In the last few days, there has been a spate of reports that the incandescent bulb is on its way back.  This relates to work by a group of authors at MIT plus one at Purdue University in Indiana, featured in a news report from MIT:

A nanophotonic comeback for incandescent bulbs?

Many of us might look forward to this, having found compact fluorescent lamps troublesome, and LED lights a bit weird.

It relates to this very recent publication,:

Tailoring high-temperature radiation and the resurrection of the incandescent source

Not being able to access the full paper, I have at least dug out the abstract, which helps me to bridge some of the gap between popular reports and scientific ones full of sold-state physics jargon.  It begins with a reference to the general problem, namely that of converting light to electricity and vice versa.

The general difficulty is a mismatch between the spectrum of what is available and what would be desirable for maximum efficiency.  With solar cells, only part of the Sun’s spectrum matches the wavelength range required for the cells to do their “solid state stuff”, which is making electrons in the material jump from one state to another in the good ol’ quantum mechanical way, then jumping back again to release electrical energy.

The reverse problem is found with incandescent light bulbs.  These use filaments made of tungsten, one of the most refractory metals known (in German it is called Wolfram, which etymologically means wolf-toughness.)  Even so, one can only heat them to about 3000 Kelvin, half the temperature of the Sun’s surface, more like the temperature of the dimmest red dwarf stars or the most puffed up red giants.  As with those stars, only a small fraction of the energy comes off as visible light, most of it appears in the infrared and is only perceptible as heat.  Which means one does not get a good return on the “electrickery” [1] that one has paid for.

The ideal thing to do would be to tailor the thermal emission spectrum of the hot tungsten to fit the visible light spectrum.  Similar tricks have been performed with warm objects (less than 1000 Kelvin) to narrow the broad thermal infrared emission into a tighter band, but this has proved difficult for objects as hot as a working light bulb filament.

The method employs materials called photonic crystals, which act as 1, 2 or 3-dimensional diffraction gratings.  One type of 3-D photonic crystal is the “inverse opal”.  Natural opals are made from lattices of uniformly size silica spheres which have slowly settled into a pattern like the atoms in a crystal on the cubic system: solid argon is an example where all the atoms are of one type.  Inverse opals are made by allowing spheres of “soft” material to settle, infiltrate them with “hard” material, and then remove the “soft material”.  This was first reported from the University of Toronto in 2000 using some rather horrendous reagents, but a much more gentle procedure with more amenable materials is Inverse Opal Photonic Crystals from the University of Minnesota Department of Chemistry makes polymethyl methacrylate (PMMA) spheres, fills the gaps between them with silica, and finally burns away the PMMA.  (Note: the Wikipedia article is unreliable at this particular point: one cannot dissolve away polystyrene spheres with hydrochloric acid!)

Because of the scale of the structure (half the wavelength of light) and the fact that diffraction gratings work by interference (similar to the thin-film interference that is used in antireflection coatings) such an application can be described as a nanophotonic interference system.

The authors have now generated a system to work at light bulb temperatures.  They go on to say:

Here, we show that a plain incandescent tungsten filament (3,000 K) surrounded by a cold-side nanophotonic interference system optimized to reflect infrared light and transmit visible light for a wide range of angles could become a light source that reaches luminous efficiencies (
40%) surpassing existing lighting technologies, and nearing a limit for lighting applications. We experimentally demonstrate a proof-of-principle incandescent emitter with efficiency approaching that of commercial fluorescent or light-emitting diode bulbs, but with exceptional reproduction of colours and scalable power.

If one were simply to take an ordinary filament carrying the current normally used with a light bulb, and instantaneously coat it with the photonic crystal, the unwanted infrared would be reflected back, and the system would cook itself in a flash.  Rather, it will run at a much lower current, since the actual energy loss from the filament will be much reduced.

And the colour control sounds good to me.  My first encounter with compact fluorescent lighting was when visiting a flat where there was some Venetian violet glass on display in a cabinet.  Presumably because of a strong spike in the mercury emission spectrum in the indigo, the glass appeared indigo rather than violet.  And I am not so sure that LED light is all that comfortable, either.

[1] as called by Catweazle.