Popa-Simil · Source catalogue

Four-page conference paper (NSTI-Nanotech 2007, Vol. 4, pp. 614-617, ISBN 1420063766) [read: p.1 footer; deep read]. It proposes a repeating conductor-insulator-conductor-insulator ('CIci') nano-hetero-structure that converts the kinetic energy of fission products directly into electricity. The mechanism as described: a fast ion crossing the stack makes electron showers of different intensity in adjacent conductors; electrons from the high-yield conductor cross a nanometre insulator and are absorbed in a low-yield collector conductor, so the conductors polarise [read: p.1, p.3]. The abstract says efficiency 'may be higher than 80%' and could be improved by quantum effects, and that the reactor 'will run cold' because less energy is left as heat [read: p.1]. The introduction reviews earlier beta-batteries and a 1997 charged-particle battery with plate pairs, cell potentials 3-10 V [read: p.1-2]. Section 2 gives a stopping-power formula (Eq.1) and an illustrative eleven-layer stack (U 100, S 30, Al 30, Li 40, Pb 30, SiO 40, Mg 30, I 100, CF2 100, Th 100, Au 100 nm) simulated with SRIM-2003 for 100 MeV Cs (Figs. 1-2); it quotes 10 keV/nm energy release and '~1200 electrons per nm3' [read: p.2]. Module voltages are stated as mV, under 10 mV, limited by insulator breakdown; a multi-level DC/AC and AC/AC converter cascade (possibly Josephson and Andreev effects) would raise this to grid level [read: p.3]. Three fission-harvesting geometries are described (fissile material outside, mixed in, or none). Voxel size: 'in the range of mm3 to cm3' with normal conductors, 'while the usage of superconductors may drive this volume up to dm3' [read: p.3] (an earlier digest wrongly gave 1-1000 mm3 and 1000 dm3). Section 4 discusses actinide superconductors PuCoGa5 (Tc ~18.5 K) and 'PuCoRh5' as printed (Tc ~8.7 K; the Tc and the Rh discussion match PuRhGa5, so the printed name is probably a slip, not confirmed against a web source), absorption cross-sections (Table 1), and radiation damage (Frenkel defects from 5 MeV alpha recoil; dynamic equilibrium of damage and repair) [read: p.3-4]. The conclusion foresees cryogenic ultra-compact reactors, space panels, fusion harvesting and self-healing structures, and raising fuel power density from ~0.2-0.5 kW/cm3 to a theoretical 5000 kW/cm3 [read: p.4]. No experimental data from the proposed structure are presented; the only results shown are the SRIM simulation. The only experiment cited is earlier recoil-implantation work (Muntele et al. 1999, ref. 11) [read: deep read p.2, p.4]. The authors have a stake in the concept (they are its proponents, at a company, LAVM LLC); this is their own characterization. Deep-read findings on p.2 (the authors' mathematics and simulation) [read: page image p.2]: (1) Eq.1 follows the Bethe-Bloch structure, but as printed r0 is called 'Bohr's electron radius' (the formula needs the classical electron radius), the log argument '2mc b^2/(1-b^2)' lacks a factor c (dimensionally inconsistent), the Barkas term has the opposite sign from the standard form, and the Bloch term uses Z2^2 where the standard form uses the projectile charge Z1^2; atom density and effective charge are not introduced. The reader's standard form is [memory, high confidence]. With bare Z1=55 Eq.1 overestimates the stopping of 100 MeV Cs by about 10x relative to SRIM (computed in the deep read). (2) The stated layer thicknesses sum to 700 nm but Figs. 1-2 run to 600 nm, and the thin layers are drawn about 10 nm thinner than printed. (3) In the SRIM stack the Cs ion loses only about 9 MeV of 100 MeV (about 9%); it is not stopped. (4) 10 keV/nm is the right order of magnitude (Fig.2 plateaus 7-26 keV/nm; U and Au about 25). (5) '<30 atoms and ~1200 electrons per nm3' matches solid iodine or UO2 molecules, but not U or Au (~4400-4700 electrons/nm3) and not most other layers. (6) Number of ions, angle, mode and densities of the SRIM run are not stated. These are checks of the paper's own numbers and are not a verdict on the concept. Deep-read findings on pp.3-4 (added after QE) [read: p.3-4]: (1) The author's own rationale for 'run cold' (Sec. 3.2): 'The analysis of the heating process in nuclear fuel driven to the conclusion that the fast moving fission products are inducing in the insulator Uranium Dioxide ceramics electron showers of few micron range. The moving electrons interact with the lattice heating it. To prevent heating it is necessary to cut the electron showers circuits and drive them through conductive materials in circuits where the total resistance to be significantly smaller.' The voxel volume is then 'given by the conductivities of the two conductive layers', which must give an internal resistance 'much smaller than the previous one of the ceramics'. No calculation, citation or data is given for this on these pages. It sits beside p.2's 'several nm' electron free path, which is a tension for the reviewers to weigh (shower range 'few micron' versus several nm). (2) The 'energetic balance' sentence, quoted as printed: 'Another condition difficult to be achieved is the energetic balance, which simply means that the electric harvested energy to be smaller than the energy consumed to remove the thermal effects inside, given by the particles direct heating and electric current heating effects.' As printed this describes a net energy loss; it is probably a slip for 'larger' (the deep reader's reading, not the author's statement). The cryogenic cooling load is not quantified anywhere in the paper. (3) Table 1 (neutron absorption cross-sections in barns of PuCoGa5 for Pu-238/239/240/242): Pu 558 / 1017.3 / 289.6 / 18.5; Co 37.18; Ga 13.75; 'Amount' 611.68 / 1068.23 / 340.53 / 69.43; 'Absorption decrease' 91% / 95% / 85% / 27%. Arithmetic check: three of four columns sum correctly; the Pu-238 column sums to 608.93, not 611.68 (+2.75 b). The 'absorption decrease' row equals the Pu share of the printed total (558/611.68 = 91%; 1017.3/1068.23 = 95%; 289.6/340.53 = 85%; 18.5/69.43 = 27%), so in this compound Pu itself dominates absorption. The Ga entry 13.75 b = 5 x 2.75 b, whereas p.4 says Ga has 'a total absorption cross-section of only 2.5 barns'. The text says Ga absorbs 'over ten times' more than O and C; the printed paper gives no O or C values. (4) Three fission-harvesting geometries: (a) fissile material outside the structure, surrounded by a harvesting layer 'thick as the range of the fission products', fissile-to-structure ratio set by criticality; (b) fissionable material mixed into the structure, mainly in the high-density conductor; (c) no fissile material, tiles that stop external radiation 'like alpha particles from fission' (the type closest to existing alpha batteries). (5) The p.3 fabrication route is binary nanoparticle superlattice self-assembly (Talapin, ref 13); the Josephson relation K_J = 483 597.879 GHz/V and an 'Andrew [Andreev] effect' are named with no efficiency numbers. (6) Prior art cited on p.1: Willson 1955 beta capacitor (US2728867), Polansky NERI 1999 ('recently retested'), DARPA LEAPS (Tsang 2004), Linder US2517120, Young WO9748105A1 / US5861701, Ritter US4178524, Rappaport US3094634, Olsen US3706893. (7) Conclusion p.4: 'The usage of the Actinides superconductors opens the way to the cryogenic nuclear reactors acting like ultra compact batteries with huge powers based on fission.' The 0.2-0.5 to 5000 kW/cm3 figure has no derivation in the paper. Supporting documents deep-read after QE (Investigator-gathered, not in the manifest; see the supplementary entries below). These add context but do not change what this paper itself contains: no measurement of any CIci stack appears in the paper, the 2010 and 2013 patent applications, or the 2012 roadmap paper.

Key claims

• A repeating conductor-insulator-conductor (CIci) nano-structure may convert the kinetic energy of fission products directly into electricity with efficiency higher than 80% (the abstract's wording is 'may be higher than 80%').
• The conversion works because adjacent conductors give different electron-shower intensities, so electrons cross a nanometre insulator and charge the low-yield conductor.
• A reactor whose fission energy is harvested as electricity would run cold, because fuel heating (electron showers of few-micron range in resistive UO2) is removed by giving the shower currents a much lower-resistance path through conductors.
• Each CIci module gives only millivolts (below 10 mV, limited by insulator breakdown), which a multi-level DC/AC and AC/AC converter cascade can raise to grid voltage.
• Superconducting actinide compounds such as PuCoGa5 (Tc ~18.5 K) can be used in the structure in cryogenic nuclear reactors, and a dynamic balance of radiation damage and repair can keep them working for the reactor lifetime.
• Nano-hetero-structure harvesting could raise nuclear fuel power density from about 0.2-0.5 kW/cm3 to a theoretical 5000 kW/cm3.
• Electrons produced by a 100 MeV fission-fragment ion have a free path of several nm and can cross nanometre insulators ballistically.
• The harvested electric energy must exceed the energy spent removing heat (the paper's 'energetic balance' condition, printed as 'smaller').

Relevance: The project's only manifest document, and the source of all central claims. It is a concept paper with a simulation, not an experimental report. Pages 3-4 carry claims 3-6, so the pp.3-4 deep read matters.

Mainstream conflict: Direct conversion of charged-particle energy (betavoltaics, secondary-emission batteries, thermionic and direct-collection concepts) is established in principle for low power. Not established: the 80% efficiency, the cold-running reactor and the 5000 kW/cm3 figure for fission fragments, and these are not supported by data in the paper. Judging them is for the reviewers.

Supplementary entry: Investigator-gathered file, not part of the MEGA source folder or the manifest. Efficiency theory for gamma-pumped layered converters (heavy-metal emitter / insulator / light-metal collector), from the same Russian group that cites Popa-Simil's patent among its theory references. Eq.(4): I = e*N = e*gamma*hbar*omega/(K+W). Eq.(5): eta = W_ext/W_abs = I^2 R/(gamma hbar omega). Eq.(6): K >= e*phi = e*I*R. Eq.(7): K = -W/2 + sqrt(W^2/4 + e^2 gamma hbar omega R). Eq.(8): K = e*sqrt(gamma hbar omega R), beyond which efficiency falls (text says as 1/K; by eq.5 it falls as 1/K^2). Ideal case: when electron kinetic energy equals the electrostatic energy across emitter-collector, 'the electrons enter the collector with zero kinetic energy, thus eliminating heat release losses' and eta approaches unity (eta = K/(K+W) at equality, deep reader's algebra). Eq.(4) assumes all absorbed energy goes into emitted electrons. Reported experiment: 'the experimental data reported in the patent [16] (Yoshida) have shown an efficiency factor of ~1e-5 %'; the same chapter gives 'less than 0.01 %' on pp.265/267, a factor of 1000 apart, and the Yoshida patent was not checked. Four loss causes: planar geometry, thick emitter layers (tens of micrometres), thick insulator (backflow), electron energy mismatch; aerogel or vacuum proposed as insulator. Efficiency of 25% is attributed to 'plane layered structures with U235 [13, 14]' (Anufrienko et al. 2006, 2008) in one sentence, with no radiation type, definition, load or voltage; the chapter variously calls refs [13,14] experimental (pp.265, 281-282) and theoretical (p.267), so measured versus estimated is unclear. Refs [49-51] (the authors' own, gamma): insulator reduced from 200 um to 1 um raised efficiency 'by 3 orders'; the starting efficiency is not given. Claimed design output: at least 20 kW electric from 100 kW through a fuel assembly, with no experimental support in the chapter. The chapter does not compute the self-bias for fission-fragment delta electrons. Deep reader's application (not the authors'): T_max about 2.3 keV (light fragment) and 1.05 keV (heavy), so the ideal condition needs about 0.1-2 kV across a nm insulator, 10-1000 times the roughly 10 MV/cm breakdown of SiO2 [memory, high]; with a few volts of bias the same equations give eta about e*phi/K, a few percent or less. The authors name the thin-for-transport versus thick-for-breakdown conflict qualitatively and give no numbers.

Key claims

• Efficiency of a radiation-to-electricity layered converter approaches 100% when emitted electrons spend all their kinetic energy climbing the emitter-collector potential (K = e*phi) and K >> W.
• Planar layered structures with U-235 reached 25% conversion (refs 13, 14; one-sentence secondary citation).
• Reducing the insulator from 200 um to 1 um raised efficiency by three orders of magnitude (gamma pumping, refs 49-51).
• Yoshida's patent device reached about 1e-5 % (or 0.01 %; the chapter gives both).

Relevance: The strongest supporting theory in the record for the energy-bound the reviews apply to CIci, and a statement by a supporting group of the thin-insulator versus voltage trade-off. The group is an interested party for its own converters.

Mainstream conflict: Direct conversion is established in principle; efficiencies far above the reported 1e-5 to 1e-2 % for layered gamma converters are not demonstrated in this chapter.

Supplementary entry: Investigator-gathered file, not part of the MEGA source folder or the manifest. Patent application, priority provisional 11/603,812 (21 Nov 2006). No measured result is reported: numbers are SRIM / e-Casino simulations or hand estimates, and the text says 'in present there is no molecular dynamics, particle interaction software to treat completely this case' (para 0084). The 84% (para 0121) is 'the efficiency of ionization energy stopped in the C layer' and an 'upper conversion efficiency estimator', and 'has to be obtained by experimental buildup'. Para 0074: layer ratio 10:1 'pushes the conversion possible limit up to 90% that may be also reduced by the effect of the insulators, down to 80%'. Para 0096: magnetic delta layers 'increase the conversion efficiency towards 99%' (no calculation). Para 0116 gives the electrical alternatives for one fission per second: 167 MeV = 26.7 pW 'harvested at 10 mV and 2.67 nA or at 1 V and 26.7 pA'; both equal full 167 MeV, an implicit 100% conversion. FIG. 10 weights 5 + 0.5 + 1 + 0.5 = about 7 units; 5/7 is 71%, not 84% (5/6 = 83%); the Pu-238 battery estimate 20 mW from 1 cm2 x 50 um is labelled a maximal value. Para 0125 and 0132: minimum structure 20-40 microns for alpha and fission products. Para 0183: voltage is a fraction (say 50%) of the insulator breakdown; series structures add voltage, parallel add current. Para 0187: a 1 micron battery 'having a 5% efficiency'. FIG. 15 is a proposed ion-beam test setup in the present tense, with no data. e-Casino run: 'a very poor statistics of about 25 cases' (para 0102).

Key claims

• The 84% figure is an upper bound on the fraction of ionization energy stopped in the C layer, not a measured electrical efficiency.
• The structure can be run at 10 mV and 2.67 nA or at 1 V and 26.7 pA for one fission per second (167 MeV = 26.7 pW).
• The minimum useful structure for alpha and fission products is 20-40 microns.

Relevance: The inventor's own more detailed statement of the CIci concept, and the place where the paper's efficiency claim is qualified.

Mainstream conflict: Same as the paper: no measured conversion efficiency; 90-99% figures are unsupported estimates.

Supplementary entry: Investigator-gathered file, not part of the MEGA source folder or the manifest. Application 13/297,243, filed 15 Nov 2011, no priority claimed. No measurement of a CIci or super-capacitor structure and no measurement of the D+Pd reaction is reported; the text has SRIM / MCNPX / e-Casino simulations, an 'upper limit' arithmetic estimate of 84% (para 0203), and test circuits and chambers (FIG. 20, FIG. 22A/B) described as procedures in the future tense. The control mechanism ('nuclear entanglement and quantum non-localities') is a verbal, non-mathematical model: a nucleus oscillating in its lattice site emits an 'entanglement or tunneling wave', a decaying oscillatory 'soliton'; overlap of the waves of neighbouring nuclei (e.g. two D and one Pd) puts them 'one into another' in 'sub-space', allowing ternary non-local fusion without the Coulomb barrier, with the rate set by lattice eigen-modes, temperature, field, hydrogen pressure and nanocluster size (claim 1).

Key claims

• The rate of a D+Pd ternary 'non-local fusion' can be controlled through lattice vibration modes, temperature, field, hydrogen partial pressure and nanocluster size.
• A CIci or super-capacitor converter can harvest nuclear reaction energy; the 84% is an upper limit.

Relevance: Shows the inventor's later extension of the CIci converter to a fusion (LENR-style) battery; the converter claims are still simulation only.

Mainstream conflict: The entanglement / non-locality mechanism has no mathematical form in the text and is outside mainstream nuclear physics; the Pd-D low-energy reaction is not an accepted result.

Supplementary entry: Investigator-gathered file, not part of the MEGA source folder or the manifest. No CIci measurement, prototype or beam test is reported in the 81 pages; Fig. 8 and slides 50-51 are simulation plots and Fig. 13 / slide 54 is a proposed Stage 1 setup. The phrase 'collateral experiments' is not tied to direct conversion in this paper: 'Expected collateral research results and applications' (p.15) lists hoped-for spin-offs, not experiments. Cited background: Moseley 1913; the Romanian fission electric cells (Ursu & Popa-Simil 1980; Ursu, Popa-Simil, Badescu-Singureanu, Vamanu 1978; Ursu, Badescu-Singureanu, Schachter 1984; cited by reference, no numbers); slide 46-47 MIT work of Gang Chen on thermophotovoltaics (not nuclear and not CIci). Slide 48: DARPA 2004 'Liquid Electronics Power Source ... Result = Negative; Issues - low efficiency - low stability', a liquid-electronics effect 'accidentally observed by my team in 1987', no data. Slides 69-70 (author's history): 'I have never published because of lack of coherent data'; 1984-1995 worked in a direct nuclear energy conversion special program (coated capacitor structures, solid structures Se, Al, Au, U); 2011 patent application for a fusion battery (US 13/297,243). Projected figures: efficiency 'as high as 95%', power density up to 1 MW/cc (p.3, p.16, slide 23), 'first reliable products may be delivered in less than 10 years', all stated as projections with funding caveats (p.1, p.12).

Key claims

• The fusion battery with a CIci converter could reach up to 95% conversion efficiency and up to 1 MW/cc.
• The author worked on direct nuclear energy conversion from 1984 to 1995 (slide 70, not documented in the PDF) and observed a 'liquid electronics' effect in 1987.

Relevance: Tells the reader that, as of 2012, the inventor himself reports no CIci measurement and says he never published because of lack of coherent data. The unidentified 'collateral experiments' of the 2011 Physics Procedia abstract are not identified here.

Mainstream conflict: Pd-D 'fusion battery' and 'new physics' sections rest on cited LENR-type results outside mainstream acceptance; not judged here.

Supplementary entry: Investigator-gathered file, not part of the MEGA source folder or the manifest. Journal Nuclear Energy and Technology 4(3):163-166. Measured, not simulated: 100 F carbon supercapacitors with organic electrolyte (3.5 g carbon, 1800 m2/g) in a neutron field (absorbed dose rate up to 0.1 Gy/s, gamma up to 0.03 Gy/s, 50 Gy total; source type and energy not stated). Symmetric cell: about 1.24 mV (Fig. 1 last point about 1.33 mV). Asymmetric cell: abstract says up to 1.15 V, charge about 5e-3 C (computed from a literature surface-charge density of 1e-8 C/(Gy m2), not measured directly); Fig. 2 shows a plateau near 0.1 V during irradiation and Fig. 3 shows about 1.1 V a day later decaying to 0.35 V in about 125 min, so the 1.15 V is a post-irradiation value. No efficiency, power or energy output is defined or reported. No unirradiated control run is described. Authors include the developers of the concept (Chernov, Stepanov) and the capacitor maker's staff. The paper's introduction summarises earlier Anufrienko / Chernov metal-dielectric-metal (MDM) converter work as giving 'significant efficiency coefficients' with no number; those earlier papers (refs 13, 14 of Klassen) were not read. No connection to Popa-Simil is stated.

Key claims

• Neutron irradiation of asymmetric supercapacitors to 50 Gy gives up to 1.15 V and about 5e-3 C of separated charge.
• The voltage depends on the absorbed dose and on plate asymmetry, not on neutron flux density.

Relevance: The only measured data in the folder from a related (Russian MDM) line, and the lead for the Klassen 25% citation; it measures charge separation, not conversion efficiency.

Mainstream conflict: Unexplained long-time relaxation; the authors say the physical picture in supercapacitors without irradiation is not completely clear.