (small) energy needed to excite an electron from top of valence band into acceptor level and thereby "bind the hole" into the valence band. At normal temperatures, the existence of extra holes in the valence band allows electrical conduction of holes.


• In summary:
  The point of doping is:
  • For photon sensitivity: Change the wavelength sensitivity of a semiconductor by effectively changing the energies that can be absorbed with the newly created bandgap energy levels.
    • Thus, we can dope standard semiconductors to increase their sensitivity to longer (redder) wavelengths.

  • For electrical conductivity: "Pre-load" a material with an excess of electrons or holes.
    • Much work with semiconductors in circuitry involves cleverly combining insulators and p-doped and n-doped substances to make a variety of junctions.

METAL OXIDE SEMICONDUCTOR CAPACITOR
(Operation of a single pixel)

• Made by covering a block of semiconductor with a thin layer of insulating material like SiO₂.
• Add a small electrode (gate) on top
• If gate is positively charged, free electrons in semiconductor move towards gate (and holes away) but they can't cross insulator. This is a capacitor.


(Note incorrect width of insulator layer shown in figure.)

• P-Type Substrate: Now if we make the substrate out of a p-doped semiconductor and put the gate at +10 V:
  • The holes move away from the gate as before, but there are virtually no free electrons to move closer to SiO₂.

  • The region near the SiO₂ layer is called a depletion region or depletion zone.
  • Thus, if there are no thermally created electron-hole pairs, i.e. the device is cold, only photoelectrons will collect near the Si-SiO₂ interface (represented by the one "-" shown below).
  
  • The depletion region can be thought of as a well, where photoelectrons are stored. The size of the well is proportional to the gate voltage.
  • The voltage applied to a gate is called the bias voltage (note connection to "bias frame" discussed later).
  • The maximum charge (number of electrons) a pixel can hold is called the well capacity.
  • Well capacity generally given in units of e^-.
    E.g. CCDs typically can hold > 150,000 e^-/pixel.
  • The well depth (and the system electronics) affect the possible dynamic range of the device.
    E.g. with the CCD pixel above with a 150,000 electron depth, at most only this many different levels could be recordable.

  • Obviously, the key to integration is being able to create and store photoelectrons in the pixel wells.

  • How do we monitor/measure the total charge collected in the pixels?
  • One way is to switch the gate voltage and drive the electrons into the substrate where it can be collected and read as a current via another gate:
    
    This readout method is called a Charge Injection Device.

ALTERNATIVE PIXEL READOUT METHOD - CHARGE COUPLING

• An alternative readout method is the heart of a Charge-Coupled Device (CCD).
• Imagine two gates on same substrate --> two depletion zones.
• If the gates are far enough, the wells are independent.

• But, if gates are placed only about 1 micron apart, the depletion zones can "communicate".
• By fiddling with voltages, one can change the size of the local wells, and manipulate these to transfer charge packets from one to the other; electrons seek the deeper well (i.e., they are attracted to the larger gate voltage).

• The gates in this configuration form the basis of the Charge Transfer Mechanism.
• Now imagine rows of gates along a substrate and the ability to move charge along the substrate from gate to gate like a "bucket brigade" --> this is the essence of the Charge-Coupled Device (CCD).