Jay Lathrop
Jay W. Lathrop was an American engineer who, together with James R. Nall, invented photolithography for semiconductor devices at the US Army's Diamond Ordnance Fuze Laboratories in Washington in 1957 and 1958. Their trick was to turn a microscope around: a large, easily drawn pattern was projected through the lens in reverse so that it landed as a tiny, sharp image on a light-sensitive coating on the semiconductor. Lathrop is the person this entry is about, and the 1957-58 demonstration is the event; the process they started has its own entry, so the focus here is on who they were, what problem they were solving, and why a fuze laboratory turned out to be the birthplace of chip patterning.
The motive was miniaturisation for military electronics, not computers. The Diamond Ordnance Fuze Laboratories, a US Army research centre (later renamed the Harry Diamond Laboratories), worked on proximity fuzes and other small, rugged electronics that had to survive being fired from guns and carried in missiles. Transistors promised to replace vacuum tubes in such devices, but the wiring and packaging of discrete parts limited how small the circuits could become. Around the time of the Sputnik launch in October 1957, the Army and its contractors were pursuing several microminiaturisation programmes, and Lathrop and Nall were seeking a way to make the transistor structures themselves far smaller than could be laid down by hand-applied masks.
The inverted-microscope idea reversed an instrument built to enlarge so that it would shrink. A microscope takes a tiny object and presents a magnified image; run the light the other way, with a bright, enlarged pattern placed at the specimen end, and the same optics produce a greatly reduced image at the other end. Lathrop and Nall used this arrangement to project a pattern onto a semiconductor slice coated with a photosensitive film of the kind used in the printing industry. Exposed areas of the film hardened, unexposed areas could be washed away, and the patterned film then protected the surface during etching. The result was a repeatable way to produce transistor features on a scale of a fraction of a millimetre, far finer than hand methods, and one that could be copied to many devices at once.
A patent and a change of employer carried the method from the Army into industry. Lathrop and Nall filed for a patent on their process of fabricating semiconductor devices, and the patent was granted some years later in the 1960s. Lathrop left the Army laboratory in 1958 and joined Texas Instruments, which was then working on Jack Kilby's integrated circuit; the company's engineers were interested in any method that could pattern many small features precisely. That timing helps explain how photolithography reached the integrated-circuit story so quickly, because Kilby's germanium demonstration came in 1958, Fairchild's planar work followed in 1959, and both were hungry for fine, repeatable patterning. He later moved into academia, at Clemson University.
The credit is real but the history is messier than a single eureka. Photographic techniques for printed circuit boards, and light-sensitive coatings like those used in engraving, existed well before 1957, so Lathrop and Nall did not invent photography or photoresists; what they showed was that these could be shrunk by optics to the dimensions of transistors and applied to semiconductors. Equally, they did not combine their work with oxide masking, which was Fairchild's step, and a common misconception is that one person or company invented the whole planar-photolithographic combination. The episode is an instructive reminder of how much of the semiconductor art came from defence-funded work in government laboratories.
The modern relevance is the scale of what grew from the inverted lens. Today's lithography uses projection optics, step-and-repeat or scanning exposure tools, and extreme ultraviolet light, to print patterns with features of a few nanometres on wafers hundreds of times larger than the original slices, but the conceptual chain is unchanged: draw big, project small, record in a light-sensitive layer, etch. Every step in transistor scaling of the Moore's law era has depended on lenses that are descendants, in principle, of the microscope used backwards in 1957.
Read Jay Lathrop as the man who turned a microscope around: with James Nall at an Army fuze laboratory in 1957-58 he showed that a large drawing could be optically shrunk into a transistor, and that simple, almost obvious reversal gave chip fabrication its method of patterning.