November 1959

MOSFET

Atalla and Kahng fabricate the first working MOS transistor

The patent drawing for what became the basic switch of modern computing was filed under the restrained title 'Electric Field Controlled Semiconductor Device.'

Late 1959: a thin layer of oxide

In late 1959 at Bell Telephone Laboratories in Murray Hill, New Jersey, Mohamed M. Atalla and Dawon Kahng fabricated the first successful insulated-gate field-effect transistor — a metal–oxide–silicon structure. The month conventionally attached to the milestone, November, is an approximate secondary-source date; the year 1959 is what the institutional record supports.

The decisive part was almost invisible: a thermally grown layer of ordinary silicon dioxide between the metal gate and the silicon. Atalla's work had shown the oxide could tame the surface states that had frustrated earlier attempts to realize the long-proposed field-effect principle — and with that stable interface, an applied gate voltage could control current in the silicon beneath.

Invented in 1959, shown in 1960

The device was not publicly unveiled that November. Atalla and Kahng presented it at the IRE Solid-State Device Research Conference in Pittsburgh in June 1960, and Kahng filed the patent application on 31 May 1960 — issued in 1963 under the title 'Electric Field Controlled Semiconductor Device.'

Nor did it transform electronics overnight. Bell Labs initially undervalued the device — slower and less mature than bipolar technology — and stability problems delayed widespread MOS adoption until later in the 1960s and 1970s.

Credit where the record puts it

Two adjacent 1950s breakthroughs often get folded into this story. Frosch and Derick's oxide masking at Bell Labs was process groundwork — their transistor was a double-diffused junction device, not a MOSFET. Jean Hoerni's planar process was a separate 1959 innovation at Fairchild. The institutional and patent records credit the working MOS device to Atalla and Kahng.

The scale of what followed has a number: a 2018 Computer History Museum estimate put cumulative transistor production at 13 sextillion, with MOS transistors in more than 99% of modern microchips.

Sources

Researched 23 Aug 2026 6 sources 3 audit passes Date disputed

How this was checked

Researched from the web into a fact sheet, rewritten from that sheet, then audited against it by a different model. How the pipeline works →

Date disputed. Sources disagree; the research places it at ~1959-11.

What the sources leave uncertain

The year 1959 is well supported by the Computer History Museum and historical literature, but the exact month “November” is not established by the accessible primary patent or conference records consulted. Those records document the May 31, 1960 patent filing and June 1960 presentation. November should therefore be treated as an approximate secondary-source date.
Sources differ in whether they label 1959 the invention or fabrication and 1960 the demonstration. The cleanest distinction is: working device achieved in 1959; public presentation and patent filing followed in 1960.