No abstract

An improved technique is described for the growth of InP by liquid-phase epitaxy. The use of gold-plated reflector tubes in the furnace design has improved the control of layer quality, and background doping levels have been consistently reduced to 3 X 10ZS/cm 3 by replacing polycrystalline InP by PH3(g) as a source of phosphorus. The behavior of the dopants S*~, Ge, S,i,, Te, Zn, Cd, and Bi in InP are discussed and the distribution coefficients "k' are found to be ksa --0.0019, kGe -"-0.005, ksi -"-4, kwe -" 0.27, kzn = 1.14, kcd _~ 0.002, km--0.002 to 0.0002The dopants Sn and Zn were found to be the most suitable for n and p doping for LED fabrication. The Group IV elements yielded n-type material only, although the mobility measurements indicated possible amphoteric behavior for Ge. The Hall mobilities at room temperature Ior the n-type samples show good agreement with theory.z Present address:

Oneofthe limiting difficulties facing infrared systems seeking to detecttargets isbackground clutter. In many interesting situations viewing dim targets, detection ispossibleon thebasisofsystem sensitivity,butpreventedbylow signal-to-clutterratios.Overthepastseveral years, we have investigatedthepotentialofinfraredpolarizedsignaturecomponents for suppressing background clutter and enhancing the detection of dim targets. In the course of this work, we have built an imaging infrared polarimeter. participated in field experiments, and compared measured results with theoretical predictions.This paper seeks to summarize this work. It includes a brief theoretical description of the effect, a description of our measurement instrument, samples of our empirical findings, and comparisons with simple theoretical predictions. We conclude with a summary of what we have learned and what issues remain unresolved. BACKGROUNDIn military applications ofinfrared technology (e.g., thermal imagers, spectrometers, seekers), polarization has largely been ignored. Although there have been priorattempts to exploitlRpolarization in the 1960's and 70's [e.g., 1-4], the efforts seem to have been relatively short lived.Recently, there has been renewed community interest [e.g., 5-8] in JR polarization. Beginning in 1985, we have studied polarization in passive LR applications. Our motivation has been to auak low contrast and clutter-limited target detection problems where radiometric conirast is an insufficientdiscriminate. This work has included a number ofstudies [8][9][10][11][12][13][14][15][16] involving laboratory measurements, field measurements, and theoretical analyses; these are summarized in this paper. BASICS OF PASSIVE INFRARED POLARIMETRIC SIGNATURESIt is well known that unpolarized light reflected from smooth surfaces (at non-normal incidence) acquires an elliptical polarized component whose major axis is oriented perpendicular to the surface normal. Said differently, the surface reflectance is greater for electric field components oriented perpendicular to the surface normal. It isless welirecognized thatKirchoff's Law applies so that infrared radiation emitted from smooth surfaces is also partially polarized, but in this case with a preference to electric field components oriented parallel to the surface nonnal. At infrared wavelengths, it is the partial polarization of emitted radiation that is of primary (but not exclusive) interest.The observed polarized signature component depends on many factors including the view angle, material refractive index, surface roughness, temperature, the radiative environment, and the surface shape seen within a single viewing resolution element. First, no effect is seen when viewing a surface at normal incidence. In practice, the greatest polarized signature components are generally observed at viewing angles in the 60-80 degree range (measured from nonnal), while useful signature levels can be expected from approximately 30 to nearly 90 degrees. Second, the material complex refractive index...

We recently demonstrated that 25 mg of bromfenac, a new nonsteroidal anti-inflammatory analgesic, is at least as effective as 400 mg of ibuprofen in relieving postoperative oral surgery pain. Our objective in this study was to determine whether higher doses were significantly more effective. Two hundred eighty (280) outpatients with postoperative pain after the surgical removal of impacted third molars were randomly assigned, on a double-blind basis, a single oral dose of 10, 25, 50, or 100 mg bromfenac; 650 mg aspirin; 400 mg ibuprofen; or placebo. Subjects rated their pain and its relief for 8 hours. All active treatments were significantly superior to placebo, and bromfenac and ibuprofen were significantly superior to aspirin. The slope of the dose-response curve of bromfenac was significant. The 100 mg bromfenac dose was significantly more effective than the 400 mg ibuprofen dose and had a significantly longer duration of analgesic action.