ToF-SIMS analysis of glass fiber cloths for PCB manufacturing Dylan Boday 1, Michael Haag 2, Joe Kuczynski 3, Markus Schmidt 2, Michael Wahl 4, Johannes Windeln 2 1 IBM Systems & Technology Group, 9000 S Rita Rd, Tucson AZ 85744-0002, dboday@us.ibm.com 2 IBM Deutschland MBS GmbH, Hechtsheimer Str. 2, D-55131 Mainz, mhaag@de.ibm.com 3 IBM Systems & Technology Group, 3605 Hwy 52 N, Rochester MN 55901-1407, kuczynsk@us.ibm.com 4 IFOS GmbH, Trippstadter Straße 120, D-67663 Kaiserslautern, wahl@ifos.uni-kl.de
content PCB (Printed Circuit Board) structure glass cloth conditioning motivation for ToF-SIMS analysis analytical setup ToF-SIMS results summary and outlook 11/11/11 2
PCB structure heat, pressure prepregs cores copper lines prepregs: glass fiber cloth pre-impregnated with resin matrix heat, pressure pictures source: www.lpkf.de cores: cured glass cloth/resin composite with structured copper layer(s) multi-functionality of composite (glass cloth, resin and optional filler) - mechanical properties (strength, CTE...) - dielectric properties (insulation, impedance...) 11/11/11 3
glass cloth conditioning glass fibers pre-treated with Silanes for - improved surface wetting (contact angle reduction) - improved matrix adhesion (organofunctional group R to match matrix resin) 11/11/11 4
glass cloth conditioning process steps critical with regard to Silanation: - starch size required for yarn processing - pyrolytic de-sizing (thorough?) - Silanation process application of starch sizing pyrolytic de-sizing source: JPS Composite Materials (wet) application of Silanes 11/11/11 5
motivation for ToF-SIMS analysis challenges for modern high performance boards - increasing # of copper layers ( 40 and above) - decreasing line width and spacing ( 50µm and below) - RoHS (Pb-free solder; solder process + 20K, new resins) known sporadic failures become epidemic! SEM micrograph: IBM Thermal cracks: CAF: Copper Anodic Filament pictures source: Dynamic Details Inc. 11/11/11 6
motivation for ToF-SIMS analysis TGA and SEM/EDX results of analysed glass cloth indicate presence of undesired material expected weight loss < 0.1%! organic debris TGA analysis and SEM micrographs: IBM 11/11/11 7
ToF-SIMS Time-of-Flight Secondary Ion Mass Spectrometry extractor accelerates secondary ions to: E = q U (Sample) mass dispersion via ion flight time: v = t = 2E m s v = = s 2q U m m 2q U 11/11/11 8
analytical setup Method: ToF-SIMS static and dynamic mode Tool: ION-TOF IV at, Germany Primary ions: 25keV Bi + 3 (clusters to promote molecular secondary ions) Samples: various samples of glass cloth supposedly de-sized and coated with coupling agent sample # 1 2 3 4 5 6 supplier A B C C C C cloth 2116 1080 1080 2113 0106 2116 11/11/11 9
ToF-SIMS - static mode raw spectra significant differences especially in the hi-mass region #6 #6 #6 11/11/11 10
ToF-SIMS - dynamic mode dynamic mode to distinguish between signals from surface coverage and bulk material glass bulk sputter removal surface coverage sputter removal 11/11/11 11
ToF-SIMS first results ToF-SIMS imaging capability (microscopic picture 300µmx300µm in the light of an ion mass ) major signals assigned to one of 4 peak groups with common appearance and behaviour group A: group B: group C: group D: 11/11/11 12
ToF-SIMS peak groups group A (glass constituents) laterally homogeneous signal glass constituents (Li, B, Mg, Al, Si, Ca, Ti, Na, K) signal increased after sputter removal of surface layer ( 30nm) confirms glass constituent different: Na and K: signal increase after sputter on #7, signal decrease on #1 sizing component? after sputter removal of surface layer: 11/11/11 13
ToF-SIMS peak groups group B (inhomogeneity I) signal limited to prominent portions of single fibers Sulfate component signal decreased after sputter removal of surface layer ( 30nm) confirms surface coverage residues from sizing or just abrasion debris from plastic bag? after sputter removal of surface layer: 11/11/11 14
ToF-SIMS peak groups group C (homogeneous organics) laterally homogeneous signal mainly hydrocarbons (C 7 H 7, C 9 H 7, C 9 H 9, C 10 H 8 ) phosphate component signal decreased after sputter removal of surface layer ( 30nm) surface coverage coupling agent? after sputter removal of surface layer: 11/11/11 15
ToF-SIMS peak groups group D (inhomogeneity II) signal confined to small areas on single fibers fragmentation pattern match: Erucamide (m=337) signal decreased after sputter removal of surface layer ( 30nm) surface coverage coupling agent or residues from sizing? after sputter removal of surface layer: 11/11/11 16
ToF-SIMS second run Method: ToF-SIMS static and dynamic mode Tool: ION-TOF IV at, Germany Primary ions: 25keV Bi + 3 (clusters to promote molecular secondary ions) Samples: 1. 2116 glass cloth w/ Silane treatment (supplier s claim) - as received - heat cleaned (matching pyrolytic de-sizing) - UV-ozone cleaned (details undisclosed) 2. pellet of glass raw material (as used to pull fibers for 2116) analyzed as a reference - as receiced - on freshly fractured surface - on sputter cleaned surface 11/11/11 17
first results 1,0E+05 Glass cloth analyses signal 1,0E+04 1,0E+03 heat treated UV ozone 1,0E+02 as received 1,0E+01 CH_3+ C_2H_3+ C_2H_5+ C_3H_5+ C_3H_7+ C_4H_7+ C_4H_9+ C_5H_7+ C_5H_9+ C_2H_3O+ C_3H_3O+ C_3H_5O+ element / fragment NO SiCH 3 O + as reported by Norrmann et al (J. Mass Spectrom. 2002, 37, 695-708) NO other SiC complexes potentially Silane related: C x H y+ and C x H y O + but no change after heat and UV ozone treatment! 11/11/11 18
first results 1,0E+06 Glass pellet analyses signal 1,0E+04 1,0E+02 fracture surface sputter cleaned as received 1,0E+00 C_2H_3+ C_2H_5+ C_3H_5+ C_3H_7+ C_4H_7+ C_4H_9+ C_5H_7+ C_5H_9+ C_2H_3O+ C_3H_3O+ C_3H_5O+ element/fragment potentially Silane related species C x H y+ and C x H y O + also present on bulk reference sample! adsorbates from the ambient 11/11/11 19
summary and outlook new ToF-SIMS application in reverse engineering: Silane treatment of glass fibre cloth first results on supplier samples: NO Silane related secondary ions detected functionalized surface not as desired - Silanes either not present - or only physisorbed instead of chemisorbed (due to insufficient fibre treatment) next steps: work with supplier to improve Silanation process Is pad cratering part of the same problem? Is this the time to involve computational materials science? 11/11/11 20