Haisch-Bernard · Source catalogue

Three-page theoretical proposal in Physical Review B 30(4), 1700-1702 (received 23 Nov 1983, revised 16 Apr 1984), funded by Air Force Rocket Propulsion Laboratory contract F04611-83-C-0013 and Hughes Aircraft Company. Forward restates Casimir's 1948 result (energy per area u = -pi*h*c/(1440 a^3), pressure P = pi*h*c/(480 a^4), with h, not hbar; the coefficients are correct for h) and the Lifshitz dielectric generalisation, notes the retarded 1/a^4 law fails below a_min = lambda_min/2pi where a nonretarded van der Waals/London law takes over, and reviews experiments: dielectrics (mica crossed cylinders, Israelachvili and Tabor) agree with Lifshitz theory from 30 to 20 nm and reach over 10 kN/m^2 at 1.4 nm, but conductor experiments (Arnold et al., silicon) held 1/a^4 only down to 300 nm and then deviated, probably from surface roughness; there were no good metal-plate data at 1-20 nm. He says it is 'highly probable' but not rigorously proven that the vacuum-fluctuation field is conservative, and uses a hydroelectric-dam analogy: a conservative field can still be used to convert stored potential energy into work. The 'vacuum-fluctuation battery' is a stack of ultrathin aluminium foil leaves separated by a few micrometres, each held apart by same-polarity electrostatic charge from an active bidirectional power supply with position sensors and feedback (he notes the suspension is unstable). Lowering the electrostatic repulsion to slightly below the Casimir force lets the leaves collapse while the supply extracts electrical energy; the battery can be 'recharged' by pushing leaves apart at the cost of supplied energy. A spiral ('Slinky') single-conductor variant is also proposed. The paper has no experiment and no numerical energy, power or efficiency estimate; a deep-read table of ideal-conductor values (e.g. about 1.3e5 Pa at 10 nm) is the reader's own calculation, not in the paper. Minor printing slip: Eq. (2) lacks a minus sign relative to Eq. (1) and (3); the OCR's '980' in Eq. (3) is 480 on the page image.

Key claims

• Two conducting plates closer than about 1 micrometre attract with the Casimir pressure P = pi*h*c/(480 a^4) arising from vacuum zero-point fluctuations.
• If the electrostatic repulsion between charged foil leaves is held slightly below the Casimir force, the leaves collapse together and the power supply extracts electrical energy from the vacuum-fluctuation field.
• The battery can be recharged by making the electrostatic force slightly larger than the Casimir force, so the process is storage and conversion of energy put into the foliated state, not a perpetual source.
• The vacuum-fluctuation field is highly probably conservative (otherwise infinite energy could be extracted), and a conservative field can still yield energy as a hydroelectric dam does with gravity.
• No good experiments exist on metal plates at 1-20 nm; if they showed the very strong predicted forces, these forces may be a new source of energy.

Relevance: Cited by the Haisch-Moddel patent as the earlier thought experiment for extracting energy from the Casimir force; it is the conservative-field, battery-style counterpart to the patent's continuous-cycle claims. The paper is a published theoretical proposal by an Air Force-funded researcher, with no experiment reported.

Mainstream conflict: None in the Casimir force itself (established and measured). The proposal is presented by Forward as a one-shot, conservative-field energy conversion. Practicality at 1-20 nm for metals was unmeasured in the paper (later measurements not covered by this document).

One-page abstract by the patent's inventors. It proposes to 'indirectly tap a new source of energy, potentially greater than nuclear', zero-point energy, which it says physicists consider not useful because it cannot be directly accessed. It cites zero-point energy preventing helium from solidifying at absolute zero and detection as circuit noise. The proposal: cycle atoms, preferably monatomic noble gases, under modest pressure through Casimir cavities. Suppressing vacuum-field modes inside the cavities would induce an orbital energy shift larger than, but analogous to, the Lamb shift, termed the 'Casimir-Lamb shift'; this should emit photons (Larmor radiation) on entry. On exit the shift would reverse by absorption from the ambient quantum vacuum field, 'without violating the second law of thermodynamics'. The abstract says recent experiments show Lamb shifts can be controlled and enhanced. A figure depicts the radiant energy being absorbed and extracted. No numbers, calculations or experiments are given.

Key claims

• Atoms moving under modest pressure through Casimir cavities undergo a Casimir-Lamb shift of their electron orbital energies, emitting photons on entry.
• On leaving the cavity the shift is reversed by absorption from the ambient zero-point field, so net energy is drawn indirectly from the quantum vacuum.
• The cycle does not violate the second law of thermodynamics.
• Zero-point energy is a source potentially greater than nuclear energy.

Relevance: The inventors' own short statement of the patent's mechanism, in their own words and with their own stake as patent inventors.

Mainstream conflict: The usual view is that zero-point energy cannot be extracted as net work from the ground state; the abstract says the physics community regards it as not directly accessible and claims an indirect route. The 'Casimir-Lamb shift' is the inventors' own term, as the abstract itself indicates.

The patent describes a system to convert electromagnetic quantum vacuum energy into heat, electricity or mechanical power. Its theory rests on stochastic electrodynamics (SED): the idea of Boyer, Puthoff, and Cole and Zou that an atomic electron's orbit is a balance between Larmor radiation (classical radiation by an accelerating charge) and absorption from the zero-point field, whose spectrum rises as frequency cubed. For hydrogen the patent gives the Bohr radius 0.529 angstrom, orbital speed c/137, and a sustaining zero-point wavelength of about 455 angstrom (0.0455 micron), with orbit frequency 6.6e15 per second. Suppressing modes at and below that wavelength in a Casimir cavity (a space between close conducting surfaces, d about 250 angstrom or larger for hydrogen, about 0.1 micron for noble gases) should make the electron settle to a lower energy state releasing an assumed 1-10 eV per atom (binding energy 13.6 eV); on exiting, the atom reabsorbs energy from the ambient field. Energy is 'extracted locally, replenished globally'. Monatomic noble gases (He, Ne, Ar, Kr, Xe) are preferred. The patent states the theory is not proven for atoms other than hydrogen. Embodiments: (1) Casimir channels, 10x10 cm plates with 5000 conducting strips, gas at 10 cm/s, 1.3e20 transitions/s giving 21-210 W (210-2100 W for a stack); (2) Casimir tunnels, a 1 cm^3 block with about 1.3 billion 0.1-micron tunnels, 500 interactions per atom, 2150-21500 W; (3) bonded sapphire wafers, 1 sq. cm, 200 nm SiO2 spacers, 40 nm platinum stripes 2 microns wide at 4 micron pitch, estimated 1-10 W at 8 atmospheres; (4) flexible photonic-crystal polymer film with stretching modulation (FIGS. 6A-6B); (5) MEMS (micro-electro-mechanical systems) pivoting plate modulating cavity spacing (FIGS. 7A-7B); (6) asymmetric entry/exit with an absorbing layer (FIG. 8); see the QE additions below. Emitted infrared is to be captured by a 1 mm water bath (absorption coefficient stated above 10 per cm over 2-200 microns). A pumping-power check uses Roy et al. FIG. 10(a): 760 torr per 5 mol/m^2-s through 60 microns; it concludes 'a pressure of 1700 Pa, corresponding to 17 atmospheres' and 1.7 mW pumping power versus 2.2-22 kW release. The deep read notes the 1700 Pa and 17 atm disagree by about 1000x; 17 atm would imply about 1.7 W rather than 1.7 mW; the check compares with the cube estimate rather than the bonded-wafer device; the Roy citation is '1993' in the body but 2003 in the references; 0.25 porosity is not used. No measurement of output is reported anywhere in the patent. Claims 1-20 cover a system with a fluid supply, a Casimir cavity in a fluid path, and means to collect released radiation; claim 15 specifies the electromagnetic quantum vacuum. References include Boyer 1975, Cole and Puthoff 1993, Cole and Zou 2003-2004, Forward 1984, Puthoff 1987, Milonni 1994. Related patents named as different: Shoulders 5,018,180; Mead and Nachamkin 5,590,031; Pinto 6,477,028, 6,593,566, 6,665,167. The assignee and inventors have a commercial stake; all output figures are the applicants' estimates, and a US patent grant is not a test of operation. QE additions (deep reads, page images). (a) Cutoff convention: the patent says modes 'whose wavelengths are longer than the separation of the plates' are excluded and that 'Mode suppression will be strongest for wavelengths of d or greater' with partial suppression at the 'stairway' d/n [read: p.10-11]; its cutoff is therefore lambda = d, whereas the textbook ideal-plate cutoff is 2d [memory, high]. For hydrogen the patent pairs the 455 A wavelength with 'a Casimir cavity of d=250 A or thereabouts (and perhaps even a larger cavity)', and says the electron's probability tail may let cavities of 0.1-0.2 micron have some effect; 0.1 micron is the figure used for the noble gases (via r^(3/2) scaling, 'a crude assumption'). The 1-10 eV release is 'we assume' [read: p.11]. (b) Maclay: 'According to a Jordan Maclay, who has done theoretical Casimir cavity calculations, a long cylindrical cavity results in an inward force on the cavity... this implies that a cylindrical cavity of diameter 0.1 micron would yield the desired decay of outer shell electrons' [read: p.11]. This is the applicants' characterisation; Maclay is not in the reference list and no publication is cited. (c) Extensions: 'An analogous process is also believed to underlie molecular bonds' (not proven; the patent says the all-atoms balance 'has not yet been proven theoretically'), and liquid working fluids (ethylene glycol to about 100 C, sodium) [read: p.12, 17-19]. The applicants also state the process is 'slow' as experienced by the electron, so heat rather than a sudden optical signature is expected [read: p.11]. (d) Related patents (Shoulders 5,018,180; Mead and Nachamkin 5,590,031; Pinto 6,477,028, 6,593,566, 6,665,167) are each dismissed on one ground, 'does not deal with energy release from atoms in Casimir cavities'; this is the applicants' characterisation of rivals' patents [read: p.12]. (e) Later embodiments, no measurement or power figure for any: FIGS. 6A-6B flexible photonic-crystal polymer film with silicon pillars, gas gaps e.g. 200 nm, stretched by a piezoelectric vibrator; FIGS. 7A-7B MEMS torsion plate rocked electrostatically between two electrodes, citing Chan et al. (Bell Labs, 2001) and stating 'we are not making use of the Casimir force to change this spacing'; FIG. 8 'Assymetric Casimir Cavity Entry and Exits Including Absorbing Means'. In FIG. 8 the gas flows past an absorbing glass layer (84) just upstream of the conducting Casimir region (86), on sapphire substrates, with no absorber downstream; the absorber is placed within roughly one emission wavelength of the emitting atoms; 'The conducting layer length is chosen so that the emerging atoms do not have substantial access to radiation emitted from the absorbing regions' [read: p.18-19]. No length, wavelength or output is given; dimensions are 'approximately the same' as FIGS. 4A-4D; the text misnumbers the absorber as 87 (should be 84). The applicants themselves write that if release and re-supply are both local, or both global, 'no net energy is provided', and that 'we provide an asymmetry in the apparatus' to avoid this [read: p.18]. In FIGS. 6-7 nothing asymmetric is described. (f) Claims (word for word in the deep read): independent claims 1, 15, 16, 18, 19, 20 require release of energy on entry into a Casimir cavity, renewed intake on exit and capture of the released energy; none contains 'net', 'gain' or a power figure, and none recites the FIG. 8 asymmetry. Claim 3-4: cavity moves relative to the fluid. Claim 20 covers a 'mechanism' changing atoms and molecules in configuration, not only Casimir cavities. (g) Reference list has 12 entries (Boyer 1975; Cole and Puthoff 1993; Cole and Zou 2003, 2004; Forward 1984; Goho 2004a,b; Milonni 1994; Park and Summers 2002; Plossl and Krauter 1999; Puthoff 1987; Roy et al. J. Appl. Phys. 93, 4870, 2003); Roy is dated 2003 in the list and '1993' in the body (typographical error). Priority: provisional application 2004-09-27 is noted by the Investigator [web], not read in this document.

Key claims

• An atom entering a Casimir cavity loses orbital energy (about 1-10 eV per transit) because suppressing zero-point modes upsets the Larmor emission versus absorption balance that holds the electron in orbit.
• On exiting the cavity the atom is re-energised from the ambient quantum vacuum field, so a gas cycled through many cavities delivers net usable energy drawn from the vacuum without violating thermodynamics.
• A Casimir block of 1 cm^3 with about 1.3 billion 0.1-micron tunnels and gas flow of 1 cm^3/s would release 2150 to 21500 W.
• A 1 sq. cm bonded-wafer device at 8 atmospheres input pressure would produce an estimated 1 to 10 W.
• Pumping power for the gas flow is about 1.7 mW, far less than the power released, so net power is produced.
• Hydrogen's ground state in SED is a Larmor emission and zero-point absorption balance at the Bohr radius, and the same balance underlies all atoms.
• In an asymmetric channel (FIG. 8) with an absorbing glass layer upstream of the conducting Casimir region and none downstream, the excess energy is delivered locally on entry and the energy deficit is supplied globally on exit, so net energy is obtained; the conducting layer length is chosen so emerging atoms lack substantial access to radiation from the absorber (no length given).
• Hydrogen in a Casimir cavity of d = 250 A or thereabouts would release about 1-10 eV (an assumption), and the stronger suppression is for wavelengths of d or greater (the patent's own cutoff convention is lambda = d).
• A cylindrical cavity of 0.1 micron diameter would yield the desired outer-shell electron decay, according to the applicants' account of Jordan Maclay's Casimir cavity calculations (no citation given).
• The same energy-release cycle is believed to apply to molecular bonds and to liquid working fluids (ethylene glycol, sodium), and can be driven by moving the walls (stretched photonic-crystal polymer, MEMS rocking plate) instead of moving the gas.

Relevance: The central document: the full statement of the inventors' device, theory, numerical estimates and claims. It is the primary text for the project's claims.

Mainstream conflict: The patent's mechanism rests on stochastic electrodynamics (a minority classical-field formulation) rather than standard quantum mechanics, in which the ground state has no lower state to decay to; continuous net power from a ground-state atom cycle conflicts with standard thermodynamics. The patent itself asserts that the second law is not violated. The reviewers judge this, not the fact-finder. Internal arithmetic inconsistencies noted in the summary. Further: the patent's cutoff convention (lambda = d) is a factor 2 looser than the textbook ideal-plate cutoff (2d) [memory, high]. The moving-wall embodiments must pay the work of re-raising the atomic level against the atom-wall force; the deep reader's estimate (reader analysis, not in the patent) is that 1-10 eV per atom would imply an extra force of tens to hundreds of atmospheres on a 200 nm gas layer. The reviewers judge this.