Poly(p-xylylene) has existed commercially for many years but has never gained widespread acceptance due to its poor thermal and optical properties. Parylene X, a new chemical vapor depositable polymer on the same process platform as poly(p-xylylene), is a copolymer of poly(ethynyl-p-xylylene) and poly(p-xylylene). It exhibits thermal stability to at least 420°C, has a low coefficient of thermal expansion of 55 ppm after it is crosslinked, and exhibits low stress. It possesses a glass transition temperature of 73°C and starts to crosslink at ∼160°C with a peak isotherm at 250°C. The ethynyl functional groups crosslink to form the very stable phenyl moiety.

Poly(p-xylylene) (also known as parylene N) has previously been used to pore seal ultralow k (≤ 2.2) (ULK) dielectrics. The parylene polymers may facilitate the integration of ULK dielectrics by: substantially improving their fracture toughness, hermetically sealing the pores, being able to use standard wet chemical cleans, and minimally impacting the observed dielectric constant, while minimally disrupting current process flow integrations. This paper introduces a new cross-linkable polymer that is deposited using thermal chemical vapor deposition (CVD) on the same tool that is used for parylene N deposition. The polymer, poly(ethynyl-p-xylylene) (parylene X), was deposited at room temperature. A series of 30 min post-deposition anneals in helium shows that the deposited material cross-linked between 200 °C and 300 °C with full conversion at 380 °C for a ~300 Å film. After the low molecular weight species out-gassed during anneals at 200 °C, there was less than a percent weight loss to 450 °C with no change in the optical constants and no optical loss. Previous work with poly(ethyl-p-xylylene) suggests that the dielectric constant of parylene X will be significantly lower than parylene N.

Using the parylene polymer platform, benzylic brominated poly(trifluoroacetyl-p-xylylene) thin films are deposited at room temperature from a novel synthesized precursor, 2-trifluoroacetyl-a,a,a′ tribromo-p-xylene using CVD. The as-deposited benzylic brominated parylene-type polymer is nearly transparent. However, upon subsequent post-deposition anneals dehydrohalogenation can take place, yielding vinyl and ethynyl moieties and a highly conjugated absorbing polymer. Previous attempts with the deposition and annealing of the analogous non-functionalized phenyl group polymer did not yield the desired structure, due to polymer crystallization that 'locked up' the structure preventing any further chemical reactions from occurring. The as-deposited films contain much hydroxyl bonding; however, after post-deposition annealing, the films dehydrate and hydrogen bromide eliminates at 200, 300, and 380°C. During HBr removal the films become much thinner, more absorbing in the visible region, and denser. After the 300°C anneal the films stay amorphous, as seen from X-ray diffraction (XRD) measurements, and some bromine still resides in the structure. 0.4 bromine atoms per repeat unit are left compared with 1.8 bromine atoms per repeat unit in the as-deposited condition.

No abstract

Disk amplifier design for inertial fusion lasers has evolved with changing fusiondriver requirements from a primary emphasis on gain to a primary emphasis on efficiency. In this paper we compare Shiva and Nova amplifiers to a developmental amplifier (SSA) and show greater than a two -fold improvement in efficiency over past designs under all operating conditions. Experiments to optimize the efficiency of the SSA show that preionization of the flashlamps produces significant benefits and that the packing fraction of lamps is more important than the flashlamp reflector shape. They also show that the optimized flashlamp pulselength and reflector geometry depend on the desired stored energy in the laser medium.We have demonstrated a 7% storage efficiency at a stored fluence per disk of 0.5 J /cm2 (stored energy density of 0.06 J /cm3) and 4% at 2.0 J /cm2 (0.25 J /cm3).Comparison of SSA measurements with storage -efficiency calculations show that our flashlamp model accurately predicts the single -pass pumping of disk amplifiers.