On this page
Almost every phone call and video stream you use travels part of the way as pulses of light inside glass threads thinner than a hair. In the early 1960s that seemed impossible over any useful distance. Charles Kao worked out why the light was being lost and what kind of glass would stop the loss. This post tells his story and works through the two calculations behind it: total internal reflection and decibels of loss.
Who was Charles Kao?
Kao was born in Shanghai on 4 November 1933 and died in Hong Kong on 23 September 2018, according to his Nobel Prize facts page. The Nobel Committee's popular information on the 2009 prize records that his family moved to Hong Kong in 1948, that he trained as an electronics engineer, and that he completed his PhD in London in 1965. By then he worked at Standard Telecommunication Laboratories (STL) in Harlow, England, studying glass fibres with his young colleague George Hockham.
What problem did Kao solve?
The light-guiding part was already understood. A fibre traps light by total internal reflection, because its glass core has a higher refractive index than its surroundings, and coating the fibre in a lower-index glass cladding had already made medical instruments possible. The problem was loss. According to the popular information, after just 20 metres only 1% of the light that entered a fibre remained.
Kao and Hockham set a goal: at least 1% of the light should remain after 1 kilometre. In January 1966 Kao presented his conclusion. The main problem was not imperfections in the fibre thread but the glass itself, which had to be purified to a transparency never achieved before. He pointed out that fused silica could be used. The work was published as "Dielectric-fibre surface waveguides for optical frequencies" in the *Proceedings of the IEE* in July 1966.
Within a few years, scientists at Corning Glass Works in the USA had made a kilometre-long fibre using chemical processes. The popular information adds that today about 95% of the light remains after a full kilometre, and that infrared light with a wavelength of 1.55 micrometres, where losses are lowest, is used for long-distance links.
Worked example 1: the critical angle in a fibre
The NCERT Class 12 Physics chapter on ray optics explains that each fibre has a core and a cladding, with the core's refractive index higher, so light striking the boundary at more than the critical angle is totally reflected again and again.
Take sample values: core n₁ = 1.48, cladding n₂ = 1.46.
Critical angle. sin C = n₂/n₁ = 1.46/1.48 = 0.9865.
Angle. C = sin⁻¹(0.9865) ≈ 80.6°.
Meaning. Light must hit the core-cladding boundary at more than 80.6° from the normal, so it travels nearly parallel to the fibre axis. Rays that are too steep leak into the cladding.
Our guide to total internal reflection and critical angle problems has more practice.
Worked example 2: what decibels of loss mean
Engineers measure loss in decibels: loss in dB = 10 log₁₀(P_in / P_out).
Kao's target. 1% left after 1 km means P_in/P_out = 100, so the loss is 10 log₁₀ 100 = 20 dB per km.
The 1960s fibres. 1% left after 20 m is also 20 dB, but over 20 m. That is 20 × (1000/20) = 1,000 dB per km.
Modern fibre. 95% left after 1 km gives 10 log₁₀(1/0.95) ≈ 0.22 dB per km.
A 100 km link. At about 0.2 dB per km, 100 km costs about 20 dB, so roughly 1% of the light arrives, the same fraction Kao hoped for over a single kilometre.
That is consistent with the Nobel press release, which says that with a fibre of the purest glass, light signals could travel over 100 kilometres, compared with only 20 metres for the fibres of the 1960s.
Common mistakes with optical fibre questions
Swapping the indices. sin C = n₂/n₁ needs the smaller index on top. If you get sin C greater than 1, you have them reversed.
Measuring from the surface. The critical angle is measured from the normal, not from the boundary.
Treating decibels as linear. 20 dB is not twice as bad as 10 dB in power terms; it is ten times worse.
Thinking fibres work by mirrors. There is no metal coating. Reflection happens because of the index difference.
Ray optics and total internal reflection are in the optics unit of the JEE (Main) 2026 syllabus; check the syllabus for your exam year.
Light bouncing down a fibre is easier to understand when you can watch it. The Fermi Simulation Lab has interactive simulations for ideas that are easier to see than read.
How did the Nobel Prize recognise him?
Kao received half of the 2009 Nobel Prize in Physics "for groundbreaking achievements concerning the transmission of light in fibers for optical communication". The other half went to Willard Boyle and George Smith for inventing the CCD sensor, an imaging semiconductor circuit. His Nobel autobiography describes moving to the Chinese University of Hong Kong in 1970 while still returning to STL in the summers to follow the fibre research. For the laser side of the story, read our post on Charles Townes and the maser.
What can a student learn from how Kao worked?
Kao asked where the light was actually going before trying to fix the fibre. Once he showed the cause was impurity, the goal became clear and measurable: 20 dB per kilometre. Find the real cause of a problem, then set a target you can measure.
Practice routine:
Find the critical angle for core and cladding indices of 1.50 and 1.45.
Convert 3 dB, 10 dB and 30 dB of loss into the fraction of power remaining.
A 50 km link loses 0.25 dB per km. What fraction of the light arrives?
Next step: solve the three practice problems and check each decibel answer by counting how many factors of ten you divided by.
