2
\$\begingroup\$

I was just doing a project for a website and my team member excavated a standard pn-junction diode to see that inside the ceramic casing, the semiconductors of p and n doping merely abut together in close alignment. Shouldn't the case be that a single length is variably doped with n and p type impurities to form the junction? If I remember correctly, then textbooks tell that simply joining the two by their boundary does not make it a p-n junction since there are always irregularities on microscopic scale. Any insights on this?

\$\endgroup\$
  • 7
    \$\begingroup\$ Doping is usually accomplished via diffusion, not by assembling separately processed pieces into contact. You may be misidentifying the structures, for example mistaking contact metalization for an added on oppositely doped piece of semiconductor. \$\endgroup\$ – Chris Stratton Aug 28 '12 at 6:23
  • 4
    \$\begingroup\$ If you provide a suitably magnified photo, I'm sure someone could identify parts for you. \$\endgroup\$ – pjc50 Aug 28 '12 at 9:46
  • \$\begingroup\$ If it was a Schottky diode, you might be right. Those are junctions between a doped semiconductor and a metal. Look for a ring of tiny bolts around the junction. \$\endgroup\$ – Olin Lathrop Aug 28 '12 at 12:27
  • \$\begingroup\$ That last one was about the tiny bolts was meant to be a joke, but from past experience I suppose I should make that clear. \$\endgroup\$ – Olin Lathrop Aug 28 '12 at 12:28
  • \$\begingroup\$ I am afraid I cannot upload the picture as of yet. It belongs to a website work, but as soon as it is up, I will post a link here. It is basically an editing job for me there. \$\endgroup\$ – Anshul Aug 28 '12 at 15:22
1
\$\begingroup\$

Separately processed pieces of semiconductor mechanically joined won't work well as a pn junction because the facing surfaces need to be perfectly joined (as the crystalline structure level), and without and intermediate impurities (oxidation of the surface). In theory it would work, if you could align the crystals perfectly, but in practice it can't be done.

Generally a pn junction is formed by taking a bulk doped semiconductor (e.g. n-type), and diffusing p-type impurity at the surface. This creates a pn junction beneath the surface of the semiconductor. The p portion is actually counterdoped - it contains n impurities, but more p impurities, giving a new p-type characteristic.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.