Evaluation of the diffusion barrier continuity on porous low-k films using positronium time of flight spectroscopy H. K. M. Tanaka, 1,2, * T. Kurihara, 3 and A. P. Mills, Jr. 1 1 Physics Department, University of California, Riverside, Riverside, California 92521, USA 2 Muon Science Laboratory, Institute of Materials Structure Science (IMSS), High Energy Accelerator Research Organization (KEK), 1-1 Oho, Tsukuba, Ibaraki 305-0801, Japan 3 Slow Positron Facility, Institute of Materials Structure Science (IMSS), High Energy Accelerator Research Organization (KEK), 1-1 Oho, Tsukuba, Ibaraki 305-0801, Japan Received 4 August 2005; revised manuscript received 30 September 2005; published 11 November 2005 A depth-profiled positronium time of flight experiment was performed to investigate a diffusion barrier on a spin-on low dielectric constant low-kporous silica film. We studied a capping layer of 50 nm SiO 2 , which was optimized for 3-keV positron e + penetration through the capping and stopping in the porous layer. Maximum positronium intensity was found with a positron implantation energy of 3 keV that reflects the sample structure and is consistent with an open pore fraction 2 10 -4 of the capping layer. DOI: 10.1103/PhysRevB.72.193408 PACS numbers: 81.20.-n, 29.40.Mc I. INTRODUCTION Bose-Einstein condensation BECeffects of many posi- tronium atoms Ps; the electron bound state with its antipar- ticle, the positronwould be interesting to observe in a ma- terial in which there are a great many interconnected pores, such that Ps can diffuse over long distances on the order of microns 1 . If we can confine the Ps in a very thin porous film by preventing escaping into vacuum, a dense Ps gas can be more easily produced. For the purpose of the production of a dense Ps gas, a diffusion barrier must be continuous and very thin, and the tradeoff between continuity and thickness of the barrier affects the Ps gas density in the porous layer under- neath. It is therefore important to establish a method to char- acterize such film properties. Positronium annihilation life- time spectroscopy PALShas been applied to determine whether a diffusion barrier on an open-pored low-k film is continuous. 2–4 From such measurements one can infer whether or not Cu atoms or ionscan penetrate into the dielectric and degrade the electrical properties. A Ta barrier only 5 nm thick is not completely effective in preventing Ps diffusion out of a low-k film. 5 On the other hand, a capping layer of 80 nm silica effectively blocks the diffusion of Ps out of a low-k film. 2 Recently, depth-profiled positronium lifetime spectroscopy is used to probe the pore characteristics in porous, low-k silica films. 1 An alternative or complemen- tary method measuring permeability of such a diffusion bar- rier would be useful. In the past several years, on the other hand, positronium time of flight spectroscopy Ps-TOFhas become an estab- lished technique for probing Ps emission from surface or Ps diffusion in matter. 6,7 Ps-TOF spectroscopy might be a complementary method to obtain further understanding by separating decays in the porous matrix from decays attrib- uted to Ps escape into vacuum. 8 In this work, we used the Ps-TOF spectrometer developed at Slow Positron Facility, High Energy Accelerator Organization KEK-SPFto inves- tigate the interrelated effects of the implantation positron en- ergy on the Ps-TOF spectrum in a capped low-k film. At KEK-SPF, the positron beam is completely depolarized. If the positrons from which the positronium is formed are not perfectly spin polarized, there could be spin exchanging trip- let positronium-triplet positronium collisions of dense Ps in the porous medium that would increase the average annihi- lation rate. 9 Our goal is to understand the structural effects of the sample that will enable us to optimize the barrier thick- ness as required for eventual measurements of spin exchange cross section and possibly the Bose-Einstein condensation of Ps atoms. II. EXPERIMENT The pore interconnectivity of the diffusion barrier was examined by Ps-TOF spectroscopy at room temperature. The present experiment was performed at Slow Positron Facility, High Energy Accelerator Research Organization. 10 The facil- ity consists of a 50 MeV linac, an assembly of slow positron generator, a slow positron transport line and an experimental station for positron time of flight Ps-TOFspectroscopy. The experimental setup is shown in Fig. 1. The principle of the Ps-TOF method has been previously discussed, 11 and is only briefly introduced here. In the TOF measurement, the sample is bombarded with a slow positron beam. The time interval between the linac signal and the detection of the gamma ray from the emitted o-Ps self annihilation is measured to obtain the energy distribution of the emitted Ps. In the present ex- periment, the lead collimator was adjusted to effectively re- duce the annihilation gamma rays from the sample region to detect only the decay events from self-annihilation of Ps in the view of the plastic scintillator through a 4.5-mm lead slit. And at the same time, however, the counter also detects a part of the decay events after passing through the lead shield from the sample region the decay events from the sample are not totally shielded by the lead blocks. If the diffusion barrier is not impervious, Ps can diffuse out of the film and it can be easily detected by the Ps-TOF spectrometer. We performed a Ps-TOF measurement experiment at the downstream end of the beam line. The pulse heights and PHYSICAL REVIEW B 72, 193408 2005 1098-0121/2005/7219/1934084/$23.00 ©2005 The American Physical Society 193408-1