Electrolysis of Water
| Scientific Paper | |
|---|---|
| Title | Electrolysis of Water |
| Read in full | Link to paper |
| Author(s) | Philipp M Kanarev |
| Keywords | Electrolysis, Water, Temperature, Hydrogen |
| Published | 2009 |
| No. of pages | 10 |
Read the full paper here
Abstract
Atomic hydrogen exists in a plasma condition at temperature 2700-5000 C. If the formation of molecules of hydrogen at electrolysis of water goes by branch of its atoms from molecules of water, in a phase of an atomic condition of hydrogen in electrolytic solution the specified temperature should be formed, but it is not present.
Overview
Kanarev opens with a question that is genuinely worth asking. Atomic hydrogen is a plasma-temperature species; if electrolysis proceeded by first breaking water into free atoms and then pairing them into H2, the solution would have to pass through thousands of degrees, which it plainly does not. His answer is that hydrogen molecules are never assembled from free atoms at all: they are released "in the synthesized condition" directly out of clusters of water molecules, already bonded. To make that picture concrete he applies his own structural model of the atom — developed at length in The Foundation Physchemistry of the Microworld — in which electrons do not orbit but interact linearly with the protons of the nucleus.
The practical target is stated in the first paragraph. The best electrolysers spend about 4 kWh of electricity per cubic metre of hydrogen, and burning that cubic metre returns about 3.5 kWh; hydrogen becomes a competitive energy carrier only if the input can be pushed down towards 1 kWh/m3. The second half of the paper claims to have gone much further than that. A "lowcurrent electrolyser" of conical steel electrodes (Russian patent no. 2227817), driven by pulses and left disconnected for most of the run, is reported to produce hydrogen with an energy content ten times — on one reading of the instruments, 291 times — the electrical energy supplied.
The argument
Structures of water and its clusters
In Kanarev's model the oxygen atom carries two electrons on its axis and six on a perpendicular ring; the ring electrons' combined field pushes the axial pair further out, making them the valent electrons. The two hydrogen electrons join those, so that the water molecule is linear, with a bare proton exposed at each end and a negatively charged ring around the middle. On cooling, the ring electrons emit photons and draw in, pushing the axial electrons out and lengthening the molecule — which he offers as "the main reason of increase in the sizes of molecules of water at their freezing". Water molecules then link into clusters either proton-to-proton (weak, because "the size of a proton on three order is less than size of an electron") or proton-to-ring-electron; the weakness of both is offered as the explanation of water's fluidity, and the six-beam version as the origin of the snowflake.
Photons, music and prayer
Heating one litre of water from 20 °C to 100 °C takes 335.2 kJ, which per molecule is
- Eb = 335.2×103 / (6.02×1023 × 1.6×10−19 × 55.56) = 0.063 eV
Dividing by 80 gives 0.00078 eV per degree, an energy Kanarev places in the "relic range" of his Table 1. The smallest step he allows is 0.000022 eV, the energy of a photon of wavelength 0.056 m, so the minimum temperature gradient of water is 0.000022/0.00078 ≈ 0.03 °C. On this basis he asserts that quiet classical music and a praying voice cause water to form symmetric six-beam clusters, that jazz forms "ugly structures" and is therefore "weighty proof of harmful influence of jazz music on health of the person", and that a mobile telephone radiates photons which destroy clusters outright.
Faraday's law and the conventional cost
The conventional calculation is done correctly. Two faradays are required per mole of hydrogen, 2 × 96,485 = 192,980 C, or 192,980/3600 = 53.6 A·h/mol; at a cell voltage of 1.70 V that is 53.6 × 1.70 = 91.12 W·h per mole, and (1000/22.4) × 91.12 = about 4.1 kWh per cubic metre. Kanarev notes that this "give the result conterminous to experiment". Separately, a cubic metre of hydrogen weighs 1000 × 0.09 = 90 g, and at 142 kJ/g contains 12,780 kJ = 3.55 kWh.
The lowcurrent electrolyser
The cell has conical steel electrodes with gaps "imitating annual rings of trunks of trees", and runs at 1.5–2.0 V and 0.02 A. Both electrodes are the same steel, which Kanarev says "excludes an opportunity of formation of a galvanic cell", yet he records a standing potential difference of about 0.1 V with no electrolyte present, rising when solution is added, always with the positive sign on the top electrode. Gas evolution continues for many hours after the supply is switched off. Table 2 reports six cycles of 10 minutes on and 50 minutes off:
- voltmeter 11.4 V, ammeter 0.020 A → P = 0.228 W·h; oscillograph 0.40 V, 0.01978 A → P′ = 0.0081 W·h
- solution mass loss 0.60 g, of which 0.06 g assigned to evaporation, leaving m″ = 0.54 g of water "passing in gases"
- hydrogen ΔM = 0.54 × 1.23 × 0.09 = 0.06 g, energy content W = 0.06 × 142/3.6 = 2.36 W·h
- efficiency W × 100/P = 1035.1 %, and W × 100/P′ = 29,135.8 %
Kanarev concludes that the cell "possesses properties of the condenser and a source of an electricity simultaneously", and that better catalytic electrode materials should allow decomposition of water "without expenses of electric energy".
Assessment
Every calculation in the first half of the paper checks out, and it is worth saying so plainly. 335.2 kJ is the right heat for 1 litre of water over 80 K; 0.063 eV per molecule is correct; 0.00078 eV per degree follows; a 0.056 m photon does carry 2.2×10−5 eV (hc/λ = 1240 eV·nm / 5.6×107 nm); the ratio is indeed about 0.03. Table 1's wavelength/energy pairs are right throughout, including the relic maximum at about 1 mm and 1.2×10−3 eV, and the visible band at 1.60–3.27 eV. The Faraday-law block is standard and correct, as is the 3.55 kWh energy content of a cubic metre of hydrogen. The opening question about atomic hydrogen is also a fair one, and the textbook answer — that discharge and recombination occur on the electrode surface, where adsorbed H atoms are stabilised by the metal and never exist as free gas-phase atoms — is a real answer that the paper does not engage with.
The trouble begins with the structure and becomes acute in Table 2.
The linear water molecule. Kanarev's H2O is a straight line with a proton at each end. The measured H–O–H angle is 104.45°, fixed by microwave rotational spectroscopy and confirmed by neutron and X-ray diffraction, and it is the reason water has an electric dipole moment of 1.85 D. A linear, symmetric molecule would have zero dipole moment, and with it none of water's dielectric constant, none of its solvent power for ions, and no hydrogen bonding of the kind that gives it its boiling point. The model contradicts the single most-measured fact about the molecule it is about. The related claim that pure water cannot carry a current because "linear clusters have on both ends the same charges" is likewise contradicted by measurement: pure water has a conductivity of about 5.5×10−6 S/m from autoionisation, with Kw = 10−14.
Music and prayer. These claims are introduced as things "experimentally established" and "already proved", with no citation of any kind. More decisively, they are excluded by the paper's own numbers. Kanarev's mechanism requires photons of the cluster-bond scale he has just calculated, 2.2×10−5 eV, corresponding to about 5 GHz. A photon at an audible frequency of 1 kHz carries hf = 4×10−12 eV — seven orders of magnitude too little — and in any case music reaches water as an acoustic pressure wave, not as photons at the acoustic frequency. His Table 1 places nothing musical anywhere near the relic range it identifies. The mobile-phone claim is at least in the right band (1.8 GHz gives 7×10−6 eV), but no measurement of it is reported.
Table 2 fails Kanarev's own Faraday's law. This is the decisive check, and it uses only material the paper supplies. The cell passed 0.020 A for 60 minutes, that is 72 coulombs. By the law Kanarev himself sets out four pages earlier — two faradays, 192,980 C, per mole of H2 — that charge can liberate 72/192,980 = 3.73×10−4 mol, or 7.5×10−4 g of hydrogen. Table 2 claims 0.06 g: eighty times more. The paper therefore contains its own refutation. Everything downstream — the 2.36 W·h of "received hydrogen", the 1035 % and 29,136 % efficiencies — rests on that eightyfold excess. Taking the Faraday-law yield instead, the hydrogen produced carries about 0.03 W·h against 0.228 W·h drawn, an efficiency near 13 %, which is what one would expect from a cell run this way.
Two further defects in the same table. First, the source of the 0.06 g is a weight loss of the solution over a six-hour run in which the cell was disconnected for five of those hours; 0.06 g is assigned to evaporation and the remaining 0.54 g to electrolysis, on no stated basis. No gas was collected, measured or analysed. An open vessel of warm electrolyte will lose far more than 0.06 g of water to evaporation in six hours, and steel electrodes in electrolyte corrode, evolving hydrogen chemically at the expense of the metal rather than of the supply — which would also explain gas continuing to bubble "within many hours" after switch-off. No electrode mass was recorded. Second, the reported gain factors K = E″/P = 5.25/0.228 = 23.03 and K′ = 5.25/0.0081 = 648.15 divide a quantity in W·h per gram by a quantity in W·h, which is not a ratio of like things; corrected to 5.25/0.420 and 5.25/0.015 they become 12.5 and 350, so even on the paper's own terms the headline numbers are overstated by 1/0.54.
The two instruments. A voltmeter reading 11.4 V and an oscilloscope reading 0.40 V on the same cell differ by a factor of 28.5, and the paper does not explain which is right or why they disagree. Kanarev adopts the smaller, from which the 29,136 % follows — but 0.40 V is below the 1.23 V reversible decomposition potential of water, and far below the 1.70 V he used in his own correct calculation, so at that voltage sustained electrolysis is thermodynamically impossible. The result he prefers is the one his own earlier page excludes. There is also a standard pulsed-power trap here: for a pulsed waveform the mean power is ⟨VI⟩, which is not ⟨V⟩⟨I⟩; multiplying two separately averaged oscilloscope readings can understate the delivered power by a large factor. Finally, the standing 0.1 V between "identical" steel electrodes, rising on addition of electrolyte and always of the same polarity, is straightforward evidence that a galvanic cell is present — differing oxide films, surface states and oxygen access on a conical pair are enough — which undercuts the sentence in which that possibility is dismissed.
One smaller point: the remark that "on what basis electrical engineers consider, that electrons move to circuits of a direct current from a minus to a pole remains a secret" is a misreading of a convention. Electrons do move from the negative terminal to the positive one in the external circuit; conventional current is defined in the opposite sense for historical reasons, and Kanarev's own account of the cathode giving electrons to protons is exactly the standard picture.
The paper is at its best where it is most conventional, and its central experimental claim is refuted by the law it correctly states in its own middle pages.