⛲ 微光花園 World 3：水泥花園Micro Bloom · World 3: Cement Garden
⭐ 0🌸 0
▼ 點一下或按 A 繼續Tap or press A
Micro BloomWorld 3 · 水泥花園World 3 · Cement Garden
⏸ 暫停Paused
⚙️ 設定Settings
語言Language
🎵 背景音樂Music
🎼 曲目Track
🕹️ 螢幕搖桿On-screen controller
💬 劇情提示Story & hints
🔔 音效Sound effects
⛶ 全螢幕Full screen
🖥️ 繪圖Renderer
✨ 畫質Quality
1280×720 · 1920×1080 · 2560×1440 (2K)
設定會記在這台裝置的瀏覽器中。若裝置不支援 WebGPU，會自動改用軟體繪圖。Settings are remembered in this browser. If WebGPU is not available, the game switches to software rendering automatically.
💾 存檔／讀檔Save / Load
自動Auto
槽位 1Slot 1
槽位 2Slot 2
槽位 3Slot 3
存檔保存在這台裝置的瀏覽器中，過關時會自動存檔。快速鍵： 存到槽位 1、 讀取槽位 1。在關卡中存檔，讀檔後會從該關開頭重新開始。清除瀏覽器資料會刪除存檔，請用「匯出」備份。Saves stay in this browser on this device; the game also autosaves after every stage. Hotkeys: saves to slot 1, loads slot 1. A save made inside a stage restarts that stage from its beginning. Clearing browser data erases saves, so export a backup.
⌨️ 自訂按鍵Custom keys
上Up
下Down
左Left
右Right
A（確定）A (confirm)
B（返回）B (back)
暫停Pause
提示Hint
快速存檔Quick save
快速讀檔Quick load
按「變更」後再按下想用的鍵（Esc 取消）。Enter 永遠可當 A，Esc 永遠可當返回／暫停。也支援 USB／藍牙手把。Press Change, then the key you want (Esc cancels). Enter always works as A, and Esc always works as Back / Pause. USB / Bluetooth gamepads also work.
⭐ Star Quiz
🌟 過關！Stage clear!
★★★
📖 精靈圖鑑與概念明信片Sprite album & concept postcards
📘 遊戲說明書How to play
⛲ 故事Story
World 2 的 X 光水晶燈塔亮起後，皮普、叮叮和夥伴們沿著光束來到最後一站「水泥花園」。花園中央的「許願噴泉」在漏水：池壁長出白色結殼，靠近水面的漿體變得又軟又多孔。After the X-ray Crystal Lighthouse lit up in World 2, Pip, Zing and friends follow its beam to the final stop: the Cement Garden. The Wishing Fountain at its centre is leaking. White crusts cover the basin, and the paste near the water line has turned soft and porous.
花園管家「凝凝」（一團 C-S-H 凝膠）請大家幫忙：用 EDS 找出結殼的成分、做出正確的試片、讀懂水泥的微結構，最後找到偷偷溶走鈣的搗蛋水滴「溶溶」。Gelly, the garden keeper and a puff of C-S-H gel, asks for help: use EDS to find what the crust is made of, prepare proper specimens, read the cement microstructure, and finally catch Drippy, the prankster droplet who is quietly dissolving the calcium away.
這是三個世界的最後一章。同一個部落格上若有 World 2（或 World 1）的存檔，開始新遊戲時會自動帶入之前所有的夥伴與花瓣。This is the last chapter of the three worlds. If this blog has a World 2 (or World 1) save in your browser, a new World 3 game brings all your earlier friends and petals along.
🎮 操作Controls
- 鍵盤：方向鍵移動／調整滑桿，空白鍵或 Enter 確定，Backspace 返回，P 暫停，H 提示（可在「自訂按鍵」更改）。Keyboard: arrow keys move / adjust sliders, Space or Enter confirm, Backspace back, P pause, H hint (change them in Custom keys).
- 滑鼠／觸控：在花園點地面，皮普就會走過去；點平台可進入關卡；滑桿可直接拖曳。Mouse / touch: tap the ground and Pip walks there; tap a pad to enter a stage; drag sliders directly.
- 螢幕搖桿：在設定或下方按鈕開啟，左邊十字鍵、右邊 A（確定）與 B（返回）。On-screen controller: turn it on in Settings; D-pad on the left, A (confirm) and B (back) on the right.
- 也支援 USB／藍牙手把。USB / Bluetooth gamepads also work.
- 💡 提示按鈕：重播本關的講解。📖 圖鑑：看精靈與明信片。💡 Hint replays the stage explanation. 📖 Album shows your sprites and postcards.
🌼 如何種花How to plant flowers
- 收集花瓣 🌸：新遊戲送 20 片花瓣，另外加上 World 2（或 World 1）剩下的花瓣。第一次過關，每顆星換 12 片；重玩過關每顆星換 4 片。左上角的 🌸 是目前的花瓣數。Collect petals 🌸: a new game starts with 20, plus any petals left over from World 2 (or World 1). A first clear gives 12 petals per star; a replay gives 4 per star. The 🌸 counter at the top left shows how many you have.
- 在花園散步：用方向鍵、搖桿或點地面讓皮普走動。每走一小段路，就會自動種下一朵花、用掉 1 片花瓣。Walk in the garden: move Pip with the arrow keys, the controller or by tapping the ground. Every few steps a flower is planted automatically, using 1 petal.
- 花的顏色：跟在皮普後面的精靈會決定花色。三個世界的精靈都會一起種花，花園就越繽紛（方方白、鐵鐵金、針針紫、板板藍……）。Flower colours: the sprites following Pip choose the colours. Sprites from all three worlds plant together, so the garden gets more and more colourful (Calci white, Ferri gold, Etty violet, Porty blue…).
- 小規則：關卡平台上不會種花；花園最多保留約 420 朵，種滿後最舊的花會讓位給新花。花瓣用完就先去過關或重玩拿星星。Small rules: no flowers are planted on stage pads; the garden keeps about 420 flowers, and the oldest make room for new ones. Out of petals? Clear or replay a stage for more stars.
- 會存起來：種下的花會跟著存檔保存，破關畫面會顯示你種了幾朵。They are saved: your flowers are stored with your save, and the ending shows how many you planted.
⛲ 關卡（對應 PCAM.03）Stages (PCAM.03)
- 3-1 叮叮射峰：EDS 定性分析，從最大、能量最高的峰開始，用整個譜線家族確認（s.5–9、14）3-1 Zing's Peak Archery: qualitative EDS; start with the largest, highest-energy peak and confirm with the whole line family (s.5–9, 14)
- 3-2 ZAF 蹺蹺板：k 比值與 Z、A、F 基材校正，分析總和的檢查（s.10–13）3-2 ZAF Seesaw: k-ratios, Z, A and F matrix corrections, and checking the analytical total (s.10–13)
- 3-3 試片三道菜：依目的選擇形貌、微結構形態、微區分析試片的製備（s.15–19）3-3 Three-Course Specimens: choose the preparation route for topography, morphology or microanalysis (s.15–19)
- 3-4 拋光與充電：研磨拋光順序、潤滑液、鍍膜與充電效應（s.20–26）3-4 Polish & Spark: grinding and polishing order, lubricant, coating and charging (s.20–26)
- 3-5 熟料馬賽克：BSE 影像中的 alite、belite、ferrite、aluminate 與孔隙（s.29–31）3-5 Clinker Mosaic: alite, belite, ferrite, aluminate and pores in BSE images (s.29–31)
- 3-6 水化森林：SEI 的 AFt、AFm、CH、C-S-H 形貌，BEI 的相與內／外部產物，孔隙（s.32–38）3-6 Hydration Forest: AFt, AFm, CH and C-S-H shapes in SEI; phases and inner/outer product in BEI; pores (s.32–38)
- BOSS 溶溶：總複習，加上水化時間軸與「五指描述法」挑戰BOSS Drippy: review quiz plus the hydration timeline and the five-descriptor challenge
關卡依序開放：過了前一層，下一層的平台才會發光。Stages open in order: a pad lights up once the previous one is cleared.
⭐ 星星與存檔Stars & saves
- 答錯不會失敗，可以一直重試。全對 3 顆星，錯 1–2 次 2 顆星，錯更多次 1 顆星；可重玩刷新最佳紀錄。Wrong answers never end the game; just try again. No mistakes = 3 stars, 1–2 mistakes = 2 stars, more = 1 star. Replay to beat your best.
- 過關會自動存檔；也可在「存檔／讀檔」存到槽位 1–3，或匯出成檔案帶到別台裝置。The game autosaves after every stage; you can also save to slots 1–3 or export a file to use on another device.
- 在關卡中存檔，讀檔後會從該關開頭開始。A save made inside a stage restarts that stage from its beginning.
📱 手機與全螢幕Phone & full screen
- 按「⛶ 全螢幕開始」最好玩；手機請橫向握持。“⛶ Full screen & play” works best; hold your phone sideways.
- 離開全螢幕：按右上角「✕ 離開全螢幕」，或在暫停選單按「✕ 離開全螢幕」，或在設定把全螢幕切成「關」。電腦上也可按 Esc。Leaving full screen: tap “✕ Exit” at the top right, or “✕ Exit full screen” in the pause menu, or switch Full screen off in Settings. On a computer, Esc also works.
- 全螢幕時畫面固定，不會被拖動或縮放。In full screen the view is locked: no scrolling or zooming.
⛲ World 3 全破！三個世界全部完成World 3 clear! All three worlds complete
恭喜成為 SEM 大師！最終星之挑戰：為許願噴泉的混凝土寫一頁 SEM 分析計畫（試片、影像、EDS、近水面與核心的微結構描述）。Congratulations, SEM master! Final Star Challenge: write a one-page SEM plan for the Wishing Fountain concrete (specimens, images, EDS, and a description of the water face vs. the core).
AEML, NTUST
遊戲無法啟動The game could not start
此遊戲由 AEML, NTUST 製作。若程式碼遭到修改，或瀏覽器封鎖了 JavaScript，將無法執行。This game was created by AEML, NTUST. It will not run if its code has been modified or JavaScript is blocked.
HINTPAUSE
BA
AEML, NTUST
操作： 移動／調整 · 或 Enter 確定 · 返回 · 暫停 · 提示。也可以直接點選或拖曳畫面。請先點一下遊戲畫面，按鍵才會送進遊戲。Controls: move / adjust · or Enter confirm · back · pause · hint. You can also tap and drag on the screen. Click the game screen first so keys go to the game.
Micro Bloom World 3 | (c) 2026 AEML, NTUST | CT5812701 PCAM
命中！Hit!
先完成前一層！Clear the previous floor first!
已存檔至槽位Saved to slot
已讀取存檔。Save loaded.
這個槽位沒有存檔。That slot is empty.
已刪除存檔。Save deleted.
已匯出存檔檔案。Save file exported.
已匯入存檔。Save file imported.
存檔檔案不正確或已被修改。That save file is invalid or has been modified.
瀏覽器儲存空間不足，請改用匯出。Browser storage is full. Use Export instead.
此裝置無法使用 WebGPU，已改用軟體繪圖。WebGPU is not available here, so software rendering is used.
已啟用 WebGPU 硬體繪圖。WebGPU hardware rendering is on.
請按新按鍵…Press a key…
按鍵已更新。Key updated.
已恢復預設按鍵。Default keys restored.
已恢復預設設定。Default settings restored.
這裡沒有額外提示，加油！No extra hint here. You can do it!
不太對喔，再試一次！Not quite. Try again!
要離開全螢幕，按右上角「✕ 離開全螢幕」或暫停選單。To leave full screen, tap “✕ Exit” at the top right or use the pause menu.
把手機橫過來，畫面會更大！Turn your phone sideways for a bigger screen!
移動：方向鍵、搖桿或點地面 · 確定：空白鍵／AMove: arrow keys, controller or tap the ground · Confirm: Space / A
就緒。Ready.
已暫停。Paused.
重新開始本關。Stage restarted.
此瀏覽器不支援網頁音訊。This browser does not support web audio.
目前：WebGPU 硬體繪圖Active: WebGPU (hardware)
目前：軟體繪圖（Canvas 2D）Active: software (Canvas 2D)
目前：軟體繪圖（此瀏覽器無法使用 WebGPU）Active: software (WebGPU unavailable in this browser)
正在啟動 WebGPU…Starting WebGPU…
目前解析度：Current resolution:
自動（依場景變換）Auto (changes with the scene)
開始新遊戲？自動存檔會被覆蓋（槽位 1–3 不受影響）。Start a new game? The autosave will be overwritten (slots 1–3 are kept).
回到標題畫面？本關未完成的進度會消失。Return to the title screen? Progress in this stage will be lost.
刪除這個存檔？Delete this save?
讀取這個存檔？目前未存檔的進度會消失。Load this save? Unsaved progress will be lost.
從頭開始新遊戲？自動存檔會被覆蓋。Start over with a new game? The autosave will be overwritten.
獲得花瓣Petals earned
新精靈孵化！New sprite hatched!
解鎖概念明信片：Postcard unlocked:
最佳紀錄Best
總星數Total stars
種下的花Flowers planted
尚未孵化Not hatched yet
尚未解鎖：過關後取得Locked: clear the stage to get it
（空）(empty)
皮普Pip
叮叮Zing
糊糊菇Blurshroom
影子菇Shadowshroom
灰階鯨Greyling
剪剪Snip
殼殼Shelly
烤烤Toasty
凝凝Gelly
方方Calci
鐵鐵Ferri
台台Stubby
啪啪Sparky
稜稜Ally
圓圓Bello
針針Etty
板板Porty
溶溶Drippy
3-1 叮叮射峰3-1 Zing's Peak Archery
3-2 ZAF 蹺蹺板3-2 ZAF Seesaw
3-3 試片三道菜3-3 Three-Course Specimens
3-4 拋光與充電3-4 Polish & Spark
3-5 熟料馬賽克3-5 Clinker Mosaic
3-6 水化森林3-6 Hydration Forest
BOSS 溶溶BOSS Drippy
A：進入A: Enter
先完成前一關Clear the previous stage first
加速電壓Accelerating voltage
一起冒險的夥伴Friends on the adventure
從最大、能量最高的峰開始Start with the largest, highest-energy peak
再用整個譜線家族確認Then confirm with the whole line family
白色結殼含有：The white crust contains:
方解石 CaCO₃Calcite CaCO₃
氫氧化鈣 Ca(OH)₂Portlandite Ca(OH)₂
石膏 CaSO₄·2H₂OGypsum CaSO₄·2H₂O
幫星號的峰取名字：先射最大、能量最高的那座Name the starred peak: shoot the largest, highest-energy one first
結殼是哪一種礦物？Which mineral is the crust?
命中！3.69 keV 是 Ca Kα，最大的峰先解決。Bullseye! 3.69 keV is Ca Kα. The biggest peak comes first.
對一下能量：K Kα 在 3.31 keV，Sn Lα 在 3.44 keV，都不是 3.69。Check the energy: K Kα is at 3.31 keV and Sn Lα at 3.44 keV, not 3.69.
同一個家族！Ca Kβ 約是 Ca Kα 的十分之一：Kα : Kβ ≈ 10 : 1。Same family! Ca Kβ is about one tenth of Ca Kα: Kα : Kβ ≈ 10 : 1.
先別急著找新元素：先檢查家族。Ca Kβ 就在 4.01 keV，高度約十分之一。Don't hunt for a new element yet: check the family first. Ca Kβ sits at 4.01 keV at about one tenth the height.
0.52 keV 是 O Kα。0.52 keV is O Kα.
N Kα 在 0.39 keV；若有 Cr，5.41 keV 還會出現 Cr Kα。N Kα is at 0.39 keV; Cr would also show Cr Kα at 5.41 keV.
0.28 keV 是 C Kα。小心：碳鍍膜也會貢獻 C！0.28 keV is C Kα. Careful: a carbon coat also adds C!
B Kα 在 0.18 keV，N Kα 在 0.39 keV。B Kα is at 0.18 keV and N Kα at 0.39 keV.
S Kα！Mo Lα（2.29 keV）在 EDS 會和它重疊，但 Mo 會帶來整個 L 家族（Lβ₁ 2.39 keV）。WDS 約 10 eV 的解析度能清楚分開。S Kα! Mo Lα (2.29 keV) overlaps it in EDS, but Mo would bring a whole L family (Lβ₁ at 2.39 keV). WDS, at about 10 eV resolution, separates them cleanly.
Mo Lα 確實在附近，但沒有看到 Mo 的其他 L 線；Si Kα 在 1.74 keV。Mo Lα is close, but none of Mo's other L lines are here; Si Kα is at 1.74 keV.
Ca + C + O，只有一點點 S：方解石結殼！被溶出的 Ca(OH)₂ 和空氣中的 CO₂ 反應，沉積成白色的殼。Ca + C + O with only a trace of S: a calcite crust! Leached Ca(OH)₂ reacted with CO₂ from the air and deposited as a white crust.
Ca(OH)₂ 沒有碳峰（H 在 EDS 看不到），但我們量到明顯的 C 峰。Portlandite has no carbon peak (H is invisible to EDS), but we found a clear C peak.
石膏的 S 峰會很大；這裡的 S 峰很小。Gypsum would give a big S peak; ours is tiny.
已知為 50 wt% Fe、50 wt% Ni 的合金（30 kV）。未校正的結果卻是 Fe 58.5、Ni 40.5。An alloy known to be 50 wt% Fe and 50 wt% Ni (30 kV). The uncorrected result says Fe 58.5, Ni 40.5.
Fe 與 Ni 原子序相近，Z 校正很小Fe and Ni are neighbours in Z, so the Z correction is small
A：Ni Kα 被 Fe 吸收 → 提高 NiA: Ni Kα is absorbed by Fe → raise Ni
F：Fe 被 Ni 的 X 光激發螢光 → 降低 FeF: Fe is fluoresced by Ni X-rays → lower Fe
A：Fe Kα 被 Ni 吸收 → 提高 FeA: Fe Kα is absorbed by Ni → raise Fe
F：Ni 被 Fe 的 X 光激發螢光 → 降低 NiF: Ni is fluoresced by Fe X-rays → lower Ni
讓蹺蹺板平衡：選出兩個真的會發生的校正Balance the seesaw: pick the two corrections that really happen
結殼的分析總和只有 62%，下一步怎麼做？The crust total is only 62%. What next?
對！Ni Kα（7.47 keV）剛好高於 Fe 的 K 吸收邊（7.11 keV），被 Fe 大量吸收，所以要補回 Ni。Yes! Ni Kα (7.47 keV) lies just above the Fe K edge (7.11 keV), so Fe absorbs it strongly and Ni must be corrected upward.
對！被吸收的 Ni X 光又激發出額外的 Fe X 光（螢光），所以要把 Fe 調低。Yes! The absorbed Ni X-rays excite extra Fe X-rays (fluorescence), so Fe must be corrected downward.
Fe Kα（6.40 keV）低於 Ni 的 K 吸收邊（8.33 keV），不會被 Ni 強烈吸收。Fe Kα (6.40 keV) is below the Ni K edge (8.33 keV), so Ni does not absorb it strongly.
Fe 的 X 光（6.40 keV）能量不夠，激發不了 Ni 的 K 層（8.33 keV）。Fe X-rays (6.40 keV) don't have enough energy to excite the Ni K shell (8.33 keV).
未正規化的分析總和Un-normalised analytical totals
拋光的 Fe–Ni 合金Polished Fe–Ni alloy
拋光漿體中的 C-S-H 點（含水）C-S-H point in polished paste (contains water)
粗糙結殼，無標準品分析Rough crust, standardless
切平、拋光、鍍碳，再分析很多點Cut flat, polish, carbon coat, and analyse many points
直接正規化到 100% 報告Just normalise to 100% and report
把電壓開到 30 kVTurn the beam up to 30 kV
對！定量需要平整、拋光、在 X 光體積內均勻的區域。總和偏低常代表粗糙表面、孔隙或水分；水化物本來就會低於 100%，所以用原子比（Ca/Si、Al/Ca）來看趨勢。Right! Quantification needs a flat, polished region that is homogeneous over the X-ray volume. A low total usually means roughness, porosity or water; hydrates are always below 100%, so look at atomic ratios (Ca/Si, Al/Ca) and trends.
正規化會把問題藏起來：62% 告訴我們表面或樣品有問題。Normalising hides the problem: 62% is telling you something is wrong with the surface or specimen.
電壓太高，X 光體積變大，粗糙與吸收問題反而更嚴重。A higher kV enlarges the X-ray volume and makes roughness and absorption worse.
表面Surface
鍍膜Coating
觀察方式View with
新鮮破斷面，不拋光Fresh fracture, no polishing
環氧樹脂、切割、拋光（不蝕刻）Epoxy, cut, polish (unetched)
拋光後再蝕刻Polish, then etch
Au／Pd 濺鍍Au/Pd sputter coat
碳蒸鍍Carbon evaporation
不鍍膜No coating
SE 影像SE image
BSE 影像BSE image
EDS 點分析（10–15 kV）EDS points (10–15 kV)
上菜！Serve!
結殼的晶體形狀（形貌）Crystal shapes of the crust (topography)
水線附近的孔隙分布圖Porosity map near the water line
C-S-H 的 Ca/Si、Al/SiCa/Si and Al/Si of C-S-H
A 餐：看結殼的晶體形狀，選表面、鍍膜、觀察方式Course A: crust crystal shapes. Choose surface, coating and view
B 餐：池壁混凝土的 BSE 孔隙分布Course B: BSE porosity map of the basin concrete
C 餐：用 EDS 量 C-S-H 的成分Course C: C-S-H composition by EDS
三排都要各選一個喔。Pick one option in each of the three rows.
晶體形狀就在原始表面上，拋光會把它磨掉！The crystal shapes live on the original surface; polishing would grind them away!
看形貌用薄薄的 Au／Pd 濺鍍：原子從各方向飛來，凹凸表面也鍍得到；不鍍會充電。For SE topography use a thin Au/Pd sputter coat: atoms arrive from all directions and cover rough surfaces. No coat means charging.
形貌用 SE 偵測器。Topography means the SE detector.
孔隙分布要平整、灌了環氧樹脂、拋光過的切面；蝕刻會改變表面。A porosity map needs a flat, epoxy-impregnated, polished section; etching changes the surface.
用薄薄的碳膜：原子序低，不會蓋掉 BSE 的成分對比。Use a thin carbon coat: its low Z does not hide the BSE compositional contrast.
BSE 提供相（原子序）對比，灌了環氧樹脂的孔隙看起來是黑的。BSE gives phase (atomic-number) contrast, and epoxy-filled pores look black.
EDS 需要平整、拋光、不蝕刻的表面；蝕刻過的試片不能定量。EDS needs a flat, polished, unetched surface; etched specimens cannot be quantified.
要碳！金會多出 Au M 線，還會吸收低能量 X 光。Carbon! Gold adds Au M lines and absorbs low-energy X-rays.
Ca/Si、Al/Si 要用約 10–15 kV 的 EDS 點分析。Ca/Si and Al/Si come from EDS points at about 10–15 kV.
A 餐上桌：破斷面 + Au／Pd 濺鍍 + SE，針狀和板狀晶體清清楚楚！Course A served: fracture + Au/Pd sputter + SE. Needles and plates in sharp relief!
B 餐上桌：環氧樹脂拋光切面 + 碳膜 + BSE，孔隙一目了然。Course B served: epoxy-polished section + carbon + BSE. Every pore in plain view.
C 餐上桌：拋光不蝕刻 + 碳膜 + EDS，記得分析很多點、看比值。Course C served: polished, unetched + carbon + EDS. Analyse many points and plot the ratios.
水泥試片研磨拋光時，要用哪一種潤滑液？Which lubricant for grinding and polishing a cement specimen?
清水Water
少量非水性潤滑液（如乙二醇）A little non-aqueous lubricant (e.g. glycol)
完全不用潤滑液No lubricant at all
選潤滑液Choose the lubricant
對！用量要少、而且不含水。Right! Minimal, and water-free.
水會溶掉 CH，還會讓水化重新開始！Water dissolves CH and restarts hydration!
還是需要一點潤滑，只是用量要少。Some lubricant is still needed, just a minimal amount.
由粗到細研磨拋光，每一步之間都要清潔Grind and polish from coarse to fine, and clean between every step
SiC 砂紙 → 鑽石膏SiC paper → diamond paste
🧴 清潔（酒精）🧴 Clean (alcohol)
換下一步之前要先清潔！Clean before the next step!
還沒清潔！粗砂粒被帶到下一步，會在表面刮出新的刮痕。Not cleaned yet! Coarse grit carried forward scratches the surface again.
這個比較粗，會倒退回去。下一步應該是That one is coarser, a step backwards. Next should be
跳太多級了，粗刮痕磨不掉。下一步應該是Too big a jump; the coarse scratches won't come out. Next should be
洗乾淨了！All clean!
已經很乾淨了。Already clean.
鏡面拋光完成！320 → 400 → 600 → 3 → 1 → ¼ µm，每步都洗乾淨。Mirror polish! 320 → 400 → 600 → 3 → 1 → ¼ µm, cleaned at every step.
未鍍膜的方解石在充電：調整電壓，讓影像穩定再拍照An uncoated calcite crystal is charging: adjust the kV until the image is stable, then snap
📸 拍照📸 Snap
電子堆積在絕緣試片上 → 亮暗帶、掃描錯位。接近 E₂ 交越電壓時，射出的電子 ≈ 射入的電子。Electrons pile up on the insulator → bright and dark bands, scan shifts. Near the E₂ crossover, emitted ≈ incident electrons.
穩定了！約 1.5 kV 接近 E₂ 交越電壓，不再累積電荷。Stable! Around 1.5 kV, near the E₂ crossover, charge stops piling up.
還在充電：電子太多、堆在晶體上，再降低一點！Still charging: too many electrons pile up on the crystal. Go lower!
結殼接下來要做 EDS，該鍍什麼？Next the crust goes to EDS. Which coating?
要做 EDS 微區分析，該鍍什麼？Which coating for EDS microanalysis?
不鍍膜，直接 15 kVNo coat, straight to 15 kV
碳：原子序低、干擾峰少、幾乎不吸收要射出的 X 光。Carbon: low Z, few interfering peaks and almost no absorption of the outgoing X-rays.
金膜適合 SE 影像，但會加上 Au M 線、吸收低能 X 光。Gold is great for SE imaging, but adds Au M lines and absorbs low-energy X-rays.
不鍍膜在 15 kV 會劇烈充電，影像和能譜都會亂掉。Uncoated at 15 kV, it charges badly and both image and spectrum go wrong.
蒸鍍留下了陰影，怎麼辦？The evaporation left a shadow. What now?
蒸鍍的原子直線飛行（視線鍍膜），粗糙試片的背面鍍不到。Evaporated atoms travel in straight lines (line-of-sight), so the far side of a rough specimen is missed.
放在行星式載台上旋轉、傾斜Rotate and tilt on a planetary holder
鍍兩倍厚Coat twice as thick
降低真空度Lower the vacuum
轉一轉、斜一斜，每一面都鍍到了！Spin and tilt, and every side gets its share!
陰影處還是鍍不到，只是讓其他地方更厚。The shadow still gets nothing; the rest just gets thicker.
蒸鍍需要較高的真空，降低真空會讓鍍膜失敗。Evaporation needs a high vacuum; lowering it ruins the coating.
鐵相ferrite
平均原子序 Z̄：越大越亮Mean atomic number Z̄: higher is brighter
點最亮的填隙相：鐵相（ferrite）Tap the brightest interstitial phase: ferrite
點稜角分明的淺灰晶體：alite（C₃S）Tap an angular light-grey crystal: alite (C₃S)
點圓圓、有條紋的晶體：belite（C₂S）Tap a rounded, striated crystal: belite (C₂S)
點較暗的灰色填隙相：aluminate（C₃A）Tap the darker grey interstitial phase: aluminate (C₃A)
點黑色的孔隙Tap a black pore
alite 和 belite 的對比 C 大約多少？About how large is the alite–belite contrast C?
鐵相 C₄AF：Z̄ ≈ 18.9、η ≈ 0.22，熟料裡最亮！Ferrite C₄AF: Z̄ ≈ 18.9, η ≈ 0.22, the brightest phase in clinker!
Alite（C₃S）：稜角分明的淺灰色晶體，Z̄ ≈ 15.1。Alite (C₃S): angular light-grey crystals, Z̄ ≈ 15.1.
Belite（C₂S）：圓形、常有層狀條紋，Z̄ ≈ 14.6，比 alite 稍暗。Belite (C₂S): rounded, often with lamellar striations, Z̄ ≈ 14.6, slightly darker than alite.
Aluminate（C₃A）：填在縫隙裡的較暗灰色，Z̄ ≈ 14.3。倍率低時常和鐵相分不清。Aluminate (C₃A): the darker grey filling the gaps, Z̄ ≈ 14.3. At low magnification it is often hard to tell from ferrite.
孔隙：沒有東西反射電子，所以是黑的。Pore: nothing to backscatter electrons, so it is black.
這是最亮的鐵相，再找找看。That's ferrite, the brightest phase. Look again.
這是 alite：稜角分明、淺灰色。That's alite: angular and light grey.
這是 belite：圓圓的、有條紋。That's belite: rounded and striated.
這是較暗的填隙相 aluminate。That's aluminate, the darker interstitial phase.
這是黑色的孔隙。That's a black pore.
這是方鎂石 MgO（Z̄ ≈ 10.4），比較暗的小晶體。That's periclase MgO (Z̄ ≈ 10.4), a small darker crystal.
對！C ≈ (0.18 − 0.17)/0.18 ≈ 6%：看得出來，但要靠好的設定，形狀和紋理也很有幫助。Right! C ≈ (0.18 − 0.17)/0.18 ≈ 6%: visible, but only with good settings, and shape and texture help too.
45% 是 Al 和 Fe 那種差很多的組合；alite 和 belite 的 Z̄ 只差 0.5。45% is for a big pair like Al and Fe; alite and belite differ by only 0.5 in Z̄.
0.4% 是 Pt／Au 那種相鄰元素；alite 和 belite 還看得出來。0.4% is for neighbours like Pt and Au; alite and belite are still distinguishable.
在 SEI 破斷面上找出水化產物Find the hydration products on the SEI fracture
在 BEI 拋光切面上找出各相Find the phases on the BEI polished section
鈣礬石 AFt：針狀Ettringite AFt: needles
CH：大塊六角板CH: massive hexagonal plates
C-S-H：纖維狀凝膠C-S-H: fibrous gel
AFm：薄薄的板片AFm: thin plates
未水化水泥：最亮Unhydrated cement: brightest
CH：淺灰色大塊CH: light-grey masses
內部產物 C-S-H：顆粒外圍的環Inner-product C-S-H: the rim around a grain
孔隙與裂縫：黑色Pores and cracks: black
點一束針狀的鈣礬石（AFt）Tap a bundle of ettringite (AFt) needles
點大塊的六角板 CHTap a massive hexagonal CH plate
點纖維狀的 C-S-H 凝膠Tap the fibrous C-S-H gel
點薄板狀的 AFmTap the thin AFm plates
點最亮的未水化水泥顆粒Tap the brightest unhydrated cement grain
點淺灰色的 CHTap the light-grey CH
點顆粒外圍的內部產物 C-S-H 環Tap the inner-product C-S-H rim around a grain
點黑色的孔隙或裂縫Tap a black pore or crack
BSE 看得到哪一種孔隙？Which pores can BSE resolve?
鈣礬石（三硫型硫鋁酸鈣水化物）：細長的針！Ettringite (calcium trisulfoaluminate hydrate): long, thin needles!
氫氧化鈣：大塊的六角板，在孔隙空間裡可以長到數十 µm。Calcium hydroxide: massive hexagonal plates, tens of µm in open space.
C-S-H：纖維狀、箔狀的凝膠，包住整片背景。C-S-H: fibrous, foil-like gel covering the background.
AFm（單硫型）：薄薄的六角板片，常排成玫瑰花狀。AFm (monosulfate): thin hexagonal plates, often in rosettes.
那是一束細針：鈣礬石 AFt。That's a bundle of thin needles: ettringite AFt.
那是大塊的六角板：CH。That's a massive hexagonal plate: CH.
那是纖維狀的 C-S-H 凝膠。That's fibrous C-S-H gel.
那是薄板片排成的 AFm。That's AFm, thin plates in a rosette.
換成拋光切面的 BEI：SEI 漂亮但只沿著弱面破開，BEI 才能公平地量比例。Now the polished BEI: SEI is beautiful but follows weak paths; BEI cuts every phase fairly for quantification.
未水化的水泥顆粒最亮。Unhydrated cement grains are the brightest.
CH：淺灰色的大塊，是卜作嵐反應會吃掉、也最先被水溶出的相。CH: light-grey masses. Pozzolanic reactions consume it, and water leaches it first.
內部產物（緻密的 C-S-H）長在原本顆粒的邊界之內；外部產物則填在顆粒之間。Inner product (dense C-S-H) forms within the original grain boundary; outer product fills the space between grains.
灌了環氧樹脂的孔隙和裂縫是黑色的。Epoxy-filled pores and cracks are black.
那是最亮的未水化顆粒。That's a bright unhydrated grain.
那是淺灰色的 CH。That's light-grey CH.
那是中灰色的 C-S-H 基質。That's the mid-grey C-S-H groundmass.
那是顆粒外圍的內部產物環。That's the inner-product rim around a grain.
那是黑色的孔隙或裂縫。That's a black pore or crack.
BSE 約在微米尺度工作。哪一種孔隙看得到？BSE works at the micrometre scale. Which pores can it see?
MIP 有「墨水瓶效應」，會把大孔算成入口的大小MIP suffers from the ink-bottle effect, assigning large pores to their small entrances
較粗的毛細孔隙（µm 級）Coarse capillary pores (µm scale)
膠體孔隙（≲ 10 nm）Gel pores (≲ 10 nm)
兩種都很容易Both, easily
對！BSE 看得到較粗的毛細孔，膠體孔太小。BSE 影像分析和 MIP 是互補的。Right! BSE resolves coarse capillary pores; gel pores are far too small. BSE image analysis and MIP complement each other.
膠體孔在 C-S-H 裡，小於約 10 nm，SEM 解析不到。Gel pores sit inside C-S-H, below about 10 nm, beyond SEM resolution.
膠體孔太小了，BSE 只看得到較粗的毛細孔。Gel pores are too small; BSE sees only the coarser capillary pores.
答對就發射彩虹光束！Right answers fire the rainbow beam!
依照時間順序點水化的各階段Tap the hydration stages in time order
這個描述屬於五指中的哪一指？Which of the five descriptors is this?
水泥顆粒的水化時間軸Hydration timeline of a cement grain
外殼慢慢長大，顆粒在裡面慢慢消失The shell grows and the grain slowly disappears inside it
約 10 分鐘：表面出現非晶質凝膠與短的 AFt 棒~10 min: amorphous gel and short AFt rods on the surface
約 10 小時：C₃S 反應，外部 C-S-H 長在 AFt 網上~10 h: C₃S reacts; outer C-S-H grows on the AFt network
約 18 小時：長 AFt 棒，內部 C-S-H 開始生成~18 h: long AFt rods; inner C-S-H starts
1–3 天：C₃A 與 AFt 反應成六角板 AFm1–3 d: C₃A reacts with AFt → hexagonal AFm plates
約 14 天：內部 C-S-H 填滿顆粒與外殼間的空隙~14 d: inner C-S-H fills the gap between grain and shell
尺寸、形狀、方向、共生關係、成分：這句話在描述哪一項？Size, shape, orientation, association, composition: which one does this describe?
未水化核心約 5–30 µmUnhydrated cores of 5–30 µm
鈣礬石是針狀、CH 是板狀Ettringite is needle-like; CH is platy
CH 在骨材界面區（ITZ）有方向性排列CH is aligned at the aggregate ITZ
內部產物環繞在 alite 外圍；CH 長在原本的水空間Inner product rims around alite; CH in former water space
EDS 量得 C-S-H 的 Ca/Si ≈ 1.7EDS gives C-S-H Ca/Si ≈ 1.7
尺寸Size
形狀Shape
方向Orientation
共生關係Association
成分Composition
時間軸排好了，內部產物把外殼填滿！Timeline complete: the inner product fills the shell!
順序不對，想想哪個先發生。Not in order. Think about what happens first.
提示：C₃A 和石膏最先反應（幾分鐘），C₃S 要等幾小時。Hint: C₃A and sulfate react first (minutes); C₃S takes hours.
五根手指全用上了，描述別人也能檢查！All five fingers used: a description anyone can check!
再想想：它說的是大小、樣子、排列、旁邊是誰，還是成分？Think again: is it about size, look, alignment, neighbours or chemistry?
提示：「旁邊是誰」就是共生關係。Hint: "what is next to what" is association.
對！Right!
走到發光的平台開始關卡，沿路會種下花朵Walk to a glowing pad to start a stage; you plant flowers along the way
移動：方向鍵、搖桿或點草地 · 確定：空白鍵／AMove: arrow keys, controller or tap the lawn · Confirm: Space / A
已帶入 World 2 的夥伴與花瓣！Your World 2 friends and petals came along!
已帶入 World 1 的夥伴與花瓣！Your World 1 friends and petals came along!
叮！燈塔的光束把我們帶到這裡：「水泥花園」，三個世界的最後一站！Ding! The lighthouse beam led us here: the Cement Garden, the last stop of all three worlds!
World 1 和 World 2 的夥伴全都來了，花瓣也帶著！咦……噴泉在漏水？All our friends from Worlds 1 and 2 are here, with our petals! Huh... is the fountain leaking?
歡迎……我是花園管家凝凝，一團 C-S-H 凝膠。許願噴泉的池壁長出白色結殼，靠近水面的漿體也變得又軟又多孔。Welcome... I'm Gelly, the garden keeper, a puff of C-S-H gel. White crusts are growing on the Wishing Fountain, and the paste near the water has gone soft and porous.
成分就交給我們吧！這次要確切知道它是由什麼組成的。Leave the composition to us! This time we'll find out exactly what it's made of.
滴答、滴答！我是溶溶，一滴一滴把鈣帶出漿體。抓得到我再說吧！Drip, drop! I'm Drippy, and I carry calcium out of the paste one drop at a time. Catch me if you can!
最後三個本領：X 光分析、試片製備、水泥微結構判讀。學會了，SEM 就再也沒有秘密！Three last skills: X-ray analysis, specimen preparation and reading cement microstructure. Master them and the SEM holds no more secrets!
走到 3-1 的平台出發吧！Let's head to pad 3-1!
叮！燈塔的光束把我們帶到這裡：「水泥花園」，三個世界的最後一站！Ding! The lighthouse beam led us here: the Cement Garden, the last stop of all three worlds!
這台裝置上沒有之前的存檔，不過夥伴們還是都來了！咦……噴泉在漏水？There's no earlier save on this device, but our friends came along anyway! Huh... is the fountain leaking?
歡迎……我是花園管家凝凝，一團 C-S-H 凝膠。許願噴泉的池壁長出白色結殼，靠近水面的漿體也變得又軟又多孔。Welcome... I'm Gelly, the garden keeper, a puff of C-S-H gel. White crusts are growing on the Wishing Fountain, and the paste near the water has gone soft and porous.
滴答、滴答！我是溶溶，一滴一滴把鈣帶出漿體。抓得到我再說吧！Drip, drop! I'm Drippy, and I carry calcium out of the paste one drop at a time. Catch me if you can!
最後三個本領：X 光分析、試片製備、水泥微結構判讀。學會了，SEM 就再也沒有秘密！Three last skills: X-ray analysis, specimen preparation and reading cement microstructure. Master them and the SEM holds no more secrets!
走到 3-1 的平台出發吧！Let's head to pad 3-1!
先定性、再定量！先問「有什麼」，再問「有多少」。漏掉或認錯一個元素，後面所有數字都會錯。Qualitative first, then quantitative! First "what is there?", then "how much?". Miss or misidentify one element and every number after it is wrong.
我來當弓箭手：先射最大、能量最高的峰，再檢查它整個家族。K 家族裡 Kα : Kβ ≈ 10 : 1。I'll be the archer: shoot the largest, highest-energy peak first, then check its whole family. In the K family, Kα : Kβ ≈ 10 : 1.
EDS 一次看到整道彩虹，只是有點模糊（約 130 eV）；WDS 一次只看一種顏色，清晰無比（約 10 eV）。EDS sees the whole rainbow at once, a bit blurry (about 130 eV); WDS looks at one colour at a time, razor sharp (about 10 eV).
我是鐵鐵！定量時先算強度比 k = I／I（標準品）。可是基材會改變有多少 X 光跑得出來……I'm Ferri! Quantification starts with the k-ratio, k = I / I(standard). But the matrix changes how many X-rays get out...
Z：原子序，影響背散射與阻擋能力；A：X 光出來時被吸收，通常最大；F：被別的元素的 X 光激發螢光，通常最小。Z: atomic number, through backscattering and stopping power. A: absorption on the way out, usually the largest. F: fluorescence excited by other elements' X-rays, usually the smallest.
這個 Fe–Ni 蹺蹺板歪了，找出真的會發生的校正！This Fe–Ni seesaw is tipped. Find the corrections that really happen!
我是台台，SEM 的試片載台！同樣的食材，不同的菜要用不同的做法：形貌、微結構形態、微區分析。I'm Stubby, an SEM specimen stub! Same ingredient, different recipe for each dish: topography, morphology or microanalysis.
形貌：保留原表面，用 SE。形態：切割、灌環氧樹脂、拋光，用 BSE。微區分析：拋光不蝕刻、鍍碳，用 EDS。Topography: keep the surface, use SE. Morphology: cut, epoxy, polish, use BSE. Microanalysis: polished, unetched, carbon coat, EDS.
啪滋！我是啪啪，堆在沒鍍膜的試片上的那一大堆電子！Bzzt! I'm Sparky, all the electrons piled up on an uncoated specimen!
先做鏡面拋光：由粗到細，少量非水性潤滑液，每一步之間都要清潔。First a mirror polish: coarse to fine, a little non-aqueous lubricant, and clean between every step.
再來讓我冷靜：鍍膜、接導電路徑，或把電壓降到 E₂ 交越電壓附近。Then calm me down: coat, add a conductive path, or lower the kV toward the E₂ crossover.
我是稜稜，alite（C₃S），稜角分明！I'm Ally, alite (C₃S), all sharp angles!
我是圓圓，belite（C₂S），圓圓的、身上有條紋。I'm Bello, belite (C₂S), round and striped.
我是最亮的鐵相！還記得 World 2 嗎？平均原子序越大，BSE 越亮。熟料就是一幅石頭馬賽克！And I'm ferrite, the brightest! Remember World 2? Higher mean Z, brighter BSE. Clinker is a stone mosaic!
針！針！我是針針，鈣礬石 AFt。Needles, needles! I'm Etty, ettringite (AFt).
我是板板，氫氧化鈣 CH，六角柱。溶溶一直在溶解我……I'm Porty, calcium hydroxide (CH), a hexagonal pillar. Drippy keeps dissolving me...
SEI 在破斷面上看形狀很漂亮；拋光切面的 BEI 才能公平地告訴我們比例。SEI shows our shapes beautifully on fracture surfaces; BEI on a polished section tells the true proportions.
滴答、滴答！水流過孔隙，把 CH 一點一點洗出去，你們的噴泉就要垮了！Drip, drop! Water runs through the pores and washes the CH out bit by bit. Your fountain will crumble!
大家一起發射彩虹光束！每答對一題，就封住一個孔隙！Everyone, fire the rainbow beam! Every right answer seals a pore!
你們一定記不得水泥顆粒是怎麼水化的！Bet you can't remember how a cement grain hydrates!
簡單！看我的外殼長大吧。Easy! Watch my shell grow.
那你們描述看看這片微結構啊！Then describe this microstructure, if you can!
五根手指：尺寸、形狀、方向、共生關係、成分！Five fingers: size, shape, orientation, association, composition!
嗚……我只是想找個地方流動而已……Sniff... I just wanted somewhere to flow...
那就從表面流過去，不要穿過漿體！漿體夠緻密，水就會留在外面。Then flow over the surface, not through the paste! When the paste is dense, water stays outside.
噴泉得救了！鈣被溶出、方解石結殼之謎解開了。The fountain is saved! Calcium leached, calcite crust explained.
從電子束之塔、對比海灣到水泥花園，三個世界全部完成，你已經是 SEM 大師了！From the Electron Beam Tower and Contrast Cove to the Cement Garden: all three worlds complete. You're an SEM master now!
大家一起跳最終綻放之舞！嗨！嗨！嗨！Everyone, the final Bloom Dance! Hi! Hi! Hi!
和 EDS 相比，WDS 的特點是？Compared with EDS, WDS…
能量解析度好得多（約 10 eV），但一次只分析一種元素Has much better resolution (about 10 eV) but measures one element at a time
速度更快，一次收集整個能譜Is faster and collects the whole spectrum at once
不需要繞射晶體Needs no diffracting crystal
WDS 用晶體依布拉格定律一次繞射一個波長，能分開重疊的譜線，適合精確定量與微量元素；EDS 快、一次全譜，但約 130–155 eV 的解析度讓低於 3 keV 的譜線擠在一起。（PCAM.03 s.6）WDS diffracts one wavelength at a time (Bragg's law), separating overlapping lines; it suits accurate quantification and trace elements. EDS is fast and takes the whole spectrum, but at ~130–155 eV the lines below ~3 keV crowd together. (PCAM.03 s.6)
提示：彩虹 vs 望遠鏡。Hint: the rainbow vs the telescope.
ZAF 三種校正中，通常哪個最大、哪個最小？Of the three ZAF corrections, which is usually largest and which smallest?
吸收 A 最大，螢光 F 最小Absorption (A) largest, fluorescence (F) smallest
螢光 F 最大，原子序 Z 最小Fluorescence (F) largest, atomic number (Z) smallest
三者總是一樣大They are always equal
內部產生的 X 光在出來的路上被部分吸收，通常是最大的校正；別的元素的 X 光額外激發元素 i 的螢光，通常最小。校正需要知道全部成分（包括以化學計量算出的 O）。（s.9–11）X-rays generated inside are partly absorbed on the way out, usually the largest correction; fluorescence of element i by other elements' X-rays is usually the smallest. The correction needs the full composition, including O by stoichiometry. (s.9–11)
提示：梅塔騎士說「什麼拖住你」最重。Hint: "what holds you back" weighs the most.
為什麼蝕刻過的試片不能拿來做 EDS 定量？Why can't an etched specimen be used for quantitative EDS?
表面不再平整，成分也不再具代表性The surface is no longer flat or chemically representative
蝕刻會讓試片導電Etching makes the specimen conductive
蝕刻會去除所有孔隙Etching removes all the pores
蝕刻能讓某一相的紋理更明顯（形態觀察），但定量需要平整、拋光、不蝕刻的表面與標準品。（s.17–18）Etching reveals one phase nicely for morphology, but quantification needs a flat, polished, unetched surface and standards. (s.17–18)
提示：蝕刻過的劍很漂亮，但量不出真正的成分。Hint: an etched blade is beautiful, but you can't measure its true metal.
試片充電時，哪一組做法能改善？Which set of actions cures charging?
鍍膜、接導電路徑、降低電壓或電流、加快掃描、或用低真空模式Coat, add a conductive path, lower the kV or current, scan faster, or use low-vacuum mode
提高電壓並放慢掃描Raise the kV and slow the scan
加大電子束電流Increase the beam current
未鍍膜的方解石在 15 kV 出現亮暗帶與掃描錯位；在 1.5 kV（接近 E₂ 交越電壓，射出 ≈ 射入）就穩定了。（s.25）Uncoated calcite at 15 kV shows bright and dark bands and scan breaks; at 1.5 kV, near the E₂ crossover where emitted ≈ incident electrons, the image is stable. (s.25)
提示：讓啪啪身上的電子少一點、跑得掉。Hint: fewer electrons on Sparky, and a way out for them.
在 BEI 拋光切面中，大多數未水化水泥顆粒是？In a polished BEI section, most anhydrous cement grains are…
多礦物顆粒：一顆裡有好幾種相Polymineralic: several phases in one grain
單一的 alite 晶體Single crystals of alite
純的鐵相Pure ferrite
水泥顆粒大小差很多，且多半是多礦物組成；倍率低時 C₃A 與鐵相常分不清。SEM 的定量相結果也不一定和 Bogue 計算一致。（s.29、31）Grains span a wide size range and most are polymineralic; at low magnification C₃A and ferrite are often hard to tell apart. SEM phase results also need not match the Bogue calculation. (s.29, 31)
提示：水泥是一群人，不是一個人。Hint: cement is a crowd, not a single guy.
要做定量影像分析，為什麼用拋光切面的 BEI，而不是破斷面的 SEI？For quantitative image analysis, why use BEI of a polished section rather than SEI of a fracture?
拋光切面是隨機切面，各相以真實比例出現；破斷面沿弱面破開，沒有代表性A polished section is a random cut showing phases in true proportion; a fracture follows weak paths
SEI 看不到 CHSEI cannot show CH
破斷面一定是平的Fracture surfaces are always flat
SEI 很適合看形貌，但只在沿弱面破開的表面上；BEI 拋光切面適合 World 2 學到的影像量化。（s.32–33）SEI shows morphology well, but only on surfaces that broke along weak paths; polished BEI suits the image quantification learned in World 2. (s.32–33)
提示：拋光切面在比例上從不說謊。Hint: a polished slice tells no lies about proportions.
K 家族中 Kα : Kβ 的強度比大約是？Within the K family, Kα : Kβ is about…
10 : 110 : 1
1 : 11 : 1
1 : 101 : 10
候選元素要連次要譜線都以正確比例出現才算確認。L 家族：Lα 1、Lβ₁ 0.7、Lβ₂ 0.2……（s.6–8）A candidate is confirmed only if its minor lines appear in the right ratio. L family: Lα 1, Lβ₁ 0.7, Lβ₂ 0.2… (s.6–8)
提示：一杯 Kα 配一小匙 Kβ。Hint: one cup of Kα to a spoon of Kβ.
微區分析為什麼鍍碳而不鍍金？Why carbon rather than gold for microanalysis?
碳原子序低：干擾峰少，也幾乎不吸收射出的 X 光Carbon's low Z: few interfering peaks and little absorption of outgoing X-rays
碳比較閃亮Carbon is shinier
金不導電Gold does not conduct
金會加上 Au M 線並吸收低能 X 光；碳蒸鍍是微區分析的標準鍍膜。（s.18、22）Gold adds Au M lines and absorbs low-energy X-rays; carbon evaporation is the standard coating for microanalysis. (s.18, 22)
提示：只撒一層薄薄的碳粉，別抹金色糖霜。Hint: a thin sprinkle of carbon, no gold frosting.
要看鈣礬石針狀晶的形貌，該用哪種表面？To see the shape of ettringite needles, which surface?
新鮮破斷面，不拋光A fresh fracture, not polished
拋光再蝕刻Polished and etched
拋光不蝕刻Polished, unetched
形貌試片：清潔、黏到載台、乾燥、接導電路徑或鍍膜，用 SE 偵測器；千萬別把要看的凹凸磨掉。（s.16）Topography: clean, attach to the stub, dry, add a conductive path or coat, and use the SE detector; never polish away the bumps you came to see. (s.16)
提示：台台的 A 餐。Hint: Stubby's course A.
水泥試片研磨拋光為什麼用非水性潤滑液？Why use a non-aqueous lubricant when polishing cement?
水會溶掉 CH，還會讓水化重新開始Water dissolves CH and restarts hydration
水太貴Water is too expensive
水會讓漿體變硬Water makes the paste harder
還要用低絨毛的拋光布減少浮凸，並在每一步之間用乙二醇或酒精清潔。（s.20）Also use a low-nap cloth to limit relief, and clean with glycol or alcohol between steps. (s.20)
提示：別讓 CH 洗澡。Hint: don't give the CH a bath.
熟料在 BSE 影像中最亮的相是？The brightest clinker phase in BSE is…
鐵相 C₄AF（Z̄ ≈ 18.9）Ferrite C₄AF (Z̄ ≈ 18.9)
Belite C₂SBelite C₂S
方鎂石 MgOPericlase MgO
亮度順序：鐵相 > alite（15.1）> belite（14.6）> aluminate（14.3）> MgO（10.4）。（s.30）Brightness: ferrite > alite (15.1) > belite (14.6) > aluminate (14.3) > MgO (10.4). (s.30)
提示：鐵鐵會發光。Hint: Ferri glows.
「內部產物」C-S-H 指的是？"Inner product" C-S-H is…
在原本顆粒邊界之內形成的緻密 C-S-HDense C-S-H formed within the original grain boundary
長在孔隙裡的 C-S-HC-S-H growing in pores
一種 CH 晶體A kind of CH crystal
外部產物是顆粒之間、皺摺薄箔狀的 C-S-H；內部產物較緻密，奈米壓痕也能分辨出不同的力學性質。（s.35）Outer product is crumpled thin-foil C-S-H between grains; inner product is denser, and nanoindentation can tell their mechanical properties apart. (s.35)
提示：舊城牆內，防禦緻密而沉靜。Hint: inside the old walls, the defence is dense and quiet.
哪一相會被卜作嵐反應消耗，也是蓄水結構中最先被溶出的？Which phase is consumed by pozzolanic reactions and leached first in water-retaining structures?
CH（氫氧化鈣）CH (portlandite)
鐵相Ferrite
AliteAlite
例如 LC3 裡的煅燒黏土會和 CH 反應；被溶出的鈣遇到 CO₂ 就形成方解石結殼。（s.33）Calcined clay in LC3, for example, reacts with CH; leached calcium meets CO₂ and forms a calcite crust. (s.33)
提示：溶溶一直在溶解誰？Hint: who does Drippy keep dissolving?
把水泥試片用高溫烘箱烘乾，會怎樣？Oven-drying cement specimens at high temperature…
會破壞鈣礬石與 C-S-H；應改用溶劑置換或冷凍乾燥Damages ettringite and C-S-H; use solvent exchange or freeze-drying instead
是終止水化最好的方法Is the best way to stop hydration
完全沒有影響Has no effect at all
每篇論文都要寫清楚乾燥與灌注方法，因為它會改變孔隙、裂縫，甚至相組成。（s.26）Report the drying and impregnation method in every thesis: it changes porosity, cracking and even the phase assemblage. (s.26)
提示：烘得太快，蛋糕就裂了。Hint: dry it too fast and the cake cracks.
為什麼單靠 MIP 常得不到真正的孔徑？Why does MIP alone often miss the true pore size?
墨水瓶效應：大孔被算成它小入口的大小The ink-bottle effect: large pores are counted at their small entrance size
MIP 只看得到膠體孔MIP only sees gel pores
水銀會溶解 CHMercury dissolves CH
所以 BSE 影像分析與 MIP 是互補的：BSE 看得到較粗的毛細孔真正的大小。（s.36）So BSE image analysis and MIP are complementary: BSE shows the true size of the coarser capillary pores. (s.36)
提示：大洞穴，小門口。Hint: a big cave with a small doorway.
定性 vs 定量 · EDS vs WDS · 辨識能峰與譜線家族 · 結殼 EDS 挑戰（PCAM.03 s.5–9、14）Qualitative vs quantitative · EDS vs WDS · identifying peaks and line families · crust EDS challenge (PCAM.03 s.5–9, 14)
k 比值 · Z、A、F 基材校正 · Fe–Ni 實例 · 分析總和與水泥 EDS 實務（s.10–13）k-ratio · Z, A, F matrix corrections · the Fe–Ni example · analytical totals and cement EDS practice (s.10–13)
形貌、微結構形態、微區分析三類試片的製備（s.15–19）Preparing specimens for topography, morphology and microanalysis (s.15–19)
切割與拋光 · 潤滑液 · 濺鍍與蒸鍍 · 假象與充電（s.20–26）Cutting and polishing · lubricant · sputter vs evaporative coating · artefacts and charging (s.20–26)
熟料相的形狀與 BSE 灰階 · Z̄ 與對比 · 未水化水泥顆粒（s.29–31）Clinker phase shapes and BSE grey levels · Z̄ and contrast · anhydrous grains (s.29–31)
SEI 與 BEI 中的水化產物 · CH · C-S-H 型態與內／外部產物 · 孔隙（s.32–37）Hydration products in SEI and BEI · CH · C-S-H types and inner/outer product · pores (s.32–37)
總複習 + 水化時間軸 + 五指描述法（s.38–40）Review + hydration timeline + the five descriptors (s.38–40)
交互作用體積精靈，三個世界的主角。The interaction-volume sprite, hero of all three worlds.
特性 X 光，這次化身弓箭手，一箭一個能峰。A characteristic X-ray, now an archer: one arrow, one peak.
花園管家，一團 C-S-H 凝膠。纖維狀、箔狀，什麼形狀都有。The garden keeper, a puff of C-S-H gel: fibrous, foil-like, every shape at once.
方解石結殼。被溶出的 Ca(OH)₂ 遇到 CO₂ 就變成我。The calcite crust. Leached Ca(OH)₂ meets CO₂ and becomes me.
鐵相 C₄AF，熟料最亮的相，也是 ZAF 蹺蹺板上的 Fe。Ferrite C₄AF, the brightest clinker phase, and the Fe on the ZAF seesaw.
SEM 試片載台。碳膠帶黏牢、銀膠接地，絕不漂移。An SEM stub. Carbon tape holds tight and silver paint grounds the sample: no drift.
充電效應。鍍膜、接地或降到 E₂ 附近，牠就安靜了。Charging. Coat, ground or go near E₂, and Sparky calms down.
Alite（C₃S）：稜角分明的淺灰色晶體。Alite (C₃S): angular light-grey crystals.
Belite（C₂S）：圓圓的、有層狀條紋。Belite (C₂S): rounded, with lamellar stripes.
鈣礬石 AFt：細長的針，早期水化的主角之一。Ettringite AFt: long, thin needles, a star of early hydration.
氫氧化鈣 CH：六角板狀，會被卜作嵐反應吃掉、被水溶出。Calcium hydroxide CH: hexagonal plates, eaten by pozzolans and leached by water.
改邪歸正的溶溶，現在只從表面流過。The reformed Drippy, who now flows only over the surface.
📮 1 · EDS 定性分析Qualitative EDS
Kα : Kβ ≈ 10 : 1 · Lα 1, Lβ₁ 0.7, Lβ₂ 0.2 · Mα 1, Mβ 0.6
先從最大、能量最高的峰開始，再用整個家族確認，標記完再找下一座。小心重疊（S Kα／Mo Lα）與逃逸峰、和峰。EDS 約 130–155 eV、一次全譜；WDS 約 10 eV、一次一個元素。Start with the largest, highest-energy peak, confirm with its whole family, mark it, then move on. Beware overlaps (S Kα / Mo Lα), escape and sum peaks. EDS: ~130–155 eV, whole spectrum; WDS: ~10 eV, one element at a time.
PCAM.03 s.5–9
📮 2 · 定量與 ZAFQuantification & ZAF
kᵢ = Iᵢ / I₍ᵢ₎ · Cᵢ / C₍ᵢ₎ = [ZAF]ᵢ · kᵢ
前提：平整拋光、垂直電子束、X 光體積內均勻，且知道全部元素。總和應約 98–102%。Fe–Ni：Ni 被 Fe 吸收（A）、Fe 被激發螢光（F）。水泥：10–15 kV、鍍碳、多點分析，用 Ca/Si、Al/Ca 等原子比。Assumes a flat, polished surface normal to the beam, homogeneous over the X-ray volume, with all elements known. Totals should be ~98–102%. Fe–Ni: Ni absorbed by Fe (A), Fe fluoresced (F). Cement: 10–15 kV, carbon coat, many points, atomic ratios such as Ca/Si and Al/Ca.
PCAM.03 s.9–13
📮 3 · 依目的選試片Specimens by goal
Topography → SE · Morphology → BSE · Microanalysis → EDS/WDS
形貌：破斷面、粒子，保留原表面，濺鍍。形態：低速鑽石切割、環氧樹脂、研磨拋光，可蝕刻（但不能定量）。微區分析：拋光不蝕刻、含標準品、鍍碳。Topography: fractures and particles, keep the surface, sputter coat. Morphology: low-speed diamond saw, epoxy, grind and polish, optional etch (no quantification). Microanalysis: polished, unetched, standards included, carbon coat.
PCAM.03 s.15–18
📮 4 · 拋光、鍍膜與充電Polish, coat & charging
SiC 320 → 400 → 600 → 3 → 1 → ¼ µm
少量非水性潤滑液、低絨毛布、每步清潔。濺鍍：多方向、包覆凹凸，適合 SE。蒸鍍：直線、會有陰影，需旋轉傾斜；碳蒸鍍是微區分析標準。充電：鍍膜、導電路徑、降 kV／電流、快掃描、低真空。Minimal non-aqueous lubricant, low-nap cloth, clean between steps. Sputtering: many directions, covers rough surfaces, good for SE. Evaporation: line-of-sight, shadows, so rotate and tilt; carbon is the microanalysis standard. Charging: coat, conductive path, lower kV/current, faster scans, low vacuum.
PCAM.03 s.20–26
📮 5 · 熟料馬賽克Clinker mosaic
Z̄: C₄AF 18.9 > C₃S 15.1 > C₂S 14.6 > C₃A 14.3 > MgO 10.4
Alite 稜角分明、belite 圓形有條紋、鐵相是最亮的填隙相、aluminate 是較暗的填隙相。Alite／belite 對比約 6%：看得出來，但要好設定，再加上形狀與紋理判斷。Alite angular, belite rounded and striated, ferrite the brightest interstitial phase, aluminate the darker one. Alite/belite contrast is about 6%: visible with good settings, helped by shape and texture.
PCAM.03 s.29–31
📮 6 · 水化漿體Hydrated paste
SEI: AFt needles · AFm plates · CH hexagons · C-S-H fibres
BEI：最亮是未水化顆粒、淺灰 CH、中灰 C-S-H 基質、黑色孔隙與裂縫。內部產物在原顆粒邊界內、較緻密；外部產物在顆粒之間。膠體孔 ≲ 10 nm 看不到，毛細孔 10 nm–10 µm 中較粗的 BSE 看得到。BEI: brightest unhydrated grains, light-grey CH, mid-grey C-S-H groundmass, black pores and cracks. Inner product lies within the old grain boundary and is denser; outer product fills the space between. Gel pores (≲ 10 nm) are invisible; coarser capillary pores (10 nm–10 µm) are resolved by BSE.
PCAM.03 s.32–37
📮 7 · 描述微結構Describing a microstructure
Size · Shape · Orientation · Association · Composition
顆粒相（phenograins）、基質（groundmass）、孔隙。水化：10 分鐘 AFt 短棒 → 10 小時外部 C-S-H → 18 小時內部 C-S-H 開始 → 1–3 天 AFm → 14 天內部產物填滿。描述要附比例尺與倍率，讓別人能檢查。Phenograins, groundmass and pores. Hydration: 10 min short AFt rods → 10 h outer C-S-H → 18 h inner C-S-H starts → 1–3 d AFm → 14 d inner product fills the gap. Always give scale bars and magnification so others can check.
PCAM.03 s.37–40
綻放早晨 · PopBloom Morning · Pop
慵懶草原 · Lo-fi 輕鬆慵懶Lazy Meadow · Lo-fi (laid-back)
廚房律動 · Funk 節奏明快Kitchen Groove · Funk (upbeat)
寶石咖啡館 · Bossa nova 慵懶Gem Café · Bossa nova (laid-back)
像素星空 · 8-bit 晶片音樂Pixel Stars · 8-bit chiptune
果凍漂流 · SynthwaveJelly Drift · Synthwave
電子衝刺 · Rock 節奏明快Electron Rush · Rock (upbeat)
糖霜千層 · Bubblegum PopSugar Layers · Bubblegum pop
溶溶大決戰 · Hard rockDrippy Showdown · Hard rock
綻放之舞 · DiscoBloom Dance · Disco
明信片圓舞曲 · WaltzPostcard Waltz · Waltz
海灣衝浪 · Surf rock 節奏明快Cove Surf · Surf rock (upbeat)
鋼鼓沙灘 · Calypso 輕鬆慵懶Steel-drum Beach · Calypso (laid-back)
午後微風 · Reggae 輕鬆慵懶Afternoon Breeze · Reggae (laid-back)
花園進行曲 · Folk pop 輕快Garden March · Folk pop (upbeat)
最終樂章 · Anthem rockGrand Finale · Anthem rock
No comments:
Post a Comment