Science Contrarian review: Pappas-PT (US 5,556,418, "PAP-IMI")

Reviewer role: Science Contrarian (STEP 2). Judges by experiment, builds the strongest honest case, examines how the claims were handled, and holds the inventor to the same standard. Tags: [read] document, [web] URL, [memory] with confidence. Source base: one patent (7 pages, OCR, no source gap), the deep read, and the Investigator's leads (sites.md, research/ notes and files). I read the deep read, the extracted patent text, the Leisner abstract, and the methods sections of Viti 2023 and Keilani 2025 myself.

Corrections to Step 1

1. facts.json says "No clinical or biological data are given" and the deep read confirms it [read: deep/ section 6]. Agreed. But claims.md and facts.json state the 10^6 V/m air-versus-membrane figures as if consistent; the deep read correctly shows the patent prints 10^6 V/m for BOTH the membrane field and air strength [read: page image p.4]. facts.json should carry this as the patent's own internal inconsistency. Real membrane field is 70 mV over about 5-10 nm (typical bilayer, [memory] high) = 0.7-1.4 x 10^7 V/m; the patent's "several Angstrom" gives about 10^8.
2. claims.md claim 2 says bursts are "tens to hundreds of MHz". The scope readout (1.3 MHz minimum, 240.3 MHz maximum, 110 MHz average) is a statistic with no stated gate, identical on FIG. 2 and FIG. 3 [read: deep section 4]. It does not show a spectrum. No vertical (volts or amps) scale is printed on either oscillogram.
3. claims.md claim 3 says "bursts shorter than the time needed for air to break down". The patent gives burst length as under 10^-6 s (description), 10^-4 to 10^-6 s (claim 14) and under 10^-4 s (claim 16) [read: deep section 5]. These differ by up to 100x and 10^-4 s is long compared with air breakdown formative times. Also claims 1 and 17 (the actual legal claims) mention none of the 1-100x gain, voltage or negative resistance; only resonance absorption.
4. facts.json "mainstream_conflict" says the excess voltage "may come from stored energy in L and C". That is right and is the heart of the matter (see claim 1 below). Fine, but it was left as a guess; below it is calculated.
5. Investigator point 3.2: the Greek application number differs (920100309 vs 920200309 printed); likely a misprint, not verifiable here. [web] Espacenet note in sites.md.
6. Investigator lead P1.6: the Physiology PDF figures (">1 GW", "888 magnetic nanopulses at about 80 J each") are the inventor's later claims. Arithmetic check: 1 GW over about 100 ns is about 100 J, consistent with ~80 J, so no internal conflict; 80 J is the same order as the manufacturer figures 60 J (Viti) and 96 J (Keilani). 888 pulses x 80 J is about 71 kJ delivered per session.

What the physics of the device actually is (my reconstruction from published numbers)

Independent papers describe the device as a capacitor discharged through a spark gap into a spiral coil: up to 40 kV, peak currents "several 10 kA" (Viti) or up to 10 kA (Keilani), about 60 J (Viti) or 96 J (Keilani) per pulse, 50-150 mT, base frequency about 240 kHz, pulse about 50 microseconds, 1-8 Hz, and "in the maxima and minima of the damped oscillation, high-frequency oscillation peaks in the megahertz to gigahertz range" [web] https://pmc.ncbi.nlm.nih.gov/articles/PMC10001461/ and https://pmc.ncbi.nlm.nih.gov/articles/PMC12534222/. These are manufacturer-supplied numbers repeated by the authors, not independent measurements.

I checked whether those numbers hang together (my calculation, [memory] high on the formulas). Take the patent's coil: 30 cm diameter (R = 0.15 m), 3 turns, wire cross-section 6 mm^2 (radius a about 1.4 mm). Loop inductance L = mu0 N^2 R [ln(8R/a) - 2] = 8.1 microhenry. 60 J at 40 kV needs C = 2E/V^2 = 75 nF. Then f = 1/(2 pi sqrt(LC)) = 205 kHz (reported: about 240 kHz), peak current I = V sqrt(C/L) = 3.9 kA, and centre field B = mu0 N I/(2R) = 48 mT (reported 50-150 mT). So a plain LC discharge reproduces the reported frequency and field. Strength for the device: the numbers are internally coherent and the energy is simply the stored capacitor energy. "Several 10 kA" would need about 400 J at this inductance (E = L I^2/2 at 10 kA), so that one published figure is inconsistent with 60-96 J unless the coil inductance is smaller. This is a reviewer inference with assumed coil dimensions; the patent gives no L, C or voltage.

Claim 1: arc plasma gives 10-100x transient voltage and current and "infinite then negative gain" (energy creation)

a. Mathematics. Internally inconsistent as an energy argument. "Gain" is defined as a ratio of resistances (infinity/0.5 = +infinity; 0.5/(-100) < 0) [read: p.5]. A ratio of two resistances is not a power ratio. A resistance of infinity before ignition just means an open circuit with zero current. The patent itself says the gain of some electrons "is achieved by a corresponding decrease in the energy of the remaining electrons or ions" [read: p.4], which is energy conservation within the plasma, not creation. A negative differential resistance (dV/dI < 0) is a real and well-documented property of arcs ([memory] high); it is not a negative resistance V/I and it supplies no net power. No energy accounting, capacitance or voltage is given anywhere [read: deep sections 2-3]. b. Experimental consistency. A transient voltage 10-100x the supply is ordinary: LC ringing, switching with a spark gap and inductive kick (L dI/dt) give large overshoot, and power at the 0.001% duty cycle is 10^5 times the average power by arithmetic alone [read: p.6 for duty cycle]. These do not need new physics. The numbers above reproduce the device with ordinary circuit theory. c. Tested? The inventor reports his own measurement (1989 Einsiedeln, 1990 Tesla Symposium, 1991 IECEC "Energy Creation in Electrical Sparks and Discharges: Theory and Direct Experimental Evidence", vol. 4 pp. 416-423) [read: p.4; web per sites.md]. I did not obtain the 1991 paper; it is the one document that could show a power-balance. No independent power-balance test is on record. His 1998 letter claims "20% excess heat on the probe alone" [web] research/files/NEN-1998-06-N199806.pdf, p.28; I did not read the measurement method (a letter, no data). Absence: a direct, instrumented energy-in vs energy-out test of the spark-gap discharge has not been done by anyone I can find. d. If true: conservation of energy would need an exception for oscillating arcs; instead, a mundane reading (stored-energy release, short duty cycle) costs nothing. Strongest honest case: the observation of high transient peak voltage/current in a pulsed arc is real and easily verified; the patent's contribution is applying it. The "gain" language is the problem, not the observation. Verdict: split. Weak form (peak voltage/current well above supply and average values in a pulsed discharge): Established (ordinary circuit physics, my calculation). Strong form (net energy gain/"infinite gain" from negative arc resistance): Refuted by calculation as argued in the patent (resistance ratio is not an energy ratio; no accounting; the patent's own text concedes a loss elsewhere), and never properly tested for energy balance by anyone, including the inventor (1991 paper not read).

Claim 2: sub-microsecond bursts at plasma eigenfrequencies, delivered by an induction coil

a. Consistent in form: damped LCR curve with RF ringing at extrema is what a spark-gap discharge does. The burst duration is stated inconsistently (10^-6 s, 10^-4 to 10^-6 s, under 10^-4 s) [read: deep section 5]. The "few Hz" charging rate versus 0.5-1 MHz base in FIG. 2 (my reading) versus 240 kHz in later papers is not reconciled by the patent. b. Consistent with experiment: yes. Keilani describes exactly this waveform [web] PMC12534222, with megahertz-to-gigahertz peaks. "Characteristic eigenfrequencies of the gas" is not shown: the 1.3-240 MHz readout is not a line spectrum and no nitrogen line is identified. A spark-gap's RF content is set largely by the circuit's parasitic L and C and the gap, not an atomic transition ([memory] medium; reviewer inference). c. Tested: waveform appears in the patent's oscillograms and is repeated in two 2020s clinical papers (manufacturer figures). No independent spectrum measurement found. d. Little would have to change. Verdict: Plausible, untested for "bursts exist"; the "characteristic eigenfrequency of the gas" part Never properly tested (no spectrum shown).

Claim 3: sub-breakdown-time pulses exceed air's dielectric strength 1-100x without ionisation; 10^6 V/m cell-scale fields non-invasively

a. Internally inconsistent argument: the patent equates 70 mV/several Angstrom with air's 10 kV/cm = 10^6 V/m [read: page image p.4]. Air actually breaks down near 3 x 10^6 V/m (DC, [memory] high). The membrane figure should be about 10^7-10^8 V/m. b. Experimental: the weak form is real. Fast-rise pulses do exceed DC breakdown strength: in nanosecond discharges in atmospheric air breakdown field about 75 kV/cm was measured versus about 30 kV/cm DC [web] search result summarising the nanosecond-breakdown literature, https://pmc.ncbi.nlm.nih.gov/articles/PMC4110921/ (and overvoltage rises with rise rate); I did not open the paper itself, so the 75 kV/cm figure is a secondary report, a factor of about 2.5, not 100. Two further errors in the premise (my calculation): (i) induction does not apply a field to air at all; the field in tissue is the induced E = (r/2) dB/dt. With 50 mT in a quarter of 240 kHz (1.04 microseconds), dB/dt = 4.8 x 10^4 T/s, so at r = 10 cm E is about 2.4 x 10^3 V/m, a thousand times below 10^6, with no air breakdown near it. (ii) A cell does not need 10^6 V/m applied: the Schwan relation gives induced membrane voltage of about 1.5 E R (R = cell radius) [web] https://pubmed.ncbi.nlm.nih.gov/2248989. With R = 10 microns, E = 2.4 kV/m gives about 36 microvolts (valid to about 1 MHz per the search summary, and falling at higher frequency), tiny compared with 70 mV. So the required-field premise is wrong in both directions, and the device's low-frequency field is far too weak to move transmembrane potential by the patent's mechanism. For GHz bursts at gigawatt peak, the unknown is the real field in tissue; the patent measures none. c. Tested: nobody has measured the induced field in tissue for this device in a source I found. d. If true: nothing; short-pulse overvoltage is standard high-voltage physics. Verdict: split. "Fast pulses exceed DC air breakdown strength": Established, but by factors of a few, not 100 (the 100x figure is unsupported). "Therefore cell-scale 10^6 V/m fields are delivered non-invasively": Internally inconsistent (membrane and air strength given as equal) and Refuted by calculation for the quoted field and frequency (induced E about 2 kV/m, membrane effect about tens of microvolts).

Claim 4: cell constituents resonate with plasma eigenfrequencies; selective absorption triggers selected endothermic reactions (nitrogen example)

a. No equations. The reaction scheme XAB + Y + plasma energy A -> XA + YB is bookkeeping, not a mechanism [read: p.5]. Claim 1 and claim 17 say atoms "similar to atoms of said specific element" in the patient resonate with the plasma atoms; but a nitrogen atom bonded in a molecule has lost the free-atom spectrum, and free nitrogen emission lines are optical/UV, not radio. The patent gives no frequency for any "nitrogen resonance" [read: deep section 7]. b. Experimental: NMR does give element-specific radio-frequency resonance, in a static field (nitrogen-14 about 3 MHz per tesla, [memory] medium; hydrogen 42.6 MHz/T, [memory] high). The inventor himself reframed the idea later as NMR/EPR with no static field (US 2006/0220648 A1, US 7,151,372 B2) [web] https://patents.google.com/patent/US20060220648A1/en. That is a legitimate-sounding resonance concept, but NMR energy per photon (about 10^-7 eV) is far below any chemical bond energy, so it cannot drive endothermic reactions ([memory] high, arithmetic: h x 100 MHz = 4 x 10^-7 eV). The 1998 letter ("C to Fe", "30 keV", "20% excess heat") and the 2006 Physiology text ("Na + O -> K", cell as "(ATP) controlled microscopic Nuclear Reactor") are further, much stronger claims by the inventor [web] research/files; no data or independent replication found. Biological transmutation literature (Kervran, Vysotskii and Kornilova, Moawad 2013) is related work, not independent confirmation of this device. Ion cyclotron resonance (Liboff, cited in the patent, US 5,160,591) is a mainstream-adjacent frequency-selective concept with its own contested literature. c. Tested: no controlled test of selective reaction activation is on record. Leisner 2002 (below) tested bone healing and was negative; it is not a test of resonance selectivity. d. If true: biological chemistry could be steered by weak RF at low photon energy; this would require a new coupling mechanism. Verdict: Never properly tested (not dismissed after a test, simply no controlled test of selectivity); the nuclear transmutation and "excess heat" extensions: Never properly tested, with no data in hand to evaluate.

Claim 5: restoring transmembrane potential (cancer 15 mV, unhealthy 50, healthy 70) is therapeutic; bursts "pasteurise" degenerate cells

a. Consistent with itself apart from the field problem in claim 3. The numbers 70/50/15 mV are cited without reference [read: p.4]. Reduced membrane potential in some tumour cells is an established observation in the cancer-bioelectricity literature ([memory] medium; the patent cites none). b. Experimental record, both signs, stated in full: - Leisner et al. 2002, Hebrew University: controlled rat fracture model, untreated controls; PAP-IMI group healed slower and had more fibrous callus; "do not support the claim that PEMF generated by the PAP-IMI stimulate osteogenesis" [read: research/files/pubmed-11913824-Leisner2002.txt; web PMID 11913824]. Limits: one species, one model, no sham device (untreated control, not sham), fresh fractures where PEMF effect is itself conflicting per the abstract. A real controlled negative result. - Viti 2023: single-blind randomised sham-controlled, n = 32 (17/15), chronic musculoskeletal pain; heart-rate-variability indices (SDNN, RMSSD, NN50, pNN50, HF) rose after PAPIMI, not after sham; authors declare no conflict [read: PMC10001461]. Measures autonomic response, not membrane potential, and a short-term surrogate. - Viti 2025 (single arm, n = 30; pain fell, no sham) and Keilani 2025 (randomised, single-blinded, sham-controlled, n = 20 post-COVID fatigue, 10 sessions, Medical University of Vienna; improvements in 6-minute walk and fatigue scales; sham applicators identical looking but no field) [read: PMC12534222; web per sites.md]. Small pilots. Funding in Keilani: Vienna city grant; no competing interests declared. - Earlier single-subject PAPIMI reports on seizures, wounds and fatigue are cited in Viti refs 40-42 and not retrieved [web per sites.md]. - Cancer and AIDS claims: none supported by controlled data. Pappas is quoted as saying the machine "can cure cancer and AIDS" and "is absolutely safe" [web] Seattle Times 19 Nov 2007, per research/ notes (newspaper's characterisation; he is the interested party). FDA 2005 letter lists 2 rat deaths and a tremor, 2 tachycardias and one fatal myocardial infarction during treatment [web] research/files/seattletimes-2007-11-19-FDA-letter-papimi.pdf (regulator's account; a stake). Seattle Times links deaths to the device; no causal investigation found. c. Replication: pain/HRV effects have been seen by two independent European groups in pilots (n = 20-32), all with the manufacturer's device; sham-controlled in two. The cancer and transmembrane-potential claims have not been tested. Not replicated in the sense of the same endpoints by a large independent trial. d. If true: a non-contact, low-field PEMF would need a mechanism for membrane-potential change that the field calculation in claim 3 argues is absent; the observed effects could be nonspecific (autonomic, sensory, placebo with imperfect sham blinding; pulsed fields of 50-150 mT can be felt as vibration). Verdict: split. Short-term pain/autonomic/fatigue effects: Plausible, with small controlled pilots showing positive results and no large independent trials (PEMF class generally has contested but real clinical literature, [memory] medium). Fracture healing: Tested, not supported (Leisner 2002, one controlled animal study). Cancer/AIDS cures and membrane-potential-restoration mechanism: Never properly tested (no tests of any kind; regulators acted on safety and marketing grounds, not on physics, so not "tested and failed"). "Pasteurisation" by pressure: Never properly tested.

How the claims were handled (rejection and suppression)

• It was not dismissed without a hearing in the patent sense: examiners in the US, EP (granted 2001, no opposition) and others granted it; a grant says nothing about efficacy [read: p.1; web per sites.md].
• US FDA: the 510(k) of 6 Nov 1995 was not cleared because clinical data were lacking; the device was later declared a significant-risk Class III device (2005 letter) and an import alert issued (89-08) [web per sites.md]. This was a regulatory refusal on the grounds of missing clinical evidence, not a refutation. The FDA inspector wrote "the science behind the device is not clear". The sponsor-run study (37 sites, IRB-terminated) was uncontrolled and mishandled per the FDA, which is why it produced no usable efficacy data; the failure of a trial to be done well is not a test. Distributor-run mislabelling (units declared as "plant and seed germinator") is a documented misconduct item by the importers [web per sites.md], not by evidence about physics.
• No sign that any negative physics experiment (a spectrum, an energy balance) was run and then withheld. The inventor also did not publish one. Funding or publication blocking: none documented. I found no evidence of suppression of data.
• The record did not go dark: after the US ban the device continued in Europe under CE marking with university trials from 2023.

Inventor audit (same standard as the critics)

• Unit and numeric errors in the patent: claim 5 "diameter ... 6 square millimeters", claim 6 contradicts the 30 cm coil, claim 2 omits its parent, membrane-air equality [read: deep section 5, 1].
• Unfalsifiable or unquantified statements: "found to be effective to the same degree as microcontact" with no data; "can reach any desired depth by increasing instantaneous power"; "selectively and at will" [read: p.5].
• Claimed an energy-creation result in 1989-1991 and in 1998 ("20% excess heat") and later "Earth unbalance device - free energy generation from zero" (GR20180200149U, 2018) [web per sites.md], yet the 1991 paper is the only one that might hold data; I did not obtain it.
• The oscillograms (FIGS. 2-3) have no vertical scale, so they cannot show any gain.
• The patent's arguments against prior art ("steps have been taken to suppress the burst energy") state a suppression claim about other plasma workers with no citation [read: p.6].

Investigator findings kept in the record (whatever the verdict)

• Family: WO 94/01176, EP 0648140 B1 (granted 2001), DE 69329908, RU 2126278 C1 (lapsed 2011), US maintenance fees paid through 2007 (expired 2013). Predecessors: Cure, Pappas, Eichler US 5,068,039 and US 5,239,442 (remote ion transport, "ideal in applying Nordenstrom's method"); lightning conductor EP 0402552 A1 (1989), the spark-gap/LC design before medical use. Later US 7,151,372 (NMR/EPR reframing).
• Inventor physics: Nuovo Cimento B 76, 189 (1983) on the Ampere force; Pappas and Moyssides Phys. Lett. A 111, 193 (1985); Moyssides and Pappas J. Appl. Phys. 1986, an experimental test of Biot-Savart-Lorentz forces at NTU Athens. These are real peer-reviewed papers by the inventor on the force law (cited 151 times for the first). I could not assess them here; they are not evidence for the medical claims.
• Imitators: Horkel DE 10 2007 030325 A1 (near copy; lists Pappas among original assignees), Pulsartec EP 2062614 A1, Healing Machines WO 98/57704. Others built the same pulsed spark-gap design.
• Supporters: Viti 2023/2025, Keilani 2025 (above); Seattle Times: Wallach and Pappas "both claimed the treatment a success"; Karampatsou memoriam 2026 (collaborator since 1996, interested) reports conference presentations (Salzburg), use in Cuba, Vietnam and Swiss clinics, no studies cited.
• Critics and regulators: FDA letters 2003 and 2005, EIR 2002, California 2005 judgment (Bio-Energy Services, marketing claims; $2,750 + $7,000; reported by Quackwatch, an anti-quackery site), Health Canada 2007 advisory, Leisner 2002.
• Death: reported 11 May 2026 (Facebook posts; memoriam 11-12 Jun 2026); date from one source.

Summary table

Open check items: the 1991 IECEC paper (power balance), the 1998 "excess heat" method, an independent induced-field and spectrum measurement of a PAPIMI discharge, and a large independent sham-controlled trial.