報告_OCP2026_KYZEN_浸沒前清洗化學_20261015

PDF 原檔：報告_OCP2026_KYZEN_浸沒前清洗化學_20261015_original.pdf

講者公開簡報（OCP 官方 Recorded Content 頁提供下載）。文字以 PyMuPDF 逐頁抽取，### p<n> 為頁碼；圖表內容見原檔。

圖片清單（已驗證 2026-10-15）

原始內容

p1

Chelsea Jewell Application Lab Manager | KYZEN Before the Fluid: How Pre-Immersion Cleaning Chemistry Shapes Long-Term Electronic Reliability Cooling Environments Immersion Project

p2

Reliability Begins Before Fluid Exposure Every assembly enters the fluid with a manufacturing and cleaning history.

p3

OCP Guidance Establishes the Context • Residual solder flux and process residues are recognized contamination sources. • Cleaning and rinse conditions influence what enters the dielectric fluid. Manufacturing Residues and Cleaning • Assemblies and supporting materials can interact with immersion fluids. • Contaminant transfer and material compatibility are reliability considerations. Material Compatibility • Fluid quality management includes contamination, fluid condition, and compatibility. • Monitoring methods support long-term immersion reliability. Fluid Condition and Lifecycle

p4

Experimental Design A TEST CONDITIONS As Built Manufacturing condition retained Robustly Cleaned Complete flux removal Chemistry Mismatched Poor flux/chemistry solubility match Poorly Rinsed Complete flux removal. Poor chemistry rinse Post-Immersion IPA Rework Manual IPA cleaning after immersion fluid exposure B C D E COMPARISON STRATEGY A B C D E B B Residue Condition Chemistry Type Rinse Completeness Post-Immersion Rework All cards: SAC305 / ROL0 No-Clean flux. B and D: Solvent-based aqueous chemistry; C: Alkaline-based aqueous chemistry.

p5

Fluid Exposure and Measurement Strategy Initial Condition A As Built B Robustly Cleaned C Chemistry Mismatched D Poorly Rinsed E Post-Immersion IPA Rework Fluid Exposure: Continuous EID Monitoring Independent fluid for each condition Two cards run in sequence; card 2 entered card 1’s heated fluid 4 channels per card, logged every 30 s Post-Exposure Analysis Assembly • Visual inspection • Localized C3 extraction • Ion chromatography Fluid • Baseline fluid sample • Post-exposure sample • GCMS / LCMS screening

p6

Electrical Response: Defined by Recovery Card 1, full 168 h exposure 1E+7 1E+8 1E+9 1E+10 1E+11 0 24 48 72 96 120 144 168 Average EID resistance, Ω Exposure time, hours A As Built B Robustly Cleaned C Chemistry Mismatched D Poorly Rinsed E Post-Immersion IPA Rework Source: Immersion Canary EID, card 1, 30 s logging. Log-average of all four channels (CH1–CH4), 1 h rolling average. C excludes CH3, which read at the instrument floor (7.3×10⁵ Ω) for the full run.

p7

Electrical Response: Defined by Recovery Card 1 Fresh fluid; heaters on 0.6 h after entry 1E+7 1E+8 1E+9 1E+10 1E+11 0 12 24 36 48 Avg. EID resistance, Ω Exposure time, hours Card 2 Same fluid, already heated; 54 h before power outage 1E+7 1E+8 1E+9 1E+10 1E+11 0 12 24 36 48 Avg. EID resistance, Ω Exposure time, hours Hours to 1×10¹⁰ Ω B Robustly Cleaned D Poorly Rinsed A As Built C Chemistry Mismatched E IPA Rework Card 1 7.3 h 16.9 h 30.5 h 32.6 h 47.4 h Card 2 3.7 h 31.0 h 21.8 h 14.5 h 23.8 h Log-average of channels, 1 h rolling average; C excludes Channel 3 on both cards.

p8

Electrical Response: Defined by Recovery 1 10 100 Hours to reach 1×10¹⁰ Ω (log scale) 3.7 h 7.3 h B Robustly Cleaned 16.9 h 31.0 h D Poorly Rinsed 21.8 h 30.5 h A As Built 14.5 h 32.6 h C Chemistry Mismatched 23.8 h 47.4 h E IPA Rework ●Card 1: fresh fluid ◆Card 2: pre-heated, used fluid Log-average of channels, 1 h rolling average on log scale; C excludes Channel 3. Card 2 window: 54 h.

p9

Post-Exposure Assembly Characterization Localized C3 Extraction + IC Channel 1 component site, after exposure 0.67 1.40 0.82 1.67 1.02 0.0 0.5 1.0 1.5 2.0 Sulfate Succinate Adipate µg/in² A As Built E IPA Rework Sulfate not detected on E. No species detected B C D Weak organic acids A 2.22 µg/in² E 2.70 µg/in² Localized extraction, not whole-board cleanliness. Post-Exposure Visual Inspection A As Built B Robustly Cleaned C Chemistry Mismatched D Poorly Rinsed E IPA Rework Representative site for each condition after exposure; no cleaning after removal from fluid.

p10

Fluid Analysis GCMS: All Five Sets After 168 h A B C D E Every set matches the clean-fluid control (repeated at 332 h). LCMS: 168 h, Zoomed 16–22.5 min Blank Clean A B C D E 19.1 min m/z 249.2065 Small peak in every exposed set; absent from the clean fluid. Present regardless of cleaning: likely a minor extractable from the part. Peaks at 18.6–18.7 min also appear in the blank. One minor common feature; no condition-specific signal. Bulk fluid organic analysis did not differentiate the tested conditions. Methods: GCMS, Thermo Trace 1310 GC / ISQ LT MS, 10× in n-hexane. LCMS, Agilent 1260 Infinity / 6545 Q-TOF, 10× in toluene. Organic screening only; not designed to quantify trace ionic contamination. Source: J. Miller, Organic Chemistry Lab, Micron Technology, September 2026. LCMS for the 332 h set was not reported due to instrument noise.

p11

Connecting Electrical, Surface, and Fluid Behavior Pre-immersion condition was intentionally varied through residues, cleaning chemistry, rinse completeness, and rework history. Incoming Assembly Condition Only robustly cleaned boards stabilized quickly on both cards. Mismatched chemistry, poor rinse, and IPA rework were no better than no cleaning. Electrical Response Electrically recovered assemblies did not present identical post- exposure surface conditions. Localized IC and visual inspection retained evidence of processing history. Surface Characterization Bulk organic analysis did not clearly distinguish the pre- immersion conditions at this test scale. A dielectric-fluid ionic test method is still needed. Fluid Analysis Cleaning for immersion is a distinct process requirement: chemistry match and rinse completeness must be specified and verified. Electrical, surface, and fluid measurements are complementary. Qualification Implication

p12

Call to Action OCP CE - IC - Fluids - Cleaning Guidelines OCP CE – Immersion Cooling How can we strengthen immersion qualification by better defining the incoming assembly condition? Join us in the OCP Immersion Cooling Fluids & Cleaning work. We're looking for partners to extend this with more boards, fluids, and labs. • Record incoming assembly condition in qualification Propose adding cleaning chemistry, rinse verification, and T₀ cleanliness to the Immersion Fluids & Cleaning Guidelines. • Close the ionic gap in dielectric fluid Organic screening did not separate the conditions; we need a method that measures ionic contamination directly in the fluid. • Align qualification practices Build a shared framework across reliability, component compatibility, fluid lifecycle, and cleaning communities.

p13

（本頁無可抽取文字，內容在圖中）