A dissertation presented in partial fulfillment of the requirements for the degree of Doctor of Philosophy

全文

(1)國立中山大學材料科學研究所博士論文. 多元鎂基非晶質金屬玻璃熱機性質與熱可塑功能性之研究 Study on the thermomechanical properties and workability of Mg-based bulk metallic glasses. 研究生：張育誠撰指導教授：黃志青博士. 中華民國九十七年七月.

(2) Study on the thermomechanical properties and workability of Mg-based bulk metallic glasses. A dissertation presented in partial fulfillment of the requirements for the degree of Doctor of Philosophy. by. Yu-Chen Chang. This work was conducted under the supervision of. Prof. Chih-Ching Huang. Institute of Materials Science and Engineering National Sun Yat-Sen University. July, 2008.

(3)

(4) 作者簡介及謝誌. 本人張育誠生於民國 67 年 9 月 14 日，高雄市人，民國 92 年進入國立中山大學材料科學與工程研究所就讀，於 96 學年完成博士學位。求學期間品學兼優，個性樂觀開朗，為人圓滑處事合宜，和朋友同儕相敬如賓，與師長成為忘年之交，亦師亦友互增所長，對於未來更充滿信心與幹勁，期待更多元挑戰。本論文之研究及撰寫，特蒙恩師黃教授志青循循善誘、誨人不倦及去惑解疑，學生誠摯感恩，對於恩師之研究精神與為人風範，深表敬佩景仰，令我如沐春風、芝蘭之化，望與恩師日浹往來，增益學生所不能。承蒙義守大學鄭教授憲清古道熱腸、鼎力相助，令本研究更添色彩。且蒙跨海相助之杜博士興蒿，令我研究更加繽紛多元、精益求精。感謝所上諸位師長們指點提攜，更衷心地感謝所上高伯威教授、謝克昌教授、陳明教授、機械與機電工程學系潘正堂教授與義守大學材料科學與工程學系鄭憲清教授不辭辛勞擔任口試委員，惠予指導，並對於本論文提出卓越見解與指正。此外，感謝國科會經費支持，並蒙中山大學電子顯微鏡技術員陳貴香與王良珠小姐、中山大學奈米中心技術員顏采蓉小姐及顏仲崑先生辛勞協助樣品測試；義守大學材料科學與工程學系鄭憲清教授與其學生們與博精儀器股份有限公司之原料及測試儀器資料方面提供，除深表感謝外，誠心祝福順心如意。深受黃谷正教官照顧與幫忙，令本人可專心論文研究，特此感謝。所辦朱惠敏、顏秀芳和陳秀玉小姐於行政公事與日常生活上之挺立拔刀相助，深表恩謝。華應麒大哥的社會大學與處事態度更令小弟我受益豐碩，一日大哥終身大哥，大哥之名永銘於心。研究期間，承蒙學長高立衡教授教誨良殷、樂育英才，令我學以致用志向得展，感恩言謝之心沒齒難忘。鄭舜宇博士耿直率性之個性與精益求精之求學精神令人佩服，小弟求學期間深受呵護照顧，真是感激不盡。博士班學長姐胡義群教授、鄭國光博士、洪英博博士、吳玉娟博士、楊凱琳博士和蔣偉仁博士之熱心教導指教，令我事半功倍百尺竿頭。.

(5) 同時感謝張量然博士、丁仕旋博士、黃惠君博士和呂旻憲博士等研究所博士班同儕好友們的支持、鼓勵及協助，令我不顯孤單寂寞。尤其無話不談的同儕知己曾婉如博士是我精神心靈的良師、休憩娛樂的伙伴，令我抒解求學時的壓力與心靈上的困境，更加讓我易於融入中山材料所的群體裡；互通心靈的同儕知己謝心心博士，妳的深切關懷、恩情鼓舞，令本人通過考驗、克服難關，完成學業，沒有妳的伴陪，更沒有此時此刻的我；好友李敬仁博士則令我瞭解當研究工作者該具備的能力與責任，並鼓勵我面對挑戰與困難，突破瓶頸與阻礙，其專注研究工作之態度更可作為研究生典範，值得好好效法；好友洪子翔博士賦予我新的研究生命，令我研究之心死灰復燃，其懇切耐心之指導、謹慎嚴謹之態度，令我獲益良多，並帶領著我朝向成功的彼端；好友羅友杰博士直言不諱、擇善固執的個性，令本人敬佩不已，並將其視為一面明鏡，來檢討改正自身缺失；好友張志溢博士的刻苦耐勞與堅毅不覺也令在下稱服，作為學習效法之範。此外，研究室陳海明、賴炎暉、鄭宇庭、周鴻昇、裴浩然、劉名哲及郭哲男等博士班學弟們勤勉不倦、不眠不休用心向學，還常抽空幫忙在下，令人銘感於心。對於湯振緯、宋大豪及孫碩陽學弟們努力不懈、實事求是之學習態度與熱誠幫助之行為，也令我感動不已，在此致以深情感謝，並祝各位盡其所長、鴻圖大展。除研究所同儕好友外，吳善業(Gohan)、劉郡怡、程心慧(Mor)、林瑾瑜、王學樑、王肇霖、陳里芳、郭逢祥、沈銘原及張依寰等知己莫逆相伴鼓舞，使我於研究時不感孤苦無依，更讓我的研究所生涯多采多姿，比他人更增添色彩，謹呈萬分謝意。我將最崇高之感謝之意呈獻給家父母－張習遠先生與謝小燕女士賢伉儷，感謝多年來栽培及支持，與關心愛護之恩情，無以為報，僅能將此論文獻於您們，這是屬於您們的榮耀。蒙乾爹娘－周正義先生與季惠琴女士賢伉儷，您們的鼓勵與關懷，給予我莫大前進的力量，讓我於成長階段勇於挑戰及創造，開創美好將來，感謝之意溢於言表。此外，大舅謝達華教授的恂恂善誘、諄諄教誨，於百忙之中還不忘提攜照顧，令外甥我於枷鎖中逃脫、低潮中走出，感激之心銘心鏤骨。再者，感謝張、謝氏歷代祖先、外公婆、姊張藝馨、乾哥(嫂)－周徹夫婦、哥張.

(6) 志豪、妹楊佳蓉、弟謝明勳，以及乾妹周黛珊(Vivian)等親戚之伴陪，讓我於求學過程不顯枯燥乏味，你們的讚許鼓舞也增強了我的信心與求學研究之興趣。最後，感謝一路走來始終伴陪於身的戴元君小姐，妳在我身後支持我的那雙手，協助著我邁向彼此共創的美妙未來，妳的疼惜關愛令我哀傷時得到安慰、生氣時獲得抒解、挫折時賦予我再度站起奮發向前的勇氣，甚至於痛苦時，妳那無人比擬的笑顏，更有著令我心曠神怡忘卻痛苦的魔力，妳永恆不變的支持與鼓勵，才令本論文可順利撰寫完成，妳可謂此論文之親生母親。在此以最感恩的心感謝所有關心我、關懷我、給予我幫助的人，相逢即是有緣，結識各位也令在下於人生路上不虛此行，感謝各位於我求學研究階段之勉勵激勵，令本人可以順利完成這本研究論文，在此以「天下無不散之宴席」及「結束即是開始」來與各位長輩、師長、學長姐、同學、學弟妹及所有知己好友道聲珍重再見，期待重逢！. 張育誠歲次戌子年暑. 謹誌. 于中山西灣.

(7) Table of Content. Table of Content .............................................................................................................. i List of Tables................................................................................................................... v List of Figures ................................................................................................................ vi 中文摘要 ......................................................................................................................... xi Abstract ........................................................................................................................... 1 Chapter 1 Introduction ............................................................................................... 3 1.1. Glass ............................................................................................................... 3. 1.2. Metallic glasses .............................................................................................. 5 1.2.1 Evolution of metallic glasses................................................................. 5 1.2.2 Manufacture methods of metallic glasses.............................................. 8 1.2.3 Factors of the glass forming ability (GFA) ......................................... 11 1.2.4 Bulk metallic glasses ........................................................................... 15 1.2.5 Characterizations of bulk metallic glasses........................................... 16 1.2.6 Application of bulk metallic glasses.................................................... 20 1.2.7 Workability of bulk metallic glasses ................................................... 23. 1.3. Mg-based bulk metallic glasses.................................................................... 24. Chapter 2 Background and literature review ......................................................... 25 2.1. The forming conditions of amorphous alloys............................................... 25 2.1.1 The empirical rules for forming amorphous alloys ............................. 25 2.1.2 The correlative theories of empirical rules .......................................... 27. 2.2. Viscous flow behavior .................................................................................. 31 2.2.1 Thermodynamics and kinetics of metallic glasses .............................. 32 2.2.2 Kinetics and viscosity of metallic glasses ........................................... 35 2.2.3 Angell plot and Vogel-Fulcher-Tammann (VFT) formula.................. 37 i.

(8) 2.3. Crystallization of supercooled metallic liquids and glasses ......................... 38. 2.4. New viscosity measurement of bulk metallic glasses .................................. 41 2.4.1 Three point beam bending method ...................................................... 41 2.4.2 Dimensional changes method.............................................................. 42. 2.5. Deformability of supercooled liquids ........................................................... 44 2.5.1 Deformation behavior.......................................................................... 44 2.5.2 Microforming ...................................................................................... 45 2.5.3 Deformability parameter...................................................................... 47 2.5.4 Deformation model.............................................................................. 47. 2.6. Partial element replacement in Mg-Cu-Y(Gd) based bulk metallic glasses. 48 2.6.1 The effect of the boron element........................................................... 49 2.6.2 The effect of the silver element ........................................................... 50. 2.7. Purpose of this work ..................................................................................... 51. Chapter 3 Materials and Experiments .................................................................... 53 3.1. The preparation of amorphous alloy specimens ........................................... 53 3.1.1 Materials .............................................................................................. 53 3.1.2 Arc melting process ............................................................................. 53 3.1.3 Melt spinning technique ...................................................................... 54 3.1.4 Injection casting process...................................................................... 55. 3.2. Microstructure and phase identification ....................................................... 56 3.2.1 XRD analyses ...................................................................................... 56 3.2.2 SEM observations................................................................................ 56. 3.3. Thermal analyses .......................................................................................... 56. 3.4. Mechanical properties .................................................................................. 57 3.4.1 Hardness .............................................................................................. 57 3.4.2 Dynamic mechanical analysis ............................................................. 57 ii.

(9) 3.4.3 Thermomechanical analysis ................................................................ 58 3.5. Imprinting (Micro forming).......................................................................... 59. Chapter 4 Results and discussions ........................................................................... 60 4.1. The Mg-based amorphous ribbons ............................................................... 60 4.1.1 Microstructure ..................................................................................... 60 4.1.2 Thermal properties............................................................................... 60 4.1.3 Mechanical analyses ............................................................................ 61. 4.2. Thermal and thermomechanical properties of Mg58Cu31Y11 bulk amorphous. alloy ........................................................................................................................ 63 4.2.1 Microstructure and thermal properties ................................................ 63 4.2.2 Thermomechanical properties ............................................................. 64 4.2.3 Viscous flow behavior ......................................................................... 65 4.2.4 Deformability ...................................................................................... 65 4.2.5 Findings for Mg58Cu31Y11 properties .................................................. 66 4.3. Micro-imprinting in Mg-Cu-Y metallic glasses ........................................... 67 4.3.1 Finite element simulation .................................................................... 67 4.3.2 Morphology of the micro-lens mold and imprinted specimens........... 68 4.3.3 Comments for micro-imprinting.......................................................... 69. 4.4. Workability and thermomechanical properties of Mg-Cu-(Ag, B)-Gd bulk. metallic glasses ....................................................................................................... 70 4.4.1 Phase demonstration and thermal properties ....................................... 71 4.4.2 Thermomechanical analysis ................................................................ 72 4.4.3 Viscous flow behavior ......................................................................... 73 4.4.4 Mechanical formability ....................................................................... 75 4.4.5 Deformation model.............................................................................. 76 4.4.6 Combined effects of B and Ag addition .............................................. 79 iii.

(10) Chapter 5 Summary .................................................................................................. 81 References ..................................................................................................................... 84 Tables............................................................................................................................. 93 Figures ......................................................................................................................... 107. iv.

(11) List of Tables. Table 1.1 The features and applications of metallic glasses....................................... 93 Table 1.2 Bulk metallic glasses and their developed year.......................................... 94 Table 1.3 Summary of ΔTx, Trg, γ, Rc and Zc for typical BMGs ................................ 95 Table 1.4 The classification of amorphous alloy systems .......................................... 96 Table 2.1 Binary amorphous systems and mixing enthalpy values calculated based on Miedema’s macroscopic model.................................................................. 97 Table 2.2 The density for various Mg-Cu-Y-B system combinations ....................... 98 Table 2.3 Thermal properties and GFA index for the Mg-Cu-Y-B glassy alloys ...... 99 Table 2.4 Thermal analysis and maximum diameter for fully amorphous phase formation (Dmax) for Mg-Cu-(Ag)-Y(Gd) alloys ..................................... 100 Table 3.1 The compositions of the metallic glasses in this research........................ 101 Table 4.1 Summary of the thermal properties of the Mg67Cu25Y8, Mg65Cu25Y10 and Mg58Y31Gd11 metallic glass ribbons, at a heating rate of 20 K/min......... 102 Table 4.2 The micro/nano- hardness of the Mg67Cu25Y8, Mg65Cu25Y10 and Mg58Y31Gd11 metallic glass ribbons......................................................... 103 Table 4.3 Viscous flow behavior of the Mg58Cu31Y11 BMGs obtained from TMA at a heating rate of 10 K/min........................................................................... 104 Table 4.4 Summary of the thermal and thermomechanical properties of the Mg-Cu-(Ag, B)-Gd amorphous samples, measured by DSC and TMA. . 105 Table 4.5 Summary of the fragility and viscous deformation related parameters of the Mg-Cu-(Ag, B)-Gd BMGs....................................................................... 106. v.

(12) List of Figures. Figure 1.1. Illustrator of different parts to form the glassy state, roughly indicating the energies of the initial states relative to the final glassy states............... 107. Figure 1.2. The arrangement of atoms in (a) crystalline and (b) amorphous states. 107. Figure 1.3. The surface of a metallic glass.............................................................. 108. Figure 1.4. Schematic drawing of the splat quenching method. ............................. 109. Figure 1.5. Schematic drawing of the two rollers quenching process..................... 110. Figure 1.6. The relationship of critical thickness and the date of discovery........... 111. Figure 1.7. Schematic drawing of (a) sputtering and (b) vacuum evaporation ....... 112. Figure 1.8. The DSC trace of an amorphous alloy.................................................. 113. Figure 1.9. The relationship of Rc and ΔTx ............................................................. 113. Figure 1.10 The relationship of Rc and Trg ............................................................... 114 Figure 1.11 The definition of the γ value ................................................................ 115 Figure 1.12 The relationship of (a) Rc and γ, and (b) Zc and γ ................................. 116 Figure 1.13 The relationship of Rc and γm ................................................................ 117 Figure 1.14 Relation between tensile fracture strength, hardness and Young’s modulus for bulk amorphous alloys and conventional crystalline alloys............ 118 Figure 1.15 Maximum bending and rotating beam fatigue stress as a function of cyclic number up to failure for bulk amorphous Zr65Al10Ni10Cu15 and Pd40Cu30Ni10P20 alloys. ......................................................................... 119 Figure 1.16 The pictures are the Baseball bat featuring based on Vitreloy (a), and tennis racket with BMGs casting in the frame (b) ................................ 120 Figure 2.1. Schematic drawing of binary phase diagram ........................................ 121. Figure 2.2. Mechanisms for the stabilization of supercooled liquid and the high glass forming ability ...................................................................................... 122 vi.

(13) Figure 2.3. Differential Scanning Calorimeter (DSC) thermogram of Zr41.2Ti13.8Cu12.5 Ni10Be22.5 from 500 to 1150 K using the scan heating rate of 10 K/min.. ............................................................................................................... 123. Figure 2.4. Heat capacity curves of the Zr41.2Ti13.8Cu12.5Ni10Be22.5 glass, the corresponding liquid and crystalline solid as a function of temperature. ............................................................................................................... 124. Figure 2.5. The Gibbs free energy difference between crystal and supercooled liquid for different various glass forming liquids. Vit1 with the smallest cooling rate shows the smallest free energy difference ..................................... 125. Figure 2.6. Fragility plot of the viscosities of the bulk metallic glasses and several non-metallic “strong” and “fragile” glasses. The data on non-metallic glasses were taken from Angell’s literature.......................................... 126. Figure 2.7. TTT Diagram for BMG showing the liquidus temperature, glass transition temperature, super cooled liquid region, and the crystalline nose. The samples have not been sheared during the experiment. The solid line and dashed line show the fits obtained by using Deff ∝ 1/η and Deff ∝ exp(– Qeff/k·T)................................................................................................. 127. Figure 2.8. Schematic figure of a typical TMA instrument .................................... 128. Figure 2.9. The TMA curve for such an experiment............................................... 129. Figure 2.10 Viscosity as a function of reduced temperature (T/Tm) for a Pd40Cu30 Ni10P20 amorphous alloy. ...................................................................... 130 Figure 2.11 Photograph revealing the external appearance of an amorphous La-based cylinder and an amorphous alloy wire and gear-shaped produced by pulling the amorphous cylinder. ........................................................... 131 Figure 2.12 Fine precision amorphous alloy mirrors prepared by die forging an amorphous Zr60Al10Ni10Cu20 alloy in the supercooled liquid region.... 132 vii.

(14) Figure 2.13 Map of the phases for bulk specimens with Ag content in Mg-Cu-Y (a) and Mg-Cu-Gd (b) alloys............................................................................. 133 Figure 3.1. The experimental flow chart. ................................................................ 134. Figure 3.2. The illustration of arc melting furnace.................................................. 135. Figure 3.3. The illustration of a single-roller melts spinning process..................... 136. Figure 3.4. The chart of an injection casting process.............................................. 137. Figure 3.5. The picture of Perkin Elmer Diamond DSC. ........................................ 138. Figure 3.6. The scheme charts of the (a) heat flux (TA 2920 DSC) and (b) power compensation (Perkin Elmer Diamond DSC) differential scanning calorimeters (DSC). .............................................................................. 139. Figure 3.7. The photo of a Shimadzu HMV-2000 Vicker’s microhardness tester.. 140. Figure 3.8. The standard Nano Indenter® XP is a complete, turnkey system consisting of the major components illustrated...................................................... 141. Figure 3.9. The photo of dynamic mechanical analyzer (DMA, Perkin Elmer Diamond DMA) .................................................................................................... 142. Figure 3.10 The picture of thermomechanical analyzer (TMA 7, Perkin Elmer Diamond) .............................................................................................. 143 Figure 3.11 Hot embossing set-up spring plate system............................................ 144 Figure 3.12 The mold used for (a) initial stage simulation and (b) real micro-imprinting experiment............................................................................................. 145 Figure 3.13 The schematic replication process on the PMMA ................................ 146 Figure 4.1. Photograph showing the surface appearance of melt spun Mg58Cu31Y11 alloy ribbons.......................................................................................... 147. Figure 4.2. X-ray diffraction patterns of the Mg-Cu-Y system amorphous alloy ribbons................................................................................................... 148. Figure 4.3. DSC scans of the Mg-Cu-Y system metallic glass ribbons .................. 149 viii.

(15) Figure 4.4. Temperature dependence of storage module and tanδ of the Mg58Cu31Y11 metallic glass ribbons obtained by DMA operated in the tensile mode at a heating rate of 10 K/min. ...................................................................... 150. Figure 4.5. The photo showing the surface appearance of injection cast Mg58Cu31Y11 alloy samples with different diameters, from 3 and 4 mm. .................. 151. Figure 4.6. (a) X-ray diffraction patterns and (b) DSC thermogram of the Mg58Cu31Y11 BMG obtained at a heating rate of 10 K/min.................. 152. Figure 4.7. Temperature dependence of relative displacement of the as-cast 4 mm bulk amorphous Mg58Cu31Y11 alloys obtained by TMA. ............................. 153. Figure 4.8. Relationship of the as-cast 4 mm bulk amorphous Mg58Cu31Y11BMG between applied stress and ΔL/L0 at (a) Tonset, (b) Tvs, and (c) Tfinish.... 154. Figure 4.9. Typical TMA and DTMA curves measured at stress level of 7.08 kPa for as-cast bulk amorphous Mg58Cu31Y11 alloys. ....................................... 155. Figure 4.10 Temperature dependence of effective viscosity for the quoted applied stresses at a heating rate of 10 K/min in the Mg58Cu31Y11 bulk metallic glass....................................................................................................... 156 Figure 4.11 The temperature dependence of ΔL/P (displacement/applied stress) at various applied stresses for the as-cast Mg58Cu31Y11 samples. ............ 157 Figure 4.12 Variation of the onset temperature of viscous flow (Tonset) with applied stresses. ................................................................................................. 158 Figure 4.13 The simulated imprinting evolution and the corresponding load prediction: (a) 60 s, pressure 31 Pa and (b) 240 s, pressure 1176 Pa. ..................... 159 Figure 4.14 The forming extent of the BMG under an applied pressure of (a) 100 kPa and (b) 300 kPa for 4 min. .................................................................... 160 Figure 4.15 The forming extent of the BMG under an applied pressure of 400 kPa for (a) 1 min and (b) 3 min. ........................................................................ 161 ix.

(16) Figure 4.16 SEM micrograph of the formed BMG, operated at 150oC for 4 minutes under a pressure of 400 kPa. ................................................................. 162 Figure 4.17 Comparison of the surface profiles of the original Ni-Co mold, Mg58Cu31Y11 mold, and the PMMA. .................................................... 163 Figure 4.18 X-ray diffraction patterns of the as-quenched Mg65Cu25Gd10, Mg65Cu15 Ag10Gd10, Mg65Cu22Ag3Gd10, and Mg65Cu22B3Gd10 samples. ............. 164 Figure 4.19 DSC thermograms of the Mg65Cu25Gd10, Mg65Cu15Ag10Gd10, Mg65Cu22Ag3Gd10, and Mg65Cu22B3Gd10 bulk metallic glasses ........... 165 Figure 4.20 The temperature and compressive strain curves of Mg65Cu25Gd10, Mg65Cu15Ag10Gd10, Mg65Cu22Ag3Gd10, and Mg65Cu22B3Gd10 bulk metallic glasses. .................................................................................... 166 Figure 4.21 Measured viscosities in the supercooled liquid for a constant applied load of 7.08 kPa under isochronal heating scan at 10 K/min. ...................... 167 Figure 4.22 (a) Temperature dependence of the viscosity of Mg65Cu25Gd10, Mg65Cu15Ag10Gd10, Mg65Cu22Ag3Gd10, and Mg65Cu22Gd10B3 bulk metallic glasses, (b) the Angell plots the current four alloys, as compared with Zr46.5Ti8.25Cu7.5Ni10Be27.5 and Au77Ge13.8Si9.4 alloys. ................... 168 Figure 4.23 Determination of the STZ sizes of the Mg-Cu-(Ag)-Gd alloys based on the TMA data after the correction for coefficient of thermal expansion. ... 169 Figure 4.24 Extraction of the activation energy of the Mg-Cu-(Ag)-Gd alloys during shear deformation within the supercooled temperature region............. 170 Figure 4.25 Determination of STZ sizes of the Mg65Cu25Gd10 and Mg65Cu22B3Gd10 alloys based on the TMA data after the correction for coefficient of thermal expansion. ................................................................................ 171 Figure 4.26 Extraction of the activation energy of Mg65Cu25Gd10 and Mg65Cu22B3Gd10 during shear deformation within the supercooled temperature region. 172 x.

(17) 中文摘要. 近年來，非晶質合金之未來潛力與實質工業應用開始被多加關注，尤其微機電系統(MEMS)零組件與光電壓印材料方面，更是投注大量的人才來研發。本論文主要以具低玻璃轉位溫度(Tg)之輕量型鎂基非質晶合金，作為新一代的壓印材料之研究。且利用熱機性分析儀(Thermomechanical analyzer)來獲得鎂基非晶質合金之粘度行為與熱機性參數，並進而探討其二次加工性，提出加工溫度區間、耐熱程度等實用之數據，作為日後工業應用之有利資訊。輕量型鎂基非晶質合金具良好之玻璃形成能力(Glass forming ability)，根據鎂基非晶質合金之發展歷程，與熱分析所得之結果，顯示鎂-銅-釔(Mg-Cu-Y)與鎂銅-釓(Mg-Cu-Gd)之金屬成分是易於製成塊狀鎂基非晶質金屬玻璃(Bulk metallic glass)。本文先以鎂-銅-釔作主軸，藉由熱分析與硬度結果作為依據，獲取最適合之塊狀非晶質合金成分(Mg58Cu31Y11)作為熱壓印材料樣本。樣本經由熱壓印機施予適當應力與時間之下，以幾近完美地形成六角鏡模具，且成功轉印於聚甲基丙烯酸甲酯(Polymethylmethacrylate)之上。有限元素分析法(Finite element simulation) 之結果，也與熱機性分析、壓印試片相互印證。此外，鎂-銅-釓之金屬成分更較鎂-銅-釔具良好玻璃形成能力，據相關文獻所言，若以少量銀(Silver)或硼(Boron)元素取代成分鎂-銅-釓(釔)中的銅(Copper)元素原子成分比，將可提升熱穩定性或機械性質，更增加金屬玻璃塊材形成大小，以便加強材料使用性。本論文也探討鎂-銅-(銀，硼)-釓多元鎂基合金之熱機械性質、粘滯流動行為和形變能力等加工性，並利用形變模型來深入探討其內部結構與能量關係之影響。由研究發現，添加銀或硼元素，雖增加塊狀大小、機械性質或熱穩定性，但卻提高粘滯流動困難性、增加活化所需能量、降低形變能力等負面加工性因子，侷限二次形變能力。但銀、硼取代銅元素之玻璃金屬材料確實顯示出較佳之機械性質和室溫下之硬度。由此可知，材料應用性不應只單探討材料之形變能力、熱性質或機械性質，而是需全方面考量其應用所需。 xi.

(18) 最後，本論文提出數個指數如粘滯性 (Viscosity) 、 VFT 溫度 (Vogel-Fulcher-Tammann temperature)、應力下之熱穩定性、形變性(Deformability) 等，以量化數據方式，來得知材料之加工性質，作為塊狀玻璃金屬加工性之參考，以便用於未來應用發展之依據。. xii.

(19) Abstract. In the near couple years, the applications of amorphous alloys have attracted great attention due to their characteristics and future potential. This research is intended to synthesis a lighter Mg-based amorphous alloy as the imprinting materials for micro-electromechanical system (MEMS) with a high glass forming ability (GFA) and lower glass transition temperature (Tg). Also, the workability of the Mg-based metallic glasses is examined in terms of several viscous flow behaviors and parameters obtained from the thermomechanical analysis (TMA). The lighter Mg-based metallic glasses exhibit their superior glass forming ability, and can be cast into bulk metallic glasses (BMGs). Based on the thermal analysis of the Mg-Cu-Y glassy materials, the evaluation of the glass forming ability and thermal stability for searching the optimum alloy composition is conducted. By using Mg58Cu31Y11 amorphous alloy with the best composition as the micro-forming specimens, imprinting was made by hot pressing at 150oC with several applied compressive stresses to form the hexagonal micro-lens arrays. Finite element simulation using 3D Deform software is also applied to trace the microforming evolution, and to compare with the experimental observations. The results demonstrate that the imprinting is feasible and promising. On the other hand, the Mg-Cu-Gd BMGs with even better GFA than Mg-Cu-Y are explored in terms of their thermomechanical properties. Extension of this study is performed partially by Cu replacing by Ag or B for the improvement of maximum diameter and thermal stability. And the workability of these Mg-Cu-(Ag, B)-Gd metallic glasses, namely, Mg65Cu25-xAgxGd10 (x = 0, 3, 10 at %) and Mg65Cu22B3Gd10 is evaluated in terming of the thermomechanical parameters, viscous flow behavior, deformability, and the deformation model. It is found the fragility for viscous 1.

(20) deformation would increase with the replacement of Ag or B, leading to the negative factors for the micro-forming and nano-imprinting practices. This conclusion is supported by the many extracted parameters. Thus, even the B-additive Mg based BMG has much higher hardness and Ag-additive Mg based BMG has the larger maximum rod diameter, they are more difficult to be formed, appearing as a negative factor in the micro-forming or nano-imprinting industry. The base Mg65Cu25Gd10 alloy stilly appears to be more promising than the Ag or B-containing alloys when the viscous forming is under consideration.. 2.

(21) Chapter 1 Introduction 1.1 Glass. Glass, in the traditionally correct conception, is a liquid that has lost its capability of flow. In some cases, the glassy structure is hardly identifiably from the fluid substance before it becomes the glass [1]. Why do these special materials or solutions abruptly occur the dramatic “slowing down” in the diffusive motions of its particles? At some exactly freezing temperatures, these substances do not form precisely ordered crystalline materials like so many other normal matters, why? These questions are brought forward by physicists as well as materials scientists. In 1995, Angell’s paper [1] was published to give an overview of many approbatory views and to highlight some of the efforts to answer these questions. The answers of these questions affect on a wide range of studies. It is known that glass occurs naturally as the volcanic glass, and a recent idea is also shown that most of the universe water exists in the glassy state like water in comets (formed by condensation from the gaseous state at very low temperatures). And the most polymers in common uses are non- or semi-crystalline solids, but these rubbers become splintered like dropped goblet when impacted at the liquid nitrogen temperature. The instance of the naturally occurring glassy water in comets demonstrates that glasses are not necessarily formed by the cooling of a liquid state. In face, the glassy state can be obtained by many different methods and behave as the same materials fundamentally. Vitreous silica, the typical glass material, can also be obtained by many ways, for instance, the cooling of liquid silica, vapor condensation, heavy particle 3.

(22) bombardment of a crystalline form, chemical reaction of organosilicon compounds followed drying, and vapor-phase reaction of gaseous molecules followed by condensation. In fact, only the latter route with the purified reactants produces the high purity glass sufficient for the requirements of the fiber optics communications technology. While the densities of the products prepared by the different processes may not be identical without an annealing process, they can still be differentiated by the characteristics of their X-ray diffraction patterns. The various methods for preparing glasses are summarized in Figure 1.1 [2]. For many matters and mixtures, there are non-crystalline (amorphous) structures for the atoms and molecules with constitutionally low energy, and these particles can easily congregate. These may not be the lowest energy packing modes, because the exothermic reaction occurs during crystallization. During the cooling state at normal rate, for the good glassy formers, the probability to nucleate a crystal rather than to form of the glassy solid is low. In the case of liquid B2O3 (ingredient of Pyrex glass) at an ambient pressure, the crystals will not grow even when the B2O3 melt is seeded with a crystallite of the stable phase, because of the small driving force and the slow kinetics. Crystallization in this case is only induced by raising the pressure [3]. In some binary solutions at low temperatures, the glassy state can be as thermodynamically stable as crystals or crystal plus solution (coagel) [4]. The most thoroughly studied method to form a glassy state is still the cooling of a liquid with diffusion slowdown. This method has been applied to many glass-forming systems, including proteins, biopolymers and metallic glasses [1, 5, 6].. 4.

(23) 1.2 Metallic glasses. Metallic glasses, in the substantial definition, are the metallic materials without the periodicity of crystal, and with a disordered atomic-scale structure. In contrast with other crystalline metals, with a highly ordered arrangement of atoms and long range order, these metallic glass alloys are non-crystalline. However, the atomic dispersion in metallic glasses may be incompletely random, and retains a greater degree of short range order than that in the liquid, and both lack of the long-range order arrangement, as shown in Figure 1.2. As a result, the metallic glasses have been called as liquid metals due to this kind of liquid-like atomic arrangement, and these liquid metals have also been called as non-crystalline metals, amorphous metals, and glassy metals. Metallic glasses are usually less brittle than oxide glasses, however, these amorphous metals are glasses without the transparency, and look like opaque shiny and smooth metals, as shown in Figure 1.3 [7]. The amorphous alloys show several special properties as compared with the corresponding crystalline alloys, including (a) lower Young’s modulus and higher tensile strength, (b) larger elastic elongation about 2%, (c) larger elastic energy up to yielding, (d) indistinct plastic elongation due to inhomogeneous deformation mode, (e) relatively greater impact fracture energy, (f) better anti-oxidation/corrosion behaviors, (g) excellent electromagnetic properties, and (h) easy shaping or forming ability [8]. The unique properties and application fields of amorphous alloys are summarized in Table 1.1 [9].. 5.

(24) 1.2.1. Evolution of metallic glasses. In 1960, Klement et al. first discovered the amorphous or metallic glassy phase [6]. The splat quenching method (Figure 1.4) was used to solidify an Au75Si25 liquid into a supercooled state with cooling rates between 106 to 1010 K/s. Then, the first major breakthrough in metallic glass formation came in 1960 when Duwez and co-workers discovered that liquid Au80Si20 [6] could be turned into an amorphous solid by direct quenching from the melt with a cooling rate of 106 K/sec. In these trailblazers’ literature, they have made a metallic glass ribbon with a thickness of 1~10 μm. After couple years, the ternary metallic glasses of Au-Si-Ge and Pd-Si-M (M=Ag, Cu or Au) were produced by Chen and Turnbull [10, 11]. In 1970, Chen and Miller [12] used two rollers quenching process (Figure 1.5) to synthesize a 2 mm wide and 50 μm thick metallic glass ribbon with several meters long. This was the first time to produce the longer length scales of metallic glasses, so the continuous liquid quenching process became the primary way to make the metallic glass ribbons. During 1970s, amorphous alloys were regarded as the new classes of soft magnetic alloys [13, 14]. Hence, using the metallic glass core in transformers became the ground rule. Regrettably, the commerciality of metallic glasses also terminated here. In the late 1980s, many new “engineering” properties of metallic glasses were discovered, leading to a revival of interest in the amorphous alloys. In Inoue’s group, the precursors have found the new multi-component systems metallic glasses consisting mainly of common metallic elements with lower critical cooling rates [9], and they have also made many valuable works on fracture characteristics, crystallization behavior and stabilization phenomena of metallic glasses. The first bulk metallic glass (BMG) was reported by Drehman and Greer [15]; and Kui et al. [16] by quenching ternary Pd-Ni-P alloy in 1984. Heterogeneous 6.

(25) nucleation sites on the surface were dissolved, by repeated melting in a boron oxide flux. As a result, glass formation at cooling rates below 10 K/s and glassy ingots with sizes up to 10 mm in diameter were synthesized. Today, the bulk metallic glass (BMG) is defined in the community as a metallic glass sample with its smallest dimension at least 1 mm thick. Hence, the size of metallic glasses is also mentioned initiatorily on the following literatures. In 1991, Inoue et al. cast the La55Al25Ni10Cu10 fully glassy alloy with a 9 mm thick by using cooling Cu molds [17]. Next year, the Mg-Cu-Y BMGs with a diameter of 7 mm were also progressed [18]. After these BMGs are discovered, the same group reported a family of Zr-Al-Ni-Cu based metallic glassy alloys with a high glass forming ability (GFA), and the Zr65Al7.5Ni10Cu17.5 amorphous alloy is seized of the largest thickness (~ 15 mm) among this family [19]. On the same year, Peker and Johnson discovered an exceptionally good glass forming system; Zr-Ti-Cu-Ni-Be. In particular Zr41.2Ti13.8Cu12.5Ni10Be22.5, which is commercially known as Vitreloy1TM (Vit 1), exhibits a critical cooling rate of 1 K/s and can be cast up to several centimeters in diameter [20]. However, the presence of toxic beryllium in this alloy limits its use. Johnson and co-workers further have developed another group of easily processible BMGs Zr-Ti(Nb)-Cu-Ni-Al alloy in 1997, namely, Zr52.5Ti5Cu17.9Ni14.6Al10 and Zr57Cu15.4Ni12.6Al10Nb5 [21]. Even though their critical cooling rate is higher than Vitreloy1TM, they still have good mechanical properties and do not contain beryllium. Other important credits include that Inoue et al. have re-investigated the first discovery of Pd40Ni40P20 BMG, and they discovered the Pd-Ni-P BMG can increase the critical thickness from 5 mm to 72 mm when 30% of nickel replaced by copper [22]. Also, Ma et al. [23] developed the Mg-based BMG with inch-diameter successfully in 2005, bringing on a momentous progress on Mg-based BMG application. Table 1.2 [24] is the list of the familiar amorphous alloys and BMG systems, and Figure 1.6 shows the relationship of critical thickness with the date of discovery [25]. 7.

(26) 1.2.2. Manufacture methods of metallic glasses. Although metallic glasses are thermodynamically metastable, there are many processing routes, in which they can be processed. We can use gases (vapor deposition), liquids, or even crystals (solid state amorphization) as the starting material. From different cooling rates, these methods can be classified into the following three types: (1) cooling the gaseous state to the solid state, (2) cooling the liquid state to the solid state, and (3) transforming the solid state to another solid state.. Cooling the gaseous state to the solid state. The amorphous alloys can be produced by depositing the gaseous alloys or metal elements onto a cool substrate with a higher cooling rate of 1010 - 1012 K/s. The sputtering [26] and vacuum evaporation [26], shown in the Figure 1.7 [27], are two kinds of processes that belong to this method. For the sputtering process, a high voltage is applied on two electrodes to create an electronic filed under a gas (nitrogen or argon) environment. Then the electrons are emitted from the cathode and accelerated in the electronic filed. These electrons will excite the gas molecules into positive gas ions and electrons. Once the gas ions impact an alloy or metal target, the alloy or metal atoms will deposit on the substrate and form an amorphous film. On the other hand, the vacuum evaporation process is necessary to use a heating and vacuum system. The alloy or element vapors which are emitted from the heating target are also deposited on the substrate to form the amorphous film. Although this method has the fastest cooling rate, there are still some disadvantages. The high vacuum equipments are expensive, and the size of amorphous samples is too small to make useful applications in bulk form. Recent studies [28] 8.

(27) suggest that the thick film amorphous layer (around 100 μm to 1 mm) on the substrate can improve the substrate material hardness and toughness. This will promote the casting applications of sputtering for making amorphous surface films.. Cooling the liquid state to the solid state. Quenching liquid with a cooling rate of 103 - 108 K/s is used extensively in fabricating the amorphous alloys, and the amorphous alloys can be made into all kinds of forms by using this method, including the splat quenching, twin roll quenching, melt spinning, planar flow casting, metallic mold casting, high-pressure die casting and spray forming, etc [29-31]. The first step in this method is to melt or to atomize the alloys, and then quench them onto a low temperature mold or substrate with high thermal conductivity. The first amorphous metal Au-Si was prepared by this method (gun quenching) [32]. In this process, the molten sample is held in a non-reactive crucible with a small hole at the bottom, and small droplets are driven out of the hole by a shock wave. The droplets then impinge onto a highly conductive metal substrate such as copper, spread out and form a film. Although this method provides the highest cooling rate of about 106 - 108 K/sec, the irregularity of the foil thickness makes the sample to be a mixture of amorphous and crystalline structure. The average thickness of the film formed by this method ranges from 5 to 25 μm. Another method to obtain foils of thickness ranging from 20 to 50 μm is the hammer-anvil method [33]. In this process, two metallic pistons are propelled towards each other at high speeds, and a molten drop of alloy is quenched into a foil between the two metallic pistons. This process also provides high cooling rates. A major advancement in processing came with the development of twin roll 9.

(28) technique. This technique was potentially the first continuous process to produce metallic glasses. The thin ribbons of metallic glasses are obtained when a molten metal is fed into the nip of two rapidly rotating wheels. However, in this process, it is difficult to keep the liquid from either solidifying too early (before the minimum separation between the wheels occurs) resulting in cold worked strips, or solidifying too late (i.e. leaving the nip partially liquid). To date, the most accepted continuous processing technique of making metallic glass ribbons is melt spinning. In this technique a stream of molten metal is directed at a rapidly rotating copper wheel. This process provides high cooling rates, and the ribbons with the thickness ranging between 20 and 100 μm can be produced [34]. The most widely used technique to process (the novel) bulk metallic glasses both in industry and research is casting. This is a relatively simple technique, in which the molten material is poured into the mold with a desired pattern. It is only used for processing bulk metallic glasses due to the limited cooling rates. This technique has been widely employed at the many university laboratories and in the industry such as Liquid Metal Technologies for the mass production of parts.. Transforming the solid state into another solid state. When using this method, the cooling rate is not necessarily critical, and the processes include severe plastic deformation, the solid-state reaction, particle bombardment, and solid-state inter diffusion, etc. The principle of severe plastic deformation is to refine the grain size of the alloy by creating a large amount of plastic deformation. When the grain size is small enough, the structure of the crystalline alloy is transformed into the amorphous state. There are many ways to achive a large amount of plastic deformation , such as cyclic extrusion or 10.

(29) cyclic compression, torsion straining under high pressure, equal channel angular pressing (ECAP), mechanical alloying (MA), and accumulative roll bonding (ARB) [35-40]. When high-energy electrons or heavy ions such as N+ impact an alloy surface, the atoms of alloy surface might rearrange to form the amorphous state. This is the basic concept of the particle bombardment method. Various processes used for this method include neutron and ion particle bombardment irradiation, electron beam radiation, ion implantation, and ion beam mixing. Finally, the solid-state interdiffusion reaction method is different from the two methods mentioned above. As two different metal layers or thin films are treated with an appropriate heat treatment process, the atoms between two metal layers will diffuse to each other and make an amorphous layer, provided that the temperature is not high enough to induce any intermetallic compound.. 1.2.3. Factors of the glass forming ability (GFA). The high GFA of cast amorphous alloy. s means a larger size of an amorphous. alloy can be fabricated easily under a lower cooling rate. Meanwhile, many researches have noticed that the GFA is affiliated with the thermal properties [41-45], atomic sizes [46-51], compositions [52-56] and electronic configurations [57-62] of the amorphous alloy. Hence, some parameters are proposed to estimate the GFA and to design the metallic glass systems. Among these parameters, the thermal properties are applied extensively due to two advantages. One is that the thermal information of amorphous alloy is unique and can be obtained easily and quickly. Another advantage is that these data can be transformed to an index by simple calculation. 11.

(30) Supercooled liquid region (ΔTx). For estimating GFA of metallic glasses, the first famous index is the supercooled liquid temperature region (ΔTx). In accordance with the DSC trace of a metallic glass in a heating process, there are two reactions before the metal is molten, as shown in Figure 1.8 [63]. One reaction is the glass transition reaction with an endothermic phenomenon, and the other is the crystallization reaction with an exothermic phenomenon. Accordingly, the supercooled liquid region is defined by the difference between glass transition temperature, Tg, and crystallization temperature, Tx. And it can be described as. ΔTx = Tx − Tg .. (1-1). This temperature region not only shows the thermal stability of amorphous alloy, but also exhibits the GFA of metallic glasses. The relationship of critical cooling rate (Rc) and supercooled liquid region (ΔTx) is shown in Figure 1.9 [64].. Reduced glass transition temperature (Trg). Another factor predicted by Turnbull [45] in improving the glass-forming ability is the reduced glass transition temperature, Trg’ = Tg/Tm or Trg = Tg/Tl, and it is an important factor to determine the critical cooling rate in the supercooled liquid among these factors. A larger Trg value would lead to a lower critical cooling rate, and a better GFA. This correspondence is shown in Figure 1.10 [64]. However, Lu et al. [65] had reported that the Trg = Tg/Tl is better than Trg’ = Tg/Tm for estimating the GFA of amorphous 12.

(31) alloys. For most alloys, the glass transition temperature is influenced slightly by the change of alloy compositions; however, the liquidus temperature changes distinctly with the change of alloy compositions. So the Trg exhibits a strong dependence in alloy compositions.. Gama value, γ. Trg and ΔTx both are usually employed to evaluate the GFA, but unfortunately some metallic glass systems cannot be satisfactorily predicted the degree of vitrifaction synchronously by using these two indexes. In some situations, the ΔTx index is better than the Trg index to judge the GFA [66-68]. Whereas, the Trg index is more suitable than the ΔTx index in some other amorphous alloys [65, 69, 70]. Thus, a new index γ was proposed by Lu and Liu [71], and the definition of γ value is shown in Fig. 1.11 [72]. There are two targets for achieving a better GFA amorphous alloy. One is the stabilization of liquid phase, and the other is the resistance of crystallization. If an alloy reveals a lower Tl, the melting liquid phase of this alloy will be maintained to a low temperature under a cooling step. Then, if an alloy shows a lower Tg, it means that the metastable glass phase is more stable at low temperatures. The both cases suggest that the liquid phase is more stable. On the other hand, while the alloy has a higher Tx, it indicates that the starting crystallization of the glass phase proceeds at a higher temperature upon heating. Accordingly, the three kinds of phase transformation temperatures are combinatorial on the γ value which can be expressed as. γ=. Tx . Tg + Tl. (1-2). 13.

(32) The critical cooling rate (Rc) and the critical thickness (Zc) of amorphous alloys can be plotted against the γ value as shown in Figure 1.12 [71], and can be expressed by the following two equations,. R c =5.1 × 10 exp(− 117.19γ ) ,. (1-3). Z c = 2.8 × 10 exp(41.7 γ ) .. (1-4). 21. and. −7. In addition, the comparison of ΔTx, Trg, γ, Rc and Zc is summarized in Table 1.3 [71].. Gama-m value, γm. The above mentioned Gama value is based on two major circumstances: the resistance of crystallization and stabilization of liquid phase [71-73]. According to this idea, in our group Du et al. [74] brought forward a new GFA index, γm. In the stabilization for the liquid phase of an amorphous alloy, the concept of lower Tl is similar to the γ value. Thus, the GFA should be inversely proportional to the liquidus temperature. In other words, GFA is proportional to 1/Tl. Other than the γ value index considering the transition temperature, the γm value index deliberates the supercooled liquid region. The amorphous alloy with a larger ΔTx usually reflects a greater stability and a better glass-forming ability (GFA). As a result, the GFA should be proportional to (Tx-Tg). The factor for rhe stabilization of a liquid phase is summarized as. GFA ∝. ΔTx Tx − Tg = . Tl Tl. (1-5) 14.

(33) Now deliberating the resistance of crystallization, the higher crystallization temperature suggests a higher crystallization resistance. By normalizing this Tx by Tl in order to make for a dimensionless parameter, it suggests that GFA is directly recited to Tx/Tl. Coupled the above two considerations, the GFA index of γm is established as [74]. γm =. Tx − Tg Tl. +. Tx 2Tx − Tg = . Tl Tl. (1-6). The critical cooling rate of amorphous alloy shows the linear dependence with the γm value, as shown in Figure 1.13, and the Rc is calculated as [74] log R c = 14.99 − 19.441γ m .. 1.2.4. (1-7). Bulk metallic glasses. Since late twentieth century, the great deals of bulk metallic glass systems have been reported. Commonly, these systems of bulk metallic glasses could be divided into two main types: ferrous and nonferrous group metals, as shown in Table 1.4 [24]. The ferrous systems include the Fe-(Al, Ga)-metalloid (P, C, B, Si, Ge), (Fe, Co, Ni)-(Zr, Hf, Nb)-B, Fe-Co-Ln-B, and Ni-Nb-(Cr, Mo)-(P,B) alloys, and the nonferrous alloy systems include the Mg-Ln-M (Ln is the lanthanide atoms, and M is Ni, Cu or Zn metallic atoms), Ln-Al-TM (TM is the VI~VIII group transition metal atoms), Ln-Ga-TM, Zr-Al-TM, Zr-Ti-Al-TM, Ti-Zr-TM, Zr-Ti-TM-Be, Zr-(Nb,Pd)-Al-TM, Pd-Cu-Ni-P, Pd-Ni-Fe-P, Pd-Cu-B-Si, and Ti-Ni-Cu-Sn systems. The ferrous alloy systems are researched after the synthesis of the nonferrous alloy systems. The alloy components are sorted out into five species, as summarized in. 15.

(34) Table 1.4 [24]. The first group is consisted of ETM (or Ln), Al and LTM as exemplified by the for Zr-Al-Ni and Ln-Al-Ni systems, where ETM is IVB~VIB group transition metals and LTM is VIIB~VIIIB group transition metals. The second group is composed of LTM, ETM and metalloid such as the Fe-Zr-B and Co-Nb-B systems. The third group is the LTM Fe-(Al,Ga)-metalloid systems, and the fourth group is indicated by Mg-Ln-LTM and ETM (Zr,Ti)-Be-LTM systems. The fifth system, such as the Pd-Cu-Ni-P and Pd-Ni-P systems, are composed of only two kinds of group element (LTM and metalloid) different from the combination of the three types of group elements for the four previous groups. Note that the bulk metallic glasses are developed from many important engineering alloys, such as the Fe, Co, Ni, Ti, Cu, Zr and Mg based amorphous alloys. Over the years, the bulk metallic glasses are researched by the investigators worldwide to increase the maximum diameter in the order of Pd-Cu > Zr > Mg (Ln) > Fe > Ni > Co (Ti) systems.. 1.2.5. Characterizations of bulk metallic glasses. The bulk metallic glasses possess some various properties due to their unique structure, relative to the traditional crystalline alloys. The primary properties and application fields of these amorphous bulk metals are summarized in Table 1.1. Some more important properties are described below.. Mechanical properties. The superior mechanical properties of bulk metallic glasses are the most favorable characteristics for many applications. Since the atomic random arrangement configuration and dense packing structure, when the applied stress is on the amorphous 16.

(35) alloys, the amorphous alloys only allow limited atomic displacements to resist deformation. Due to the absence of dislocation mechanisms for plastic deformation, the amorphous alloys are always high strength materials. The bulk metallic glasses have unique mechanical properties which are different from those of the crystalline alloys, i.e., the bulk metallic glasses have higher tensile strength and lower Young’s modulus. The difference can reach about three times [75]. The relation between tensile fracture strength (σf), Vickers hardness (Hv) and Young’s modulus (E) for the various bulk metallic glasses is shown in Figure 1.14 [64], together with the data on the conventional crystalline alloys. The bulk metallic glasses have high σf of 840 - 2100 MPa combined with E of 47 - 102 GPa, depending on alloy compositions [64]. Thus, the bulk metallic glasses tend to exhibit higher σf, higher Hv and lower E than the corresponding crystalline alloys. The three-point bending flexural stress and deflection curves of the Zr-Al-Ni-Cu and Zr-Ti-Al-Ni-Cu bulk amorphous alloys have been measured [76]. These bulk amorphous alloys have high bending flexural strength values of 3000 - 3900 MPa which are 2.0-2.5 times higher than those for crystalline Zr- and Ti-based alloys. Figure 1.15 summarizes the bending and rotating beam fatigue strength as a function of fatigue cycle up to failure for the bulk amorphous Zr65Al10Ni10Cu15 [77] and Pd40Cu30Ni10P20 [78] alloys, respectively, together with the data under tensile stress conditions for various melt spun amorphous ribbons. It is confirmed that these Zr- and Pd-Cu-based bulk metallic glasses have good combination of various mechanical properties, which could not be obtained from conventional crystalline alloys.. Magnetic properties. Applications of magnetic ferrous glasses are mainly based on their superior soft 17.

(36) magnetic properties. Sensors for electronic article surveillance have become a large application of amorphous materials. Thin sheets of glassy materials were first produced by melt spinning and subsequent annealing, resulting in the formation of nanocrystals in a glassy matrix. These nanocomposites still sustain their soft magnetic properties, but have a higher saturation magnetization and lower magnetostriction than the purely amorphous base materials. The most famous example is the alloy with the tradename FINEMET [79], which is a Fe-rich Fe-Si-B alloy with a small addition of Cu and Nb. The use of thin sheets in high frequency applications is of advantage, since the eddy currents can be reduced. Thin ribbons of magnetic nanocomposites and purely metallic glasses are currently used in transformer cores, in magnetic sensors, and for magnetic shielding. Although some magnetic bulk glasses have recently been discovered, so far none of these materials show promising applications. The difficulty in the development of magnetic bulk glasses results from the fact that ferromagnetic elements like Fe, Co, Ni, and Gd are prone to oxidation, which increases the probability for heterogeneous nucleation. It was shown that the nucleation rate in Fe–Ni–P–B (Metglass) alloys can be reduced by four orders of magnitude when heterogeneous nucleation is minimized by the use of fluxing techniques [80]. Commonly, the types of magnetic amorphous alloy systems could be divided into the two categories of TM-TM and TM-M (TM = Fe, Co, Ni, Zr, Hf, etc; M = B, C, Si, P, Ge, etc) systems.. Chemical properties. When bulk metallic glasses for their good static and dynamic mechanical properties are used as structural materials, it is essential for the bulk amorphous alloys to have good corrosion resistance in various kinds of corrosive solutions. Due to the 18.

(37) homogeneous single phase structure, which is lack of grain boundaries and dislocations in crystals, the amorphous alloys have superior corrosion resistance. In addition, the corrosion resistance of amorphous alloy could be improved by adding some kinds of corrosive solute. The corrosion resistance is remarkably improved by the dissolution of Nb or Ta. Asami and Inoue [81] have examined that the corrosion resistance of the melt-spun Zr-TM-Al-Ni-Cu (TM=Ti, Cr, Nb, Ta) alloys in HCl and NaCl solutions, and found that the Nb- and Ta-containing amorphous alloys exhibit good corrosion resistance in their solutions at room temperature. The corrosion resistance is largest for the Nb-containing alloy, followed by the Ti-containing alloy and then the Zr-Al-Ni-Cu alloy. The corrosion resistance of the Nb-containing alloy is also superior to that of the pure Zr metal, indicating the remarkable effectiveness of Nb addition on the improvement of corrosion resistance even in the NaCl solution.. Other properties of amorphous alloys. Recently, the Zr–Al–Ni–Cu bulk metallic glasses with a wide supercooled liquid region before crystallization were found to exhibit a distinct plateau stage in the hydrogen pressure–concentration–isotherm relation, though the desorption ratio of hydrogen gas is very low (15%) [82]. It has subsequently been found that the desorption ratio increases remarkably in the Mg-based amorphous alloys and the desorption ratio reaches approximately 100% [83]. It is therefore said that the Mg-based bulk amorphous alloys are good candidate for hydrogen storage materials. On the other hand, the Pd-based bulk metallic exhibits very high efficient ratios for the generation of chlorine gas as an electrode material [77]. The high efficiency above 90% remains unchanged even after 500 cycles, though the efficiency of the 19.

(38) commercial pure Pt crystal decreases significantly after 20 cycles. The remarkable improvement of the generation efficiency of chlorine gas has already enabled us to use the Pd-based bulk amorphous alloy as a practical electrode material for the generation of chlorine gas.. 1.2.6. Application of bulk metallic glasses. For the past years, bulk metallic glasses have been developed in many alloy systems. As a result, the largest diameter of metallic glass exceeding 10 mm in many alloy systems, and the research and development research of metallic glass has advanced to a new era. So far, the principal areas of BMG products are sports, luxury goods, electronics, medical, defense, etc. Liquidmetal Technologies [7], which was founded in 1987 as Amorphous Technologies International, was the first company to produce amorphous metal alloys in viable bulk form. The first application was as golf club heads. Twice as hard and four times as elastic as Ti drivers, 99% of the impact energy from a BMG head is transferred to the ball (compared to 70% for Ti). Higher strength-to-weight ratio allows mass to be distributed differently, enabling various shapes and sizes of head. However, high production costs led to Liquidmetal Technologies to terminate manufacture in favor of licensing the technology to established club makers. Vitreloy can also yield stronger, lighter, and more easily molded casings for personal electronic products. In 2002, Liquidmetal began making components for liquid crystal display casings on cell phones, but the costs became a problem again. However, now a few BMGs casings have been still chosen by some technological companies for their personal computer screens, digital cameras, or laptops. Since the cast produces interest other valuable customers, the new manufactures 20.

(39) started produce the baseball bats featuring and the four areas of the tennis-racket frame (Figure 1.16). The increased stiffness enhances energy return with 29% more power [84]. Other applications in sporting goods include skis, fishing equipments, bicycle frames, and marine tools, etc. Most of BMGs provide the anti-scratch, dent-resistant and high-gloss properties. Thus, jewelry division and luxury watchmaker also use these BMGs for their valuable goods. Some BMGs also have a highly biocompatible, nonallergenic form, wear-resistant, anticorrosion and other special medical properties. For example, DePuy Orthopaedics is using these materials in knee-replacement devices. And Surgical Specialties begin producing ophthalmic scalpel blades using these BMGs. The scalpel blades are higher quality but less expensive than diamond, sharper and longer lasting than steel, and more consistently manufacturable, since they are produced from a single mold ready for use. Other applications include pacemaker casings, knives and razor blades. Some lightweight and inexpensive bulk metallic glasses [84] (e.g. Al-based alloys to replace Ti) are being developed for applications by multi-institution US Department of Defense (DoD) programs, including: the Defense Advanced Research Projects Agency (Fe, Al, Ti, and Mg based BMGs); the Caltech Center for Structural Amorphous Metals (Mg and Al based BMGs); the University of Virginia, University of Connecticut, and US Air Force (Al-based BMGs); the Texas A&M University (Zr-based BMGs and composite with crystalline phase powder); the Center for Science and Engineering of Materials at Caltech (Zr-based BMGs and other two-phase glassy alloys); the National Aeronautics and Space Administration (Zr-based BMGs and secret amorphous metals), and Liquidmetal Technologies (Zr and W based BMGs and composites), etc. 21.

(40) The DoD wants to develop military materials that are stronger, lighter, and more effective at high temperatures and stresses. For instance, in-situ W-reinforced BMG-composites can replace depleted uranium penetrators in antitank armor-piercing projectiles, because of their similar density and self-sharpening behavior. It is unlike most crystalline metal projectiles which flatten on impact, the sides of BMG-composites sheer away under dynamic loading. Other martial applications include lightweight fragmentation bombs, Lockheed Martin missiles and ceramic-BMG composite armor tiles. There are lots of good prospects for BMG materials whose properties favor easier and cheaper processing for more common-place applications. For example, a micro/nano forming or imprint technology [85-88] for micro electromechanical systems (MEMS) and microstructure fabrication is introduced where the bulk metallic glass (BMG) is formed at a temperature where the BMG exist as a viscous liquid under an applied pressure into a mold. This thermoplastic forming is carried out under comparable forming pressure and temperatures that are used for plastics. It also allows to net-shape three-dimensional parts on the micron scale. The technology can be implemented into conventional MEMS fabrication processes. The properties of BMG as well as the thermoplastic formability enable new applications and performance improvements of existing MEMS devices and nanostructures. These promising works, together with developments in America, Japan, European Union, Taiwan, and other countries greatly improves the prospects for the discovery of new BMGs with properties that will enable practical manufacturing. In turn, this is likely to open up a new space of potential applications.. 22.

(41) 1.2.7. Workability of bulk metallic glasses. The deformation mode of metallic glasses is divided into two modes, namely, inhomogeneous and homogeneous modes [89-93]. Metallic glasses usually exhibit inhomogeneous deformation at temperatures about 100 K below the glass-transition temperature. This inhomogeneous deformation is localized in discrete and thin shear bands, resulting from its non-hardenable nature. In contrast, higher temperatures lead to a homogeneous deformation at lower strain rates below 10-3 s-1. In this homogeneous deformation region, each volume element of the material contributes to the strain, resulting in a uniform deformation for a uniformly stressed specimen. However, the elongation on tensile testing is limited to below about 100% and the minimum strength is more than one-third of the room temperature strength which is much higher than that of ordinary crystalline alloys. Moreover, the metallic glasses usually crystallize at a lower temperature of one-half of the melting temperature. It is therefore, difficult to hot work the metallic glasses using a conventional method. Recently, a number of glassy alloys with a wide supercooled liquid region (above 50 K) and high glass-forming ability have been discovered [15, 16, 94, 20, 91, 95]. These glassy alloys allow the production of large-scale bulk glassy materials by consolidation of the glassy powders and casting at low cooling rates. Also, availability of the bulk glassy alloys enables unique approaches for forming complex-shaped components through significant viscous flow inherent to the supercooled liquid. The workability of the supercooled liquid in metallic glasses needs to be examined after preliminary investigations of its thermal and mechanical properties. Based on the many literatures, it is demonstrated that the determination of the workability of a BMG within the supercooled liquid regime is not straightforward. Several parameters, including the viscosity level, the supercooled temperature range, the 23.

(42) workability index, and the thermal stability, must be sought simultaneously. These factors must be considered when the BMGs are applied for industrial applications.. 1.3 Mg-based bulk metallic glasses. Among a large number of alloys, the Mg-based alloys have attracted attention especially due to a high strength to weight ratio and a low glass transition temperature. From the industrial needs of developing high strength materials with light weight, the Mg-based alloys have attracted more and more interest due to the lowest specific weight among all structural metallic materials. The Mg-based BMGs are regarded as a new family of promising materials with excellent specific strength and good corrosion resistance [96]. New Mg-based ternary amorphous alloys with a wide supercooled liquid region and a high glass formation ability have been developed in a number of alloy systems, such as Mg-TM-Y (TM = transition metal such as Cu, Ni, Zn). The formation of the Mg-based BMGs was firstly reported in 1991. Then, Inoue et al. [97] found that the Mg-Cu-Y alloys exhibited a high glass formation ability (GFA), which made it possible to produce bulk amorphous samples with a diameter of 4 mm by the copper mold casting method. Furthermore, the Mg-based BMGs exhibited high compressive fracture strength of 800-900 MPa [98] which is more than twice as high as the highest strength of conventional Mg-based crystalline alloys. Next year, Inoue’s group [18] succeeded in fabricating bulk Mg65Cu25Y10 metallic glasses with increased diameter up to 7 mm by using high-pressure die casting method. Since them, a new family of light amorphous alloy systems became available.. 24.

(43) Chapter 2 Background and literature review. 2.1 The forming conditions of amorphous alloys. A rapidly solidified amorphous phase is essential to suppress the nucleation and growth reactions of a crystalline phase in the supercooled liquid region between the melting temperature (Tm) and the glass transition temperature (Tg). In other words, the critical cooling rate (Rc) plays an important role in fabricating amorphous alloys. Furthermore, the critical cooling rate is also related with the alloy compositions.. 2.1.1. The empirical rules for forming amorphous alloys. Take a binary system as an example. Most amorphous alloys have the compositions lying near the deep eutectic points. The reason for this is easy to understand. A schematic binary phase diagram in Figure 2.1 [99] compares the freezing of the two compositions. Composition 1 passes through the liquidus line at a high temperature (point a); thus the melt has to be cooled over a very larger temperature range over which it is possible for crystalline phases to form and grow. On the other hand, the liquidus line of composition 2 lies at a much lower temperature (point b) so that the thermal energy available for crystal growth is smaller, and at the same time the temperature interval between the Tm and the Tg is small. Frankly, the amorphous structure can be obtained with this latter composition using a slower rate of cooling. The Inoue’s group had developed a series of amorphous alloy systems with high GFA in the past decade, and they summarized and proposed three simple empirical rules for the alloy design of BMGs with high GFA [64, 77, 95, 100]. The three empirical rules are presented as below: 25.

(44) (1) Multi-component system consisting of more than three main elements:. For the alloy system with more than three main elements, the difference of atomic sizes causes a retardation of atoms moving when quenching this melt into a solid state. Consequently, the atoms of the solid phase exhibit a short-range order arrangement, which leads to denser random packing. For this reason, the GFA of amorphous alloys will be improved with increasing categories of main elements that consist of the multi-component system.. (2) Above 12% difference in atomic size ratios among the main elements:. In the Hume-Rothery criterion [99], the forming condition of the solid solution is that the difference of atomic sizes should less than 15% between the solute and solvent. If the difference of atomic sizes is lager than 15%, the moving atoms will be retarded accordingly. Thus the viscosity of the melting alloy also increases, and the atoms cannot rearrange easily in the solidified process. With the increase of atomic sizes difference, the GFA of amorphous alloys can be largely improved.. (3) Negative heats of mixing among the main elements:. According to the thermodynamic theory, the heat of mixing is regarded as the ability of atomic bonding between two atoms in an alloy system. Larger negative heat of mixing refers to the fact that the bonding ability of the same atoms is very weak, but the bonding ability for distinct atoms is strong. When the liquid alloy is solidified with a larger negative heat of mixing, the distinct atoms tend to attract together and arrange in a random way. Accordingly, the amorphous alloys form easily when the distinct main 26.