Follow-up Q02: intrinsic and extrinsic infrared detectors in the 1968 Air Force proposal (p.1)

Date: 2026-10-06. Follow-up investigator (STEP F). Web time: about 3 minutes (Chrome; timer 1791342924 to 1791343096). Stopped early: the detector history and the CMB facts needed were found, and two exact-phrase searches for an outside source of the paragraph returned only copies of Moray's own proposal.

The question

In his 7 August 1968 proposal to the Air Force Systems Command ("Proposal for Development Plasma Probe", with Stan Taylor, p.1), Moray contrasts intrinsic and extrinsic semiconductor infrared detectors: extrinsic detectors reach beyond 6 microns but need cooling below 40 K, while a narrow-gap intrinsic detector for longer wavelengths is "not beyond the realm of possibility." Was he hinting at detecting the cosmic microwave background (discovered 1965) or another background radiation, perhaps the "oscillations from the Universe" of p.5? What was the state of intrinsic and extrinsic long-wavelength detectors in 1968, and how does the passage fit the plasma probe and energy converter he proposed?

What the proposal says

Source: _text/gov-Airforce-Materials-Command-Direct-Energy-Conversion-from-Radioactive-Materia-f21d20.txt (13 pages; the PDF is a modern retyped transcription, laid out in a word processor, so typing slips may be the transcriber's or the original's) [read: page image p.1 checked].

• p.1, the passage in full context. It comes straight after the "Direct conversion, Electrical Power Cavity Resonator" is introduced and the remark that "this arrangement might be transposed to become a device for measuring currents." Then: "All objects at temperatures above absolute zero give off radiations and the lower the temperature the longer the wave length. Infrared detectors use either intrinsic or extrinsic semiconductors. Intrinsic detectors use electron transitions within the atoms that make up the semiconductor material itself. The extrinsic type utilizes electron transitions that occur due to the presence of impurity atoms ... While the extrinsic materials permit detection of infrared radiations beyond 6 microns, these materials require temperatures below 40° K. This called for bulky, heavy apparatus and difficulty in designing into multi-element detectors. An intrinsic semiconductor that will detect photons in the longer wave lengths is not beyond the realm of possibility. In an intrinsic detector the more narrow the energy gap between the valence band 40° K, and the conduction band, the easier it is to excite an electron across the gap. Excitation may occur in two ways, by photon excitation and by thermal excitation. The problem is to synthesize a material with a gap arrow enough to respond to long wave lengths (i.e., low energy protons) yet wide enough that reasonable cooling temperatures will be sufficient to reduce thermal excitation." [read: p.1]. Transcription slips, as printed: a stray "40° K" inside "valence band ... conduction band"; "gap arrow" for "gap narrow"; "low energy protons" for "low energy photons".
• p.1, the next sentence, which ties the passage to the proposal: "Experiments at the Research Institute confirm our concept that our semiconductor vacuum and gas pressure efforts are the answer to energy transponders that will be capable of being both energy measuring and counting devices as well as energy converting devices." [read: p.1]
• p.1, the stated objective is an instrument for the "inter-planetary plasma": an "energy analyzer" for ion and electron energy spectra and directions, and a "mass analyzer" for the abundance of elements, in "quiet time" and "dispersed [disturbed] time" [read: p.1]. Nothing in the objective concerns infrared or microwave radiation.
• p.4. The aim is "an energy generator transponder which will be kept alive by the energies of the Universe"; by "a process not unlike the Compton and Cebye [Debye] premise, if the energy from the Universe is a scattered one ... they will become concentrated in secondary radiations, showing not only the same frequency as in the primary source but also a long wave or lower frequency" [read: p.4]. p.5 continues: the scattered quantum's "wave length will be longer and can be made compact by repetition and still lower frequencies obtained" [read: p.5].
• p.5, "oscillations from the Universe". The "pressure transmitters" and "specially constructed valve oscillators" have an "automatic variable relation to the oscillations from the Universe and capable of setting up within their circuits initial oscillations which coincide and will be in resonance with the oscillations of the Universe"; and "it is possible that a device can derive energy from a direct conversion of electricity from radiation environments and from radiations of the Cosmos through synchronized harmonics and resonance" [read: p.5].
• pp.6-7. "Radiation" is defined across the whole spectrum, "from the highest energy (shortest wave lengths) of secondary cosmic rays, gamma rays ..., x-rays, on down through ultraviolet rays, visible light, radio or Hertzian waves, and very long electric rays" [read: pp.6-7]. Infrared and microwaves are not singled out.
• p.10. Van Allen belts, the third (Arnoldy, Hoffman and Winckler) belt, and a UPI item on Atlas nose-cone emulsions recording "44,000 protons of the Van Allen belt" per cosmic ray [read: p.10].
• p.12. "it is, therefore, possible that some 'particle' from the Cosmos with greater penetrating power than Alpha particles, can penetrate blown quartz, semiconductors, or various substances and set up a decided energy reaction. Just as sodium, potassium, cesium, rubidium, barium, strontium, react to visible light, ... might not certain other substances or substance react to oscillations from the Cosmos and artificially produced radiations?" [read: p.12]
• The words "microwave", "background radiation", "3 K", "Penzias" or "Dicke" do not appear anywhere in the extracted text [read: grep of the whole file].

What was searched

What was found

1. No outside source for the paragraph was found. Exact-phrase searches for two of its sentences return only copies of the proposal itself (Scribd) [web Google, see table]. Its textbook tone (the opening "All objects at temperatures above absolute zero give off radiations", the definitions of intrinsic and extrinsic) and the unrelated stray "40° K" read like a summary of a contemporary infrared-detector article. Since Q01 found other pages of Moray's work copied from Le Bon, borrowing is possible here too, but it is not shown. Open.
2. The numbers match the extrinsic detectors of the 1960s, mercury-doped germanium (Ge:Hg) above all. Rogalski's history cites Burstein et al. (1953, impurity photoconductivity in silicon), Borrello and Levinstein, "Preparation and properties of mercury doped germanium", J. Appl. Phys. 33, 2947 (1962), and Soref on doped silicon (1967) [web https://link.springer.com/article/10.2478/s11772-012-0037-7, reference list]. Search snippets: "Ge : Hg detectors were cooled to 30 K to achieve maximum sensitivity" (Rogalski 2005, Rep. Prog. Phys. 68, 2267) [web https://scispace.com/pdf/hgcdte-infrared-detector-material-history-status-and-outlook-timx0lgh3u.pdf, Google snippet only]; "cooling to below 30 K. Ge:Hg detectors were the first to be used in real-time thermal imaging employing linear arrays in the 8-13 μm band" (Capper and Elliott, eds., Infrared Detectors and Emitters, 2001) [web Springer PDF, Google snippet]; a detector table listing "Ge:Cu(PC) 4.2K. Ge:Hg(PC) 28K. HgCdTe(PV) 77K" (Rogalski, Infrared detectors: status and trends, 2003) [web http://antonirogalski.com, Google snippet]; "the first forward looking infrared (FLIR) systems operating in the LWIR spectral window using linear arrays. Ge:Hg with a 0.09 [eV ...]" (Rogalski 2012, ResearchGate) [web Google snippet]; the first FLIR "was introduced in 1969, based on Ge:Hg linear arrays" (Madni thesis, UWA 2017) [web Google snippet]. So in 1968 a detector reaching past 6 µm meant, in practice, an extrinsic germanium detector cooled to about 30 K (Ge:Hg, cutoff about 14 µm) or 4 K (Ge:Cu, longer cutoff) with liquid neon, hydrogen or helium [memory, medium for the coolants]. "Below 40 K", "bulky, heavy apparatus" and "difficulty in designing into multi-element detectors" match this.
3. The "material with a gap narrow enough" was the 1960s narrow-gap intrinsic programme, which was well under way. HgCdTe (mercury cadmium telluride): "Discovery of variable band gap HgCdTe ternary alloy by Lawson and co-workers in 1959 opened a new area in IR detector technology" (Lawson, Nielson, Putley and Young, Royal Radar Establishment, J. Phys. Chem. Solids 9, 325, 1959) [web Springer abstract]. Mixing CdTe (gap about 1.5 eV) with semimetal HgTe (gap zero) gives "any bandgap between 0 and 1.5 eV"; with about 20% cadmium it covers the 8-12 µm window; long-wave HgCdTe "need[s] cooling to temperatures near that of liquid nitrogen (77 K), to reduce noise due to thermally excited current carriers" [web https://en.wikipedia.org/wiki/Mercury_cadmium_telluride]. Milestones in Rogalski's references: Kruse et al. (Honeywell), long-wavelength photoeffects in HgTe-CdTe, Infrared Phys. 2 (1962); Verié and Granger, HgCdTe p-n junctions, C. R. Acad. Sci. 261 (1965); Melngailis and Harman (MIT Lincoln Laboratory), lead-tin chalcogenides (PbSnTe), reviewed 1970; Bartlett et al., "Background limited photoconductive detectors for use in the 8-14 micron atmospheric window", Infrared Phys. 9 (1969) [web Springer reference list]. The proposal's statement of "the problem" (gap narrow enough for long waves, wide enough that "reasonable cooling" suppresses thermal excitation) is the design aim of that work: about 77 K instead of about 30 K. In August 1968 this was a live, largely military, development, not a speculation. Moray and Taylor do not name any of these materials.
4. The cosmic microwave background (CMB) lies far outside the passage's range. Penzias and Wilson measured "a 3.5 K residual background noise" at 7.35 cm (4,080 MHz) with a horn antenna and a helium-cooled maser receiver (published 1965); Dicke, Peebles, Roll and Wilkinson interpreted it as Big Bang radiation [web https://en.wikipedia.org/wiki/Discovery_of_cosmic_microwave_background_radiation]. Worked numbers (Wien's law, λ_peak = 2898 µm·K / T, and photon energy E = 1.24 eV·µm / λ): a room-temperature (300 K) object peaks at about 9.7 µm, photon energy about 0.13 eV; the 2.7 K background peaks at about 1,070 µm (1.07 mm), photon energy about 0.0012 eV. Ge:Hg's impurity level is about 0.09 eV (cutoff about 14 µm). A photon detector for the CMB peak would need a gap about 100 times smaller than Ge:Hg's and operation well below 1 K; 1960s CMB measurements used radio receivers and bolometers, not semiconductor photon detectors [memory, high for the physics; medium for the instruments]. The passage's "beyond 6 microns" and "low energy photons" point at thermal infrared from warm bodies (the 8-14 µm atmospheric window), the military and space-sensing use of the day.
5. Stan Taylor, Ph.D. No document was found (one Google query, no results). Still open; listed in sites.md "Not yet tried" since run 1.

Reasoning: how the passage fits, with competing readings

Known: the paragraph is a correct, compact summary of the 1968 state of infrared photon detectors. Extrinsic germanium detectors reached beyond 6 µm but needed about 30 K (Ge:Hg) or 4 K (Ge:Cu). The aim of the moment was an intrinsic narrow-gap alloy (HgCdTe, PbSnTe) that would work at about 77 K. The proposal names neither the materials nor a source, and nowhere mentions microwaves, the CMB or a 3 K background.

Inferred readings (the record allows more than one): - A. A CMB hint. Little supports it. The wavelengths are wrong by a factor of about 100 (6-14 µm against about 1 mm). The text never mentions the 1965 discovery, though it cites other recent space results (Van Allen belts, p.10). Its "Universe" passages (pp.4-6, 12) are about cosmic rays, Compton-scattered X-rays and "particles from the Cosmos", not a thermal background. The one point of contact: p.4's Compton argument (scattered "energies of the Universe" re-emitted at "still lower frequencies") implies a receiver for long-wave, low-energy radiation, and a narrow-gap absorber is such a receiver. On this reading the IR paragraph supports the idea that a semiconductor can be made to take in very low-energy quanta. Nothing in the text links this to 3 K radiation specifically. - B. Credentials and relevance to the Air Force. The paragraph shows the reviewers that the authors know current detector physics: cooling weight, multi-element arrays, the narrow-gap problem, all live Air Force concerns in 1968. It then hands over, in the next sentence, to the claim that the Institute's "semiconductor vacuum and gas pressure efforts" give "energy transponders" that are at once detectors, counters and converters. On this reading the paragraph is background (possibly summarised from a trade or review article, like the Le Bon pages in Q01). It does not describe something the Institute built. - C. A bridge from detector to converter. The proposal treats detecting and converting as one act: a doped semiconductor that absorbs radiation and gives up charge (pp.1-3: "charge multiplication", donor levels "near the conduction band", cadmium sulphide "internal amplification"). A narrower gap or shallower donor level lets lower-energy quanta free carriers. The passage's own last sentence states the counterweight: a narrow gap is also crossed by thermal excitation, so it must be cooled. How that bears on harvesting low-energy "radiations of the Cosmos" at ordinary temperatures is a question for the reviewers (Step 2); the proposal does not address it.

Fit with the plasma probe: the probe measures charged-particle fluxes (ion and electron energy and mass spectra), which a semiconductor particle detector or an electrostatic analyser does. Infrared detection plays no part in that objective. The paragraph sits between the converter's introduction and the "energy transponders" sentence, so it belongs to the detector-and-converter argument, not to the plasma measurement. For scale: the interplanetary plasma frequency, at about 5 electrons per cm³, is about 20 kHz (f_p ≈ 8.98 kHz × √n, n in cm⁻³) [memory, high]. That is the kind of "oscillation" a plasma probe meets, and it is unrelated to the 3 K background.

Still open: the source (if any) of the paragraph; who Stan Taylor was and whether he wrote this section; the Air Force's reply (none in the folder; NARA record group for AFSC proposals not searched).