We pointed a very powerful computer at some glowing dye molecules and found something that musicians have known about for 2,000 years hiding inside the chemistry.
Have you ever used a highlighter marker? The bright, glowing color comes from dye molecules that absorb certain colors of light and bounce others back at your eyes. Cyanine dyes are a special family of these molecules used in medical imaging, solar cells, and bioscience they glow in the near-infrared, which is the light just beyond red that cameras can see but human eyes can't.
What makes them special is their shape: they are long chains of carbon atoms, alternating between single and double bonds, with a positive electrical charge at each end. Think of them like a molecular tug-of-war two charged ends pulling on a chain of electrons in the middle.
Imagine a guitar string. The longer the string, the lower the note. Cyanine dyes work the same way: the longer the chain, the redder (lower-energy) the light it absorbs. A short Cy7 molecule absorbs around 750 nm (deep red). A longer chain should keep shifting toward infrared. That was the prediction. That is NOT what happened.
For the first four molecules Cy7⁺, Cy9⁺, Cy11⁰, and Cy11⁺ everything worked exactly as predicted. Longer chain = redder light. The computer confirmed real minimum-energy structures (no imaginary wiggles, which means the geometry is stable). We also found something interesting in the orbital energies.
In the molecule chain, the bonds aren't all the same length. Single bonds are longer (~1.43 Å), double bonds are shorter (~1.37 Å). The Bond Length Alternation (BLA) is just the average difference between them. Think of it as: how much does the molecule "breathe" how alternating is it? A perfectly even chain would have BLA = 0. A very alternating chain has a high BLA.
BLA is like wave height. A perfectly calm ocean = BLA of 0. Rough seas = big BLA. These dye molecules live somewhere in between, and exactly where they sit tells us how their electrons are distributed and what color light they absorb.
When we divided the average BLA by a very specific number the Pythagorean Comma (0.013643...) we got 3.685. All four confirmed molecules land in the same band between 3× and 4× the comma. The comma is a music theory number. More on this later.
Here is where the story gets strange. Based on the pattern from Cy7–Cy11, the next molecule (Cy13, a longer chain) should absorb around 1200 nm very deep infrared. Instead, the DFT calculation produced something completely unexpected.
Here's the CPCS explanation: the bond alternation at Cy13 (BLA/δ = 4.602 ≈ 9/2) means the molecule has a kind of internal "resonance frequency" that matches its fifth energy level. It's like a tuning fork that only vibrates at a specific pitch. When the molecule's structure hits this ratio, light couples most strongly to the fifth rung of the energy ladder not the first. The number 9/2 is related to the perfect fifth in music (the ratio 3:2, squared). That's not a coincidence according to the CPCS framework.
After the Cy13 shock, we ran Cy15 an even longer chain. The calculation took 16 hours and 2 minutes on the laptop, completed normally, and produced 71 geometry steps and 70 snapshots of the absorption spectrum.
The chain BLA collapsed almost to zero: 0.01933 Å = 1.417× the comma. The bridge was nearly fully smoothed out all the C–C bonds were almost the same length. And the dominant absorption:
Red light. The molecule came back from blue toward red.
f = 5.495 means Cy15 absorbs light more strongly than any other molecule in the series. When the bridge delocalizes (BLA → 0), the electrons can slosh freely across the whole chain and that gives the maximum possible "brightness." The resolution of the anomaly is the loudest moment.
Put all five molecules side by side and you see a shape that no one predicted:
The series starts in the NIR (invisible), jumps 86% higher in energy at Cy13 (blue-green, visible), then partially returns to red at Cy15. But it doesn't make it all the way back. The tonic band center would be around 843 nm. Cy15 lands at 683 nm 160 nm short. It can't fully return. That gap is the point.
Here is the strangest part. The bond-length alternation ratios at the two weird molecules are not random numbers. They are very close to two specific musical fractions:
In music, if you keep going up by perfect fifths from C, you get: C → G → D → A → E → B → F# → ... After 12 steps you almost get back to C but not quite. You overshoot by a tiny amount called the Pythagorean Comma (about 1/4 of a semitone, or 23.46 cents). The spiral never closes perfectly. Musicians have struggled with this for 2,000 years.
In music this is called I → V → II: you leave home (tonic), go to the dominant (perfect fifth), then land on the supertonic. It's a classic pre-cadence move you're setting up to return home, but the series stops at D. It never completes the journey back to C. Just like the Pythagorean spiral you can never fully close it.
The cyanine series at Cy15 lands 0.47 eV above where it started. To fully return to the tonic band, it would need to be at 843 nm. It's at 683 nm. The difference that gap is what this paper calls the molecular Pythagorean comma: a permanent energy offset left behind by traversing the 3:2 ratio once. Just as the musical comma is the error left by stacking 12 fifths, the molecular comma is the error left by one molecular fifth-step.
In January 2026, before any of the Cy13 or Cy15 calculations existed, a notebook sketch was drawn showing exactly this arc: red (NIR) dipping to blue (Cy13 anomaly) and returning to red (Cy15). The sketch even labeled it as "Cyanine dye = Electromagnetic Spectrum?"
The Cy15 calculation completed March 31, 2026 three months after the sketch. The DFT output matched the sketch in both direction and ordering. The prediction preceded the computation.
If the cyanine series is genuinely following the spiral of fifths, then Cy17⁺ should NOT return to the NIR band. It should continue spiraling landing at a BLA/δ ratio of about 1.688 (spiral step 3), and a wavelength different from everything before.
Cy17⁺ dominant bright state: NOT in the 760–920 nm range. Should continue to a new position on the harmonic spiral. BLA/δ ≈ (3/2)³ octave-reduced ≈ 1.688. If this is wrong, the CPCS framework needs revision. If this is right, the cyanine series is the first molecule shown to traverse the Pythagorean spiral of fifths in its light absorption properties as its chain length grows.