The Sudbury Neutrino Observatory: Counting the Neutrinos Nobody Could See

Why Heavy Water Was the Whole Idea

Ordinary water contains hydrogen, whose nucleus is a single proton. Heavy water contains deuterium, whose nucleus is a proton bound to a neutron. That extra neutron is what made SNO different from every solar neutrino experiment before it.

A deuteron can be broken apart by any of the three neutrino types, through a reaction that does not care about flavour at all. It can also undergo a reaction that only electron neutrinos can drive. So a single detector, running continuously, produced two independent counts from the same neutrino flux: how many electron neutrinos arrived, and how many neutrinos arrived in total.

A third channel, elastic scattering off electrons, responds mostly but not exclusively to electron neutrinos, giving a cross-check that sits between the other two. Earlier experiments, including Super-Kamiokande, had access to only that kind of channel, which is why they could establish a deficit but not its cause.

The heavy water was not bought. A thousand tonnes of it, worth several hundred million Canadian dollars, was lent from the national reserve held for nuclear reactors, used for seven years, and returned. There was no other way to obtain that quantity.

Two Kilometres Down, and Exceptionally Clean

The detector sat at the 6,800-foot level of an active nickel mine, about 2,070 metres below the surface - roughly 6,000 metres of water equivalent in shielding, deeper than any comparable detector of its era. Physicists and construction crews reached it by the miners' cage and then a walk of more than a kilometre.

The heavy water was held in a transparent acrylic vessel 12 metres across, itself suspended inside a barrel-shaped cavity filled with 7,000 tonnes of ultrapure ordinary water. Around 9,600 photomultiplier tubes on a geodesic support structure watched the inner volume for Cherenkov light.

Contamination discipline was severe. Every component was cleaned to standards closer to semiconductor manufacturing than to mining, because a trace of uranium or thorium in the acrylic or the water would have produced neutrons indistinguishable from the flavour-blind signal the experiment depended on.

The experiment ran in three phases, each using a different method to detect the neutrons released when a deuteron broke apart: first the heavy water alone, then with dissolved salt added to improve neutron capture, then with dedicated neutron detectors installed. Three independent methods, three consistent answers.

The Observatory in Numbers

SNO took data from 1999 to 2006. The site is now SNOLAB, an expanded underground facility hosting several other experiments.

• Location: Creighton mine, Sudbury, Ontario, Canada, about 2,070 metres underground
• Target: 1,000 tonnes of heavy water on loan from Atomic Energy of Canada Limited
• Vessel: acrylic sphere 12 metres in diameter, surrounded by 7,000 tonnes of ultrapure light water
• Photosensors: approximately 9,600 photomultiplier tubes
• Channels: charged current (electron neutrinos only), neutral current (all three types), elastic scattering
• Operating period: 1999 to 2006, in three phases using different neutron detection methods
• Recognition: Arthur McDonald, Nobel Prize in Physics 2015, shared with Takaaki Kajita

How the Puzzle Was Closed

The result, published in 2001 and confirmed in 2002, was arithmetic of unusual clarity. The flavour-blind channel measured a total neutrino flux from the Sun that matched the standard solar model closely. The electron-neutrino-only channel measured roughly a third of that number.

The neutrinos were therefore all arriving. Two thirds of them had simply stopped being electron neutrinos somewhere between the Sun's core and Ontario. The solar models had been right all along, and Raymond Davis Jr., whose chlorine experiment first found the deficit in the 1960s, had been measuring something real.

This is neutrino oscillation, and it requires that neutrinos have mass. Super-Kamiokande had reached the same conclusion in 1998 from a different direction, using atmospheric neutrinos and the distance they travelled through the Earth. Two experiments, different sources, different technologies, different systematic errors, one answer.

That convergence is why the result was accepted quickly and has not been seriously challenged since. It is also a useful standard to hold other claims to: a measurement becomes a discovery when an independent experiment that could have contradicted it does not.

What SNO Did Not Establish

It did not measure the neutrino masses, only demonstrated that at least two of them are nonzero. The absolute scale remained open, and is the question KATRIN was later built to attack directly.

It also did not, and could not, say anything about extracting usable energy from the solar neutrino flux. The scale of the experiment makes the point: a thousand tonnes of borrowed heavy water, two kilometres of rock overhead, seven years of running, to count a few thousand interactions in total.

That number deserves to be stated whenever the abundance of solar neutrinos is invoked. Around 65 billion of them cross every square centimetre of Earth per second, and SNO - purpose-built, exquisitely shielded, using the most reactive practical target available - recorded a few per day. The flux is enormous and the coupling is negligible, and both facts are true at once.

The separate question addressed by neutrinovoltaic research concerns ambient radiation and thermal fluctuations at an engineered material surface, not the capture of solar neutrinos. SNO speaks to the second and not to the first.