On this page
Diodes, solar cells and many transistors depend on a boundary between two kinds of doped silicon. That boundary, the p-n junction, was first noticed by accident in 1940 in a silicon sample with a crack across it. This post covers what Russell Ohl saw, why the two sides behaved differently, a worked example from his own patent, and what a student can take from the way he followed up a strange result.
Who was Russell Ohl?
The Computer History Museum's Silicon Engine people pages give his dates as 30 January 1898 to March 1987. He earned his bachelor's degree at Pennsylvania State University and joined Bell Labs in 1927, working at its laboratory in Holmdel, New Jersey. During the late 1930s, while developing detectors for short-wave and microwave communications, he turned to crystal rectifiers made of silicon.
Published detail on his life outside the lab is thin, so this post sticks to the discovery, which is well recorded.
What happened on 23 February 1940?
The museum's account, 1940: Discovery of the p-n Junction, tells the story. Ohl had found that purer silicon made better radar detectors. On 23 February 1940 he tested a small silicon slab with strange results. When it was exposed to bright light, the current through it jumped. Different parts of the crystal also gave opposite electrical effects when touched with a fine wire probe, a "cat's whisker".
Ohl and his colleague Jack Scaff found that a seam in the slab separated two regions with different impurities. One impurity, phosphorus, gave a slight excess of electrons. The other, boron, gave a slight deficiency, later recognised as "holes". They called the regions n-type and p-type, and the boundary became known as the p-n junction. Light striking the junction drove electrons across it, producing a current.
His patent, Light-Sensitive Electric Device, was filed on 27 May 1941 and issued on 25 June 1946. It describes silicon that was "sensitive to visible light, generating an electromotive force independently of any applied voltage". It also explains the naming: the region above the barrier developed a positive thermoelectric potential against a copper electrode and so was called the P zone.
Why do phosphorus and boron behave differently?
Silicon has four valence electrons. The NCERT Class 12 chapter on semiconductor electronics explains that pentavalent dopants, such as phosphorus and arsenic, have five, and trivalent dopants, such as boron and aluminium, have three. A phosphorus atom leaves one electron spare, making n-type silicon. A boron atom leaves one bond short, a hole, making p-type silicon. Where the two meet, electrons and holes cross and cancel, leaving a thin region with an electric field across it. Light that frees an electron near this region gets it pushed one way and the hole the other, and that is a current.
Worked example: what did Ohl's cells deliver?
The patent reports that a bank of eight cells "will develop a potential difference of approximately 2 volts at 2 lumens and a maximum difference of potential of about 2.4 volts."
The cells are in series, so their voltages add: V_total = 8 × V_cell.
At the maximum, V_cell = 2.4 V ÷ 8 = 0.3 V.
At 2 lumens, V_cell = 2 V ÷ 8 = 0.25 V.
To reach a sample target of 6 V with the maximum figure, you need 6 ÷ 0.3 = 20 cells in series.
The silicon cells Bell Labs built in 1954 gave about 0.52 V each, as our post on Chapin, Fuller, Pearson and the silicon solar cell shows. The American Physical Society's history of the solar cell puts Ohl's cell at about one percent efficiency.
Common mistakes:
Adding voltages for parallel cells. Series cells add voltage; parallel cells add current at the same voltage.
Calling holes particles of the nucleus. A hole is a missing electron in a bond, which behaves like a positive charge.
Mixing p-type and n-type dopants. Trivalent boron gives p-type; pentavalent phosphorus gives n-type.
Our guide to p-n junction diode characteristics takes the junction further, and Fermi's Simulation Lab has interactive simulations for ideas, like charge crossing a junction, that are easier to see than read.
What came from the discovery?
The museum's account says William Shockley's 1948 idea for the junction transistor derived from Ohl's 1940 discovery, and that the p-n junction became the most common rectifier in electronics and a basic building block of semiconductor devices. Our post on John Bardeen and the transistor covers the next step.
What can a student take from his way of working?
Treat a strange reading as data. A jumping current could have been dismissed as a faulty sample.
Find a partner to test it. Working with Scaff, Ohl traced the effect to the seam and its impurities.
Write it down precisely. His patent records methods, purities and voltages that we can still check today.
Try it now: work out how many of Ohl's 0.3 V cells and how many 0.52 V cells you would need for 12 V.
