下载文档原格式
二硼化钛陶瓷在不同温度下的氧化行为黄飞，傅正义，王为民，王皓，王玉成，张金咏，张清杰(武汉理工大学，复合材料新技术国家重点实验室，武汉 430070)摘要：采用静态氧化法对不同温度下TiB2陶瓷的氧化行为进行研究，利用X射线衍射仪、扫描电镜、X射线光电子能谱仪对氧化前后的样品进行表征。
结果表明：低温下TiB2陶瓷氧化动力学满足抛物线规律，并在表面形成液相B2O3，阻止氧化反应的进一步进行，冷却后B2O3以玻璃态覆盖在表面。
高温下TiB2氧化反应在4h前满足抛物线规律，表面形成一层TiO2多孔结构；氧化4h后，随着氧扩散距离的延长，扩散阻力加大，从而使氧化速率降低，氧化反应不再满足抛物线规律。
关键词：二硼化钛；氧化动力学；微观结构中图分类号：TF123；TB332 文献标识码：A 文章编号：0454–5648(2008)05–0584–04OXIDATION BEHA VIOR OF TITANIUM DIBORIDE CERAMIC AT DIFFERENT TEMPERATURES HUANG Fei，FU Zhengyi，W ANG W eimin，W ANG Hao，W ANG Yucheng，ZHANG Jinyong，ZHANG Qinjie(State key Laboratory of Advanced Technology for Materials Synthesis and Processing, Wuhan University ofTechnology, Wuhan 430070, China)Abstract: The oxidation behavior of TiB2 ceramics at different temperatures was investigated using the static oxidation kinetic method. The samples before and after oxidation have been characterized by X-ray diffractometer, scanning electron microscope and X-ray photoelectron spectrometer. The results show that the oxidation kinetics appear the parabolic law at low temperature. A liquid B2O3 coating on the surface of TiB2 ceramic could prevent from further oxidation. After the ceramic samples were cooled, their sur-faces were covered with glassy B2O3. At high temperature, the oxidation reaction of TiB2 ceramics showed the parabolic law only before 4h. Porous rutile TiO2 formed on the surface. But the oxidation behavior with the parabolic law for the TiB2 ceramics was not observed after oxidation for 4h because of the long path of diffusion, strong diffusion resistance and low reaction rate.Key words: titanium diboride; oxidation kinetics; microstructureTitanium diboride with P6/mmm structure is a uniquely stable compound of the boron element and tita-nium element.[1] TiB2 based materials have received wide attention because of their high hardness and elastic modulus, good abrasion resistance and superior thermal and electrical conductivity.[2–3] Potential applications in-clude high temperature structural materials, cutting tools, armor, electrodes in metal smelting and wear parts. De-spite its useful properties, the application of monolithic TiB2 is limited by poor sinterability, exaggerated grain growth at high temperature and poor oxidation resistance above 800.℃[4–5]The starting temperature to oxidize TiB2 ceramics is about 400℃ and oxidation kinetics is controlled by outward diffusion of interstitial titanium ions and inner diffusion of oxygen ions.[5–6] But there are conflicting viewpoints about the detailed oxidation process, for ex-ample, about the oxidation products and oxidation mechanism. Koh et al.[7] investigated the oxidation be-havior of dense TiB2 specimens with 2.5% in mass (the same below) Si3N4 and found that TiB2 exhibited two distinct oxidation behaviors depending on the tempera-ture. At temperatures below 1000℃, the oxidation layer comprised two layers: an inner layer of crystalline TiO2 and an outer layer mainly composed of B2O3. When the oxidation temperatures were higher than 1000℃, the收稿日期：2007–09–23。
铜与钨反应1. 引言铜和钨是两种重要的金属元素，在工业上具有广泛的应用。
本文将探讨铜与钨的反应及其相关性质。
2. 铜与钨的特性铜是化学元素周期表中的一种金属元素，其原子序数为29，原子量为63.546。
铜具有良好的导电性和导热性，在电子技术、建筑工程和制造业中广泛应用。
钨是周期表中的一种过渡金属元素，其原子序数为74，原子量为183.84。
钨具有高熔点、高密度、高硬度等特点，在电热器件和合金制造领域得到广泛应用。
3. 铜与钨的化学性质铜和钨都是金属元素，它们具有活泼的化学性质，在适当的条件下可以与其他元素发生化学反应。
3.1 铜的氧化反应铜可以与氧气反应生成氧化铜，化学式为2Cu + O2 -> 2CuO。
这是一种氧化反应，铜原子氧化成了Cu离子，同时氧气还原成了氧离子。
3.2 钨的氧化反应钨可以与氧气反应生成氧化钨，化学式为2W + 3O2 -> 2WO3。
这也是一种氧化反应，钨原子氧化成了W离子，氧气还原成了氧离子。
3.3 铜与钨的反应铜与钨可以发生反应，生成铜钨合金。
铜钨合金具有良好的导电性和导热性，广泛应用于电子元器件的制造。
4. 铜与钨反应的条件铜与钨发生反应需要一定的条件，包括温度、压力和反应物质的配比等。
4.1 温度的影响温度是影响铜与钨反应速率的重要因素之一。
一般情况下，较高的温度可以加快反应速率，使反应更加迅速进行。
然而，过高的温度可能导致反应过程不可逆，甚至破坏反应体系的稳定性。
4.2 压力的影响压力也是影响铜与钨反应速率的因素之一。
一般情况下，较高的压力会使反应体系达到更高的活性状态，从而加快反应速率。
然而，过高的压力可能对反应器具和反应条件造成不良影响，因此选择适当的压力是十分重要的。
4.3 反应物质配比反应物质的配比对铜与钨反应的结果有着重要的影响。
不同的配比可能导致不同的反应物种生成，因此需要根据具体实验要求和反应机理选择适当的配比。
5. 铜与钨反应的应用由于铜与钨反应生成的铜钨合金具有良好的导电性和导热性，被广泛应用于电子元器件的制造。
硒化锡合成反应方程式1. 引言硒化锡（SnSe）是一种重要的半导体材料，具有优异的热电性能和光学性能。
它在太阳能电池、热电材料、光电器件等领域具有广泛的应用前景。
硒化锡的合成方法有多种，其中最常用的是化学气相沉积（CVD）和溶剂热法。
本文将重点介绍硒化锡的溶剂热法合成反应方程式。
2. 实验原理溶剂热法合成硒化锡主要是利用一种含有硒源和金属锡源的溶剂，在高温下进行反应生成硒化锡。
常用的硒源有硒粉、硒酸等，金属锡源可以是金属锡粉、氧化锡等。
在反应过程中，首先将硒源和金属锡源加入适当的溶剂中，并进行搅拌混合，形成均匀的混合物。
然后将混合物转移到高温容器中，在一定温度下进行反应一段时间。
最后，通过冷却和过滤等步骤，得到硒化锡产物。
3. 实验步骤1.准备硒源、金属锡源和溶剂。
硒源可以选择硒粉，金属锡源可以选择金属锡粉，溶剂可以选择有机溶剂如乙二醇。
2.将硒粉和金属锡粉按一定比例加入乙二醇中，并进行搅拌混合，使其均匀分散。
3.将混合物转移到高温容器中，并加热至一定温度（通常为200-300摄氏度）。
4.在一定时间内保持恒定温度下反应，搅拌反应体系以促进反应进行。
5.反应结束后，将反应体系冷却至室温。
6.用滤纸或其他适当的方法过滤产物，将硒化锡分离出来。
7.对得到的硒化锡进行干燥处理，得到最终产物。
4. 反应方程式根据实验原理和步骤，可以写出硒化锡的合成反应方程式：Sn + Se → SnSe其中，Sn表示金属锡，Se表示硒。
5. 实验条件合成硒化锡的实验条件包括温度、时间和溶剂选择等。
通常情况下，温度在200-300摄氏度之间，反应时间可以根据需要进行调整。
乙二醇是常用的溶剂选择，也可以根据实际情况选择其他有机溶剂。
6. 结果与讨论通过溶剂热法合成的硒化锡具有良好的结晶性和纯度，可以用于进一步的物理性质测试和材料应用研究。
此外，通过调节反应条件和原料比例等因素，还可以得到不同形貌和尺寸的硒化锡纳米材料。
7. 结论本文介绍了硒化锡的溶剂热法合成反应方程式及实验步骤。
Trans.Nonferrous Met.Soc.China31(2021)586−594Mechanical and thermo-physical properties of rapidly solidifiedAl−50Si−Cu(Mg)alloys for thermal management applicationJun FANG,Yong-hui ZHONG,Ming-kuang XIA,Feng-wei ZHANGThe43Research Institute of China Electronic Technology Group Corporation,Hefei230088,ChinaReceived20April2020;accepted30October2020Abstract:Al−high Si alloys were designed by the addition of Cu or Mg alloying elements to improve the mechanical properties.It is found that the addition of1wt.%Cu or1wt.%Mg as strengthening elements significantly improves the tensile strength by27.2%and24.5%,respectively.This phenomenon is attributed to the formation of uniformly dispersed fine particles(Al2Cu and Mg2Si secondary phases)in the Al matrix during hot press sintering of the rapidly solidified(gas atomization)powder.The thermal conductivity of the Al−50Si alloys is reduced with the addition of Cu or Mg,by only7.3%and6.8%,respectively.Therefore,the strength of the Al−50Si alloys is enhanced while maintaining their excellent thermo-physical properties by adding1%Cu(Mg).Key words:Al−50Si alloy;rapid solidification;thermal management material;mechanical property;thermo-physical property1IntroductionAl−Si alloys containing high Si contents,also called as Al−high Si alloys or Si p/Al composites, exhibit an excellent combination of thermo-physical properties and mechanical properties,such as low density,excellent thermal conductivity,tailorable coefficient of thermal expansion,and high specific strength[1−4].Additionally,Al−high Si alloys also have good plating ability and laser weldability. There characteristics make Al−high Si alloys attractive for electronic packaging applications in the field of thermal management,especially for chip boxes to protect electronic devices from outdoor environments[5].It is well known that the properties of Al−high Si alloys are determined by the size,shape and distribution of Si phase,including primary Si and eutectic Si phase[6,7].The application of ingot metallurgy(IM)Al−high Si alloys is highly limited by the formation of the coarse and irregular primary Si phase as well as the lager needle-like eutectic Si phase.These microstructural characteristics lead to stress concentration and are detrimental to the mechanical properties and laser weldability. Therefore,a simple and effective route to refine and modify the Si phase is essential to the wide application of Al−high Si alloys.Lots of methods have been employed in the preparation of Al−high Si alloys,such as semi-solid forming[8],melt infiltration[9],ingot metallurgy with modifiers[10,11],powder metallurgy[12], rapid solidification[13]and the recently developed selective laser melting[14,15].According to the literatures,the rapid solidification route is more feasible for mass manufacturing of Al−high Si alloys for thermal management due to the advantages of high efficiency,remarkable refinement effect and ingots with large size.JIA et al[13]reported that the spray deposited Al−50Si alloy can be completely densified by hot isostaticCorresponding author:Jun FANG;Tel:+86-551-65748315;E-mail:******************DOI:10.1016/S1003-6326(21)65521-81003-6326/©2021The Nonferrous Metals Society of China.Published by Elsevier Ltd&Science PressJun FANG,et al/Trans.Nonferrous Met.Soc.China31(2021)586−594587 pressing(HIP)at570°C.Al alloys with Si contentof22%−50%were prepared by gas atomizationfollowed by hot pressing,and near fully densemicrostructure and excellent properties wereobtained[16].Al−30Si alloy prepared by spraydeposition can also be densified by hot pressing,and a continuous network of globular Si phase andan interpenetrating Al matrix were achieved[17].The Al−50Si alloy is widely used as electronicpackaging boxes,which has a high volume fractionof Si and approximately pure Al matrix.However,its strength should be improved in order to expandits application[5].The previous works of Al−highSi alloys for thermal management have beenfocused on the manufacturing technologies,parameters,and the subsequent properties.Generally,the properties of ingot metallurgyAl−high Si alloys can be modified through alloying,such as the A356,A380,and A390alloys[18].BEFFORT et al[19]reported that mechanicalproperties of the squeeze cast60vol.%SiC p/Alcomposites were also highly determined by the Zn,Cu and Mg elements in the Al matrix.However,less attention has been paid to the alloy compositionand the relationship between microstructuralevolution and properties of the Al−50Si alloy.Accordingly,in this work,Al−50Si,Al−50Si−1Cu and Al−50Si−1Mg alloys for electronicpackaging in thermal management weresuccessfully fabricated by rapid solidification(gasatomization)and powder metallurgy(hot pressing)route,and the microstructural characteristics,mechanical properties(tensile and bendingstrength)and thermo-physical properties wereparisons between the effect of Cu andMg addition on the Al−50Si alloys were analyzed based on the microstructural observations and macro-property tests.2ExperimentalPolycrystalline pure Si(99.9%,all the alloy compositions are in mass fraction unless otherwise mentioned)and pure Al(99.95%)were inductively melted at approximately1250°C.Then,Al−50Si pre-alloy powder was fabricated through a nitrogen gas atomization process,and the morphology of the powder particles is shown in Fig.1(a).After mechanical sieving,the Al−50Si pre-alloy powder with particle size less than74μm was mixed with Fig.1SEM morphologies of gas-atomized Al−50Si pre-alloy powder(a),electrolytic Cu powder(b)and inert gas-atomized Mg powder(c)with different shapes 1wt.%electrolytic Cu powder and1wt.%inertgas-atomized Mg powder,respectively.Mechanical mixing was applied for6h in the atmosphere of Ar with the mass ratio of ball to powder of4:1.The Cu and Mg powders having dendritic and spherical shapes are displayed in Figs.1(b)and1(c), respectively.The mixed powder was cold compacted at300MPa and hold for20s,and billets with relative density of approximately78%were obtained.Hot press sintering was employed on the cold compacted billets and held at560°C forJun FANG,et al/Trans.Nonferrous Met.Soc.China31(2021)586−594 58860min at45MPa.Finally,the samples with dimensions of d50mm×10mm were obtained. The hot-pressed alloys were solid solutionized at 500°C for4h and then aged at160°C for24h. Details of the fabrication process is reported in the previous work[16].Chemical compositions of the as-fabricated Al−50Si−X(X=0,Cu,and Mg)alloys were detected using an inductively coupled plasma optical emission spectrometer(IC-OES),and the results are illustrated in Table1.Morphologies of the Al−50Si pre-alloy powders,Cu powder and Mg powder were detected using a scanning electron microscope(SEM,Quanta−200).Hot-pressed samples for microstructural characterization were cut,ground,polished,and etched with Keller’s reagent.Field emission scanning electron microscope(FESEM,Sirion200)equipped with an energy dispersive spectroscopy(EDS)detector was used in the observation of microstructural details. The sizes of Si phase and secondary phases were measured using ImageJ software.The phases present in the Al−high Si alloys were further analyzed using X-ray diffraction(XRD)at a scanning angle of25°−80°.The room temperature tensile and three-point bending tests of samples were carried out on an electronic universal material testing machine (MTS850).The tensile specimens were made into a dumbbell shape according to the standard GB T228—2010with a gauge diameter of6mm. The dimensions of the three-point bending specimen are3mm×10mm×50mm.The tensile fractured surfaces of the specimens were observed using SEM.The Brinell hardness test of the alloy was performed at a load of7.35kN for30s on the polished samples.All the tensile and bending tests were repeated three times to obtain good reproducibility of data.Under the argon atmosphere,coefficient of thermal expansion of the Al−50Si−X alloys was measured in the temperature range of25−300°C using laser flash and calorimetric methods (NETZSCH LFA427/3/G).The sample has a size of 20mm×5mm×5mm and was required to be parallel and smooth at both ends.Thermal conductivity of the three kinds of alloys was performed on cylindrical slice specimens with dimensions of d10mm×3mm using NETZSCH DIL402C.Density of the alloys was measured by Archimedes method using a balance with the accuracy of0.1mg.3Results3.1Microstructural characteristicsTypical microstructures of the as-atomized Al−50Si pre-alloy powder and the hot-pressed Al−50Si−X alloys are shown in Fig.2.It can be seen from Fig.2(a)that the primary Si phase is highly refined to have a block-like morphology due to the large solidification rate and undercooling nature of gas atomization.The eutectic Si phase is also refined remarkably and its shape changes from needle-like with large aspect ratio in the as-cast alloy to bar-like with a low aspect ratio in the as-atomized powder.However,the primary Si seems to distribute mostly at the periphery of powder particles owing to the solidification sequence[20].After hot press,the gas-atomized Al−50Si pre-alloy powder is well densified and a pore-free microstructure is obtained,as shown in Figs.2(b−d). High density of defects,such as pores and cracks were observed in the Al−50Si alloy prepared by ingot metallurgy[21].Consequently,the measured density of the hot-pressed samples is near to the theoretical value.As the density of Cu(8.9g/cm3) is higher than that of Al(2.7g/cm3)while the density of Mg(1.7g/cm3)is lower than that of Al, the addition of Cu or Mg leads to a slight variation of density in the Al−50Si−X alloys.Table1Compositions of rapidly solidified(gas-atomized)and hot-pressed Al−50Si−X alloys measured by ICP-OES (wt.%)Material Si Mg Cu Zn Fe Mn Ti AlAl−50Si50.5<0.01<0.01<0.010.040.02<0.01Bal.Al−50Si−1Cu50.30.05 1.03<0.010.030.01<0.01Bal.Al−50Si−1Mg49.7 1.030.02<0.010.050.01<0.01Bal.Jun FANG,et al/Trans.Nonferrous Met.Soc.China31(2021)586−594589 Fig.2SEM morphologies of gas-atomized Al−50Si pre-alloy powder(a)and as-fabricated Al−50Si alloy(b),Al−50Si−1Cu alloy(c)and Al−50Si−1Mg alloy(d)having similar characteristics of Si phaseIt is seen that a semi-continuous networkstructure with smooth surface of the Si phase isformed in the Al matrix,as seen in Figs.2(b−d).The distribution of Si phase is quite homogeneousas compared with that of the as-atomized powder.Such characteristics of Si phase are highly differentfrom those of the as-cast Al−high Si alloys whichhave coarse and irregular(bar-like,plate-like,star-like,etc)primary Si with sharp corners as wellas needle-like eutectic Si with a large aspectratio[11,21].Furthermore,it is interesting to findthat the eutectic Si is completely absent in thehot-pressed samples due to the diffusion-controlled growth of Si phase and the Si−Si phase clustering in the solid-state sintering.There is no obvious change of the Si phase in the fabricated Al−50Si alloys with and without Cu(Mg)addition besides a little lower degree of the semi-continuous structure.X-ray diffractions were performed to detect the phases presented in the hot-pressed Al−50Si−X alloys,and the results are displayed in Fig.3.It is seen that the diffraction peaks ofα(Al)and Si phase are clearly observed in the samples.With the addition of Cu or Mg,small amounts of Al2Cu and Fig.3XRD patterns of as-fabricated Al−50Si−X alloys showing Al2Cu and Mg2Si secondary phases formed in Al−50Si−Cu/(Mg)alloys:(a)Al−50Si;(b)Al−50Si−1Cu;(c)Al−50Si−1MgMg2Si secondary phases are formed in the Al−50Si−Cu(Mg)alloys.It is noted that,different from the Al−50Si−1Cu alloy,no AlMg secondary phases are formed in the Al−50Si−1Mg alloy. However,as the content of Cu or Mg is only1%, the diffraction peaks of the Al2Cu and Mg2Si phases are not remarkable.Jun FANG,et al/Trans.Nonferrous Met.Soc.China 31(2021)586−594590To further investigate the secondary phases formed in the Al−50Si−Cu(Mg)alloys,magnified SEM observations were conducted and the results are shown in Fig.4.Other than the large Si particles,small needle-like Al 2Cu phase and bar-like Mg 2Si phase are present in the Al−50Si−Cu(Mg)alloys.This result is in consistent with the XRD patterns presented in Fig.3.Although the average sizes of the Al 2Cu and Mg 2Si secondary phases are less than 1μm,most of the Mg 2Si phase is larger than the Al 2Cu phase.Additionally,most of the Al 2Cu phases are dispersed in the center of the Al matrix.However,the Mg 2Si phase seems to distribute mostly near the surface of Si particles.This phenomenon can be attributed to the larger diffusion rate and supersaturation of Mg than those of Si in the Almatrix.Fig.4SEM morphologies and distribution of Al 2Cu (a)and Mg 2Si (b)secondary phases present in Al−50Si−Cu(Mg)alloys3.2Mechanical propertiesThe room temperature tensile tests were performed on the hot-pressed Al−50Si alloys with and without Cu(Mg)addition,and the tensile curves are depicted in Fig.5.The stress−strain response of the Al−50Si alloy is different from that containing Cu and Mg.A very slight plastic deformation of approximately 0.5%strain isobserved in the Al−50Si alloy.Remarkably enhanced ultimate tensile strength (UTS)is achieved in the Al−50Si−1Cu and Al−50Si−1Mg alloys.The plastic behavior is less evident,approximately 0.3%strain to fracture,with the addition of Cu or Mg.This phenomenon indicates that the addition of Cu(Mg)is beneficial to improving the strength of Al−50Si alloy but detrimental to the plasticity of the alloy.Additionally,the slope of the tensile stress−strain response of the Cu(Mg)-contained alloys becomes flatter and higher than that of the Al−50Si alloy,suggesting that the addition of Cu(Mg)also enhances the elastic modulus of thealloy.Fig.5Tensile stress−strain response of rapidly solidified Al−50Si−X alloys at room temperatureAverage values of the tensile strength,bending strength and hardness of the Al−50Si−X alloys were obtained from five parallel tests,and the results are shown in Fig.6.The strength of the Al−50Si alloy is significantly improved with the addition of Cu(Mg).Compared with the reference sample,the addition of 1%Cu raises the tensile and bending strength from 185.7and 288.6MPa to 236.2and 390.5MPa,with increments of 27.2%and 35.3%,respectively.Similarly,the addition of 1%Mg results in an enhancement of tensile and bending strength by 24.5%and 29.0%,respectively.At the same time,the addition of alloying elements also increases the hardness of the Al matrix.From Fig.6,it is also found that the strengthening effect of Cu is slightly higher than that of Mg.This phenomenon can be attributed to the fine and homogeneous distribution of the Al 2Cu secondary phase at the center of the Al matrix.Additionally,according to the image analysis from SEM results,the average size of Al 2Cu phase is a little smallerJun FANG,et al/Trans.Nonferrous Met.Soc.China31(2021)586−594591Fig.6Tensile strength,bending strength and hardness of rapidly solidified Al−50Si−X alloysthan that of the Mg2Si phase,which may also contribute to the higher strength of the Al−50Si−1Cu alloy.Tensile fractured morphologies of Al−50Si−X alloys are displayed in Fig.7.All samples show a clear brittle fracture feature.It is seen that the fracture planes of the alloys are vertical to the tensile direction and no visible macro-ductility fracture is observed.As seen from Fig.7(a),the crack source of the alloy with rather flat morphology is clearly observed.The crack progresses rapidly in a linear way through the sample when external pressure is applied.Figures 7(b−d)show that the Al matrix fractures by ductile rupture with tearing ridge while the Si phase fractures by cleavage surface.As there is a high volume fracture of Si phase(approximately53.7%) with semi-continuous structure,the Si particle dominated brittle fracture is the main mode of the Al−50Si alloys.The previous observation suggests that the crack tip moves through brittle fracture of the Si particles and finishes by ductile fracture of the Al matrix[22].Generally,metal matrix composites(MMCs)reinforced with high volume of reinforcement fracture in such particle dominatedFig.7Low magnification micrograph showing crack source of Al−50Si alloy(a)and high magnification micrographs of Al−50Si alloy(b),Al−50Si−1Cu alloy(c)and Al−50Si−1Mg alloy(d)Jun FANG,et al/Trans.Nonferrous Met.Soc.China 31(2021)586−594592mode [23,24].Additionally,dimples with small size are observed in the alloys due to the refined microstructure as a result of rapid solidification and solid-state sintering.However,three kinds of alloys show typical brittle fracture,and the difference among fractured morphologies is less visible.3.3Thermo-physical propertiesVariations of coefficient of thermal expansion (CTE)of the Al−50Si−X alloys as a function of temperature in the range of 25−300°C are shown in Fig.8.It is observed that the coefficient of thermal expansion increases linearly with the increase of testing temperature.The Al−high Si alloys can be regarded as Si particle reinforced Al matrix composites (Si p /Al)and the coefficient of thermal expansion of the alloy is mainly determined by the properties of the Al matrix and Si phase and the volume fraction of the Si phase according to the rule of mixture (ROM).As seen from Fig.2,there is little deviation of the volume fraction,size,and morphology of Si phase.Consequently,the coefficients of thermal expansion of the Al−50Si−X alloys are determined mainly by the properties of Al matrix.Owing to the presence of Al 2Cu and Mg 2Si secondary phase having lower coefficient of thermal expansion,the total thermal expansion of Al−50Si alloys is reduced.JIA et al [13]reported that no plastic deformation occurs in the Al matrix at low temperatures.The expansion of the alloys is caused by the combined expansion of the Al matrix and Si phase and results in the linearly increased coefficient of thermal expansion with increasingtemperature.Fig.8Coefficient of thermal expansion of rapidly solidified Al−50Si−X alloys in temperature range of 25−300°CThermal conductivity of the Al−50Si−X alloys is illustrated in Fig.9.Owing to the rapid solidification nature of gas atomization and the diffusion-controlled growth of Si phase during hot pressing,the Si phase has a semi-continuous structure with smooth surface,which contributes to the excellent thermal conductivity of the Al−50Si alloy,146.2W·m −1·K −1.At the same time,Si has low solid solubility in the Al matrix with equilibrium state,and a near pure Al matrix after hot pressing may also help for achieving high thermal conductivity of the alloy.However,the formation of the Al 2Cu and Mg 2Si secondary phases in the Al−50Si−Cu(Mg)alloys has a scattering effect on the free electron motion and hinders the thermal conduction [25].Consequently,the thermal conductivities of the Al−50Si alloy containing 1%Cu and 1%Mg are reduced by 7.3%and 6.8%,respectively.In comparison with the exceptionally improved strength of the Al−50Si alloy,this reduction of thermal conductivity is within the acceptable limit (≥120W·m −1·K −1).Fig.9Thermal conductivity of rapidly solidified and hot-pressed Al−50Si−X alloys at room temperature4Conclusions(1)Gas atomization endows the pre-alloyed Al−50Si alloy powder with highly refined primary and eutectic Si phase,and in combination with the subsequent solid-state hot-pressing,the Si phase with semi-continuous network structure is obtained.By adding 1%Cu or 1%Mg,Al 2Cu or Mg 2Si secondary phases are observed,respectively,but the influence on the Si phase characteristics is limited.(2)Tensile strength,bending strength and hardness of the Al−50Si alloys are significantlyJun FANG,et al/Trans.Nonferrous Met.Soc.China31(2021)586−594593improved with the addition of Cu or Mg, respectively,which is attributed to the strengthening effect of the fine secondary phases.The effect of Cu on mechanical properties is more remarkable compared with that of Mg.All the Al−50Si−X alloys show typical brittle fracture features having a clear cleavage surface.(3)The addition of Cu(Mg)is helpful for reducing the coefficient of thermal expansion of the Al−50Si−X alloys,but detrimental to the thermal conductivity.However,negligible difference in thermo-physical properties is observed in the Al−50Si−Cu(Mg)alloys.References[1]HOGG S C,LAMBOURNE A,OGILVY A,GRANT P S,Microstructural characterisation of spray formed Si−30Al for thermal management applications[J].Scripta Materialia, 2006,55(1):111−114.[2]KIMURA T,NAKAMOTO T,MIZUNO M,ARAKI H.Effect of silicon content on densification,mechanical and thermal properties of Al−x Si binary alloys fabricated using selective laser melting[J].Materials Science and Engineering A,2017,682:593−602.[3]ZHANG Wen-long,DING Dong-yan,GAO Ping.Highvolume fraction Si particle-reinforced aluminium matrix composites fabricated by a filtration squeeze casting route[J].Materials&Design,2016,90:834−838.[4]WANG Qian,ZHU Lin,CHEN Xiao-guang,YAN Jiu-chun,XIE Rui-shan,LI Pei-hao,WANG Zhi-hua,WANG Zhi-qi, LI Yun-tao,ZHOU Xiao-yu.Si particulate-reinforced ZnAl based composites joints of hypereutectic Al50Si alloys byultrasonic-assisted soldering[J].Materials&Design,2016, 107:41−46.[5]MAUDUIT D,DUSSERRE G,CUTARD T.Probabilisticrupture analysis of a brittle spray deposited Si−Al alloy under thermal gradient:Characterization and thermoelastic sizing guidelines[J].Materials&Design,2016,95:414−421.[6]CHEN Xu,ZHONG Yun-bo,ZHENG Tian-xiang,SHENZhe,WANG Jiang,FAN Li-jun,ZHAI Yong,PENG Ming-hu,ZHOU Bang-fei,REN Wei-li,LEI Zuo-sheng, REN Zhong-ming,HE Qiong.Refinement of primary Si in the bulk solidified Al−20wt.%Si alloy assisting by high static magnetic field and phosphorus addition[J].Journal of Alloysand Compounds,2017,714:39−46.[7]WANKHEDE D M,NARKHEDE B E,MAHAJAN S K,CHOUDHARI C M.Influence of pouring temperature and external chills on mechanical properties of aluminum silicon alloy castings[J].Materials Today:Proceedings,2018,5(9): 17627−17635.[8]WARD P J,ATKINSON H V,ANDERSON P R G,ELIAS LG,GARCIA B,KAHLEN L,RODRIGUEZ-IBABE J M.Semi-solid processing of novel MMCs based onhypereutectic aluminium-silicon alloys[J].Acta Materialia,1996,44(5):1717−1727.[9]CHEN Yu-yong,CHUNG D D L.Silicon-aluminiumnetwork composites fabricated by liquid metal infiltration[J].Journal of Materials Science,1994,29(23):6069−6075. [10]RAGHUKIRAN N,KUMAR R.Effect of scandium additionon the microstructure,mechanical and wear properties of the spray formed hypereutectic aluminum−silicon alloys[J].Materials Science and Engineering A,2015,641:138−147.[11]YU Wen-hui,ZHANG Yong,YAN Ting-liang,LIU Yue,JIANG Ao-lei,ZHENG Hong-liang,TIAN Xue-lei.Enhanced nucleation of primary silicon in Al−20wt.%Si alloy with Ni−Si inoculation[J].Journal of Alloys and Compounds,2017,693:303−307.[12]LIU Y Q,WEI S H,FAN J Z,MA Z L,ZUO T.Mechanicalproperties of a low thermal-expansion aluminum/siliconcomposite produced by powder metallurgy[J].Journal ofMaterials Science&Technology,2014,30(4):417−422. [13]JIA Yan-dong,CAO Fu-yang,SCUDINO S,MA Pan,LIHai-chao,YU Lei,ECKERT J,SUN Jian-fei.Microstructure and thermal expansion behavior of spray-deposited Al−50Si[J].Materials&Design,2014,57:585−591.[14]JIA Qing-bo,ROMETSCH P,KURNSTINER P,CHAO Qi,HUANG Ai-jun,WEYLAND M,BOURGEOIS L,WU Xin-hua.Selective laser melting of a high strengthAl−Mn−Sc alloy:Alloy design and strengtheningmechanisms[J].Acta Materialia,2019,171:108−118. [15]KANG Nan,CODDET P,LIAO Han-lin,CODDET C.Macrosegregation mechanism of primary silicon phase in selective laser melting hypereutectic Al−high Si alloy[J].Journal of Alloys and Compounds,2016,662:259−262. [16]CAI Zhi-yong,ZHANG Chun,WANG Ri-chu,PENGChao-qun,QIU Ke,FENG Yan.Preparation of Al−Si alloys by a rapid solidification and powder metallurgy route[J].Materials&Design,2015,87:996−1002.[17]WANG Feng,XIONG Bai-qing,ZHANG Yong-an,ZHUBao-hong,LIU Hong-wei,WEI Yan-guang.Microstructure, thermo-physical and mechanical properties of spray-deposited Si−30Al alloy for electronic packaging application [J].Materials Characterization,2018,59(10):1455−1457. [18]DAMAV ANDI E,NOUROUZI S,RABIEE S M,JAMAATIR.Effect of ECAP on microstructure and tensile properties of A390aluminum alloy[J].Transactions of NonferrousMetals Society of China,2019,29(5):931−940.[19]BEFFORT O,LONG Si-yuan,CAYRON C,KUEBLER J,BUFFAT P A.Alloying effects on microstructure and mechanical properties of high volume fraction SiC particle reinforced Al-MMCs made by squeeze casting infiltration[J].Composites Science and Technology,2007,67(3):737−745.[20]KALAY Y E,CHUMBLEY L S,ANDERSON I E,NAPOLITANO R E.Characterization of hypereutectic Al−Si powders solidified under far-from equilibrium conditions[J].Metallurgical and Materials Transactions A,2007,38(7): 1452−1457.[21]CAO Fu-yang,JIA Yan-dong,PRASHANTH K G,MA Pan,LIU Jing-shun,SCUDINO S,HUANG Feng,ECKERT J, SUN Jian-fei.Evolution of microstructure and mechanical properties of as-cast Al−50Si alloy due to heat treatment and P modifier content[J].Materials&Design,2015,74: 150−156.Jun FANG,et al/Trans.Nonferrous Met.Soc.China31(2021)586−594 594[22]CAI Zhi-yong,ZHANG Chun,WANG Ri-chu,PENGChao-qun,WU Xiang.Effect of copper content on microstructure and mechanical properties of Al/Si p composites consolidated by liquid phase hot pressing[J].Materials&Design,2016,110:10−17.[23]ZHANG Qi-guo,ZHANG Hong-xiang,GU Ming-yuan,JINYan-ping.Studies on the fracture and flexural strength of Al/Si p composite[J].Materials Letters,2004,58: 3545−3550.[24]BUFFIERE J Y,MAIRE E,VERDU C,CLOETENS P,PATEYRON M,PEIX G,BARUCHEL J.Damage assessment in an Al/SiC composite during monotonic tensile tests using synchrotron X-ray microtomography[J].Materials Science and Engineering A,1997,234−236: 633−635.[25]BLANK M,SCHNEIDER G,ORDONEZ-MIRANDA J,WEBER L.Role of the electron-phonon coupling on the thermal boundary conductance of metal/diamond interfaces with nanometric interlayers[J].Journal of Applied Physics, 2019,126:165302.热管理用快速凝固Al−50Si−Cu(Mg)合金的力学和热物理性能方军，钟永辉，夏明旷，张凤伟中国电子科技集团第四十三所，合肥230088摘要：通过加入Cu或Mg合金元素提高高硅铝合金的力学性能。
硫化铜纳⽶棒的低热固相合成及其光学性能V o l.26⾼等学校化学学报 N o.4 2005年4⽉ CH E M I CAL JOU RNAL O F CH I N ESE UN I V ER S IT IES 620～622 硫化铜纳⽶棒的低热固相合成及其光学性能周 杰,贾殿赠,刘 浪(新疆⼤学应⽤化学研究所,乌鲁⽊齐830046)摘要 在表⾯活性剂PEG2400存在的条件下,以醋酸铜和硫代⼄酰胺为原料,利⽤低热固相化学反应,⼀步制备出分散均匀的硫化铜纳⽶棒.X射线粉末衍射和能量散射X射线能谱分析证明,产物为纯六⾓相的硫化铜.透射电镜和扫描电镜形貌分析表明,产物为棒状,直径为80～100nm,长度为200～500nm.紫外2可见光谱和光致发光光谱表明,硫化铜纳⽶棒的紫外和荧光最⼤吸收和发射波长与常规硫化铜相⽐均发⽣了明显的蓝移,表明所制备的硫化铜纳⽶棒具有良好的光学性质.关键词 纳⽶棒;硫化铜;固相反应;光学性能中图分类号 O613.5;O614.24 ⽂献标识码 A ⽂章编号 025120790(2005)0420620203过渡⾦属硫化物因在半导体、发光装置及超导等⽅⾯具有潜在的应⽤价值⽽引起关注[1～3].纳⽶材料由于其较⼤的⽐表⾯以及量⼦尺⼨效应等在光吸收、发射、热电及光电等⽅⾯表现出与常规体相材料不同的特殊性能.硫化铜作为⼀种重要的半导体材料,被⼴泛地应⽤于热电偶、光记录、滤光器和太阳能电池等[4～6]⽅⾯.⽬前,硫化铜纳⽶材料的制备主要有化学沉积法[7]、声化学法[8]、⽔热合成法[9]和微波辐射法[10],这些⽅法⼀般需要较长的反应时间、较⾼的温度和压⼒以及特殊的反应装置.因此,寻找反应条件温和,易于操作,适⽤范围较⼴,成本低的制备⼀维纳⽶材料的新⽅法尤为重要.低热固相化学反应法近年来⽇益受到重视,并在材料合成⽅⾯显⽰出较⼤的优势[11].⽬前,已⽤低热固相化学反应法和固相反应前驱体法成功地制备了纳⽶氧化物、复合氧化物、硫化物和卤化物等[12～15].本⽂报道了在表⾯活性剂PEG2400存在的条件下,以醋酸铜和硫代⼄酰胺为原料,利⽤低热固相化学反应,⼀步制备出分散均匀的硫化铜纳⽶棒.制备的硫化铜纳⽶棒直径为80～100nm,长度为200～500nm.产物具有较好的晶型和较⾼的产率.1 实验部分1.1 试剂与仪器所⽤试剂均为分析纯.醋酸铜[Cu(A c)2?H2O,北京试剂⼚];PEG2400和硫代⼄酰胺(CH3CSN H2)(上海化学试剂公司);⼄醇(中国医药公司).⽤⽇本⽇⽴公司的H2600型透射电⼦显微镜(T E M)(加速电压为100kV)和德国L EO公司的L EO1430V P型扫描电⼦显微镜(SE M)对样品进⾏形貌分析;⽤⽇本马克公司的M XP18A H F 型X射线衍射仪(XRD)(铜靶,扫描电压为40kV,扫描电流为100mA,使⽤⽔平单⾊器,扫描步长为0101°,扫描范围为10°～80°)进⾏成分和物相分析;⽤⽇本⽇⽴U23010型紫外2可见光谱仪(UV2V is)和上海分析仪器总⼚的棱光970CR T型荧光光谱仪(PL)进⾏光学性质的分析.1.2 实验⽅法称取319930g Cu(A c)2?H2O(0102m o l),将其置于玛瑙研钵中充分研磨后,加⼊5mL PEG2 400混匀,再加⼊充分研磨的310052g CH3CSN H2(0104m o l),将混合物研磨反应5m in,体系由蓝⾊收稿⽇期:2004205220.基⾦项⽬:国家⾃然科学基⾦(批准号:20161003,20366005)资助.联系⼈简介:贾殿赠(1962年出⽣),男,教授,主要从事低热固相反应及纳⽶材料研究.E2m ail:jdz@/doc/ae5277aecd22bcd126fff705cc175********e88.html变成墨绿⾊,并有醋酸⽓味放出,继续研磨体系颜⾊加深,40m in 后完全变成⿊⾊.将其置于80℃恒温⽔浴中加热12h ,⾃然冷却,⽤蒸馏⽔和⽆⽔⼄醇交替洗涤3次,真空⼲燥,得⿊⾊产物.2 结果与讨论2.1 XR D 分析F ig .1 XR D pattern of the as -preparedcopper sulf ide nanorods 由产物的粉末X 射线衍射图(图1)可见,所有的衍射峰都指标化为纯六⾓相的CuS (JCPD S N o .0620464),未观察到如Cu 2S ,CuO 和S 等的杂峰.计算晶胞参数为a =013796nm ,c =11631nm ,符合⽂献值(a =013792nm ,c =11634nm ).固相产物的衍射峰均有明显的宽化,说明产物具有⼩的尺⼨.产物的能量散射X 射线能谱(EDX )分析结果表明,产物仅有Cu 和S 存在,在样品表⾯未检测到其它物质,表明样品是纯的硫化铜.2.2 TE M 和SE M图2(A )为表⾯活性剂PEG 2400存在的条件下,以醋酸铜和硫代⼄酰胺为原料制得的固相反应产物的T E M 照⽚.可见,在表⾯活性剂存在下,固相反应产物为平均直径约80～100nm ,长约200～500nm 的棒.图2(B )为固相反应产物的SE M 照⽚,进⼀步证明了CuS 的棒状结构.与不加表⾯活性剂所得的固相反应产物的形貌相⽐[11],表⾯活性剂聚⼄⼆醇在该固相化学反应制备硫化铜纳⽶棒的过程中起到关键作⽤.众所周知,聚⼄⼆醇具有长的链状结构,反应物可能在聚⼄⼆醇的链状结构表⾯进⾏反应,表⾯活性剂聚⼄⼆醇修饰了原来的固态反应界⾯,并诱导产物粒⼦沿某⼀⽅向定向⽣长,最终得到纳⽶棒.F ig .2 TE M (A )and SE M (B )i m ages of the copper sulf ide nanorods2.3 紫外和荧光分析F ig .3 UV -V is spectra (A )and f luorescen t spectra (B )of as -prepared copper sulf ide nanorods图3(A )为利⽤固相化学反应制得的产物的紫外2可见光谱图(谱线a ),可见,产物的吸收边和吸收峰分别在348和335nm 处,其相应的带隙能分别为3157和3171eV ,相对于本体相的吸收峰(谱线126N o .4周 杰等:硫化铜纳⽶棒的低热固相合成及其光学性能226 ⾼等学校化学学报V o l.26b,344nm,3161eV)有⼀定程度的蓝移,表明存在纳⽶材料的量⼦限域效应.图中产物的紫外吸收性能增强的主要原因是产物的粒径减⼩,其⽐表⾯积增⼤,表⾯缺陷数增多.图3(B)为利⽤固相化学反应制备的产物(谱线a)和本体相(谱线b)的荧光发射光谱图,当激发波长为350nm时,产物在640nm 处有发射峰,保留了半导体的荧光发光特性,并有别于其本体相.综上,在表⾯活性剂PEG2400存在的条件下,以醋酸铜和硫代⼄酰胺为原料,利⽤低热固相化学反应,⼀步即可制得⼀维硫化铜纳⽶棒.这是⼀种简单、快速、有效的制备⼀维硫化物半导体纳⽶材料的⽅法.样品的紫外和荧光分析结果表明,产物具有良好的光学性能,在发光装置⽅⾯具有潜在的应⽤前景.参 考 ⽂ 献[1] Shao M.W.,W u Z.C.,Gao F.et a l..J.C rystal Grow th[J],2004,260:63—66[2] W ang S.H.,Yang S.H..Chem.Phys.L ett.[J],2000,322:567—571[3] D I N G Yi(丁 轶),L I U R en2M ao(刘仁茂),ZHAN G Zu2D e(张祖德)et a l..Chem.J.Ch inese U niversities(⾼等学校化学学报)[J],2000,21(3):344—346[4] Ch ivers T..J.Chem.Soc.D alton T rans.[J],1996,7:1185—1194[5] L ak shm ikum ar S.T..So l.Energy M ater.So l.Cell[J],1994,32:7—10[6] Ko rzhuev.A.A..F iz.Kh i m.O brab.M ater.[J],1991,3:131—136[7] W AN G J iu(王 九),CH EN Bo2Shui(陈波⽔),HOU B in(侯 滨)et a l..Ch inese J.Ino rg.Chem.(⽆机化学学报)[J],2001,17:275—278[8] Kum ar R.V.,Palch ik O.,Ko ltyp in Y.et a l..U ltrason.Sonochem.[J],2002,9:65—70[9] Chen X.Y.,W ang Z.H.,W ang X.et a l..J.C rystal Grow th[J],2004,263:570—574[10] L iao X.H.,Chen N.Y.,Xu S.et a l..J.C rystal Grow th[J],2003,252:593—598[11] Ye X.R.,J ia D.Z.,Yu J.Q.et a l..A dv.M ater.[J],1999,11:941—942[12] J ia D.Z.,Yu J.Q.,X ia X..Ch inese Sci.Bull.[J],1998,43(7):571—573[13] SH EN R u2Juan(沈茹娟),J I A D ian2Zeng(贾殿赠),Q I AO Yong2M in(乔永民)et a l..J.Ino rg.M ater.(⽆机材料学报)[J],2001,16(4):625—629[14] L i F.,Zheng H.G.,J ia D.Z.et a l..M ater.L ett.[J],2002,53:282—286[15] Cao Y.L.,J ia D.Z.,L iu L.et a l..T rans.M ater.R es.Soc.Jap.[J],2004,29:103—105Optica l Properties and Syn thesis of Copper Sulf ide Nanorods bySol id-sta te Reaction a t L ow Hea ti ng Tem pera tureZHOU J ie,J I A D ian2Zeng3,L I U L ang(Institu te of A pp lied Che m istry,X inj iang U niversity,U rum qi830046,Ch ina)Abstract W ell2dispersed copper acetam ide su lfide nano rods w ere p repared by m ean s of the so lid2state reacti on of copp er acetate and th i oacetam ide in the p resence of su rfactan t PEG2400at a low heating tem p eratu re.T he pow der X2ray diffracti on pattern and EDX indicated that the p roduct w as pu re hexagonal p hase copper su lfide nano rods.T he m o rpho logy of the nano rods w as characterized by tran s m issi on electron m icro scop y(T E M)and scann ing electron m icro scopy(SE M).T he diam eter and length of CuS nano rods ob tained are80—100nm and200—500nm,respectively.A nd the m ax i m um p eak po siti on s of ab so rp ti on sp ectrum and excited spectrum of CuS nano rods w ere ob served to sh ift tow ards h igher w ave num ber com pared w ith tho se of the bu lk CuS.T he resu lts clearly show the fine op tical p roperties of CuS nano rods.Keywords N ano rod;Copper su lfide;So lid2state reacti on;Op tical p roperty(Ed.:K,G)619 Syn thesis of M oS2 Carbon Co m posite NanotubesSON G Xu2Chun3,ZH EN G Y i2Fan,HAN Gu i,Y I N H ao2Yong,CAO Guang2ShengChe m.J.Ch inese U niv.,2005,26(4),617_619622 Optica l Properties and Syn thesis of Copper Sulf ide Nanorods by Sol id-sta te Reactiona t L ow Hea ti ng Te m pera tureZHOU J ie,J I A D ian2Zeng3,L I U L angChe m.J.Ch inese U niv.,2005,26(4),620_622627 Prepara tion of Nano-iron N itr ide by Te m pera ture-programm ed Reaction fro mNano-iron Ox idesZH EN G M ing2Yuan,CH EN G R u i2H ua,CH EN X iao2W ei,L I N ing,CON G Yu,W AN G X iao2Dong,ZHAN G T ao3Che m.J.Ch inese U niv.,2005,26(4),623_627637 Starch Nanoparticle a s Tran sgen ic Veh icle M ed i a ted by Ultra soundL I U Jun,L I U Xuan2M ing3,X I AO Su2Yao, TON G Chun2Y i,TAN G Dong2Y ing,ZHAO L i2J ianChe m.J.Ch inese U niv.,2005,26(4),634_637641 Ana lysis of Nord iterpeno id A lka lo ids i n Roots of A con itum kusnezof f ii by Electrospray Ion iza tion Tande m M a ss Spectrom etryXU Q ing2Xuan,W AN G Yong,L I U Zh i2Q iang,L I U Shu2Y ing3,T I AN ChengChe m.J.Ch inese U niv.,2005,26(4),638_641646 M olecular I m pr i n ti ng-che m ilu m i nescenceD eter m i na tion of Ana lg i nH E Yun2H ua,L U J iu2R u3,ZHAN G Hong2Ge,DU J ian2X iuChe m.J.Ch inese U niv.,2005,26(4),642_646。
表面技术第52卷第11期硫醇改性超疏水铜表面的耐候性及失效机理杜文倩1，王德辉1,2，余华丽1，李罗慧子1，罗静1，邓旭1*（1.电子科技大学 基础与前沿研究院，成都 610054；2.中国空气动力研究与发展中心 结冰与防除冰重点实验室，四川 绵阳 621000）摘要：目的探究户外环境中导致硫醇改性超疏水铜表面失效的因素及其超疏水性失效的机制。
方法通过化学刻蚀法在铜表面构筑纳米结构，利用正十二硫醇进行表面改性，得到具有超疏水性的铜表面。
将该表面置于户外进行耐候性研究，并通过4种模拟户外环境实验探究超疏水性失效的原因，包括组合循环实验（循环条件含紫外辐射、淋雨和凝露）、紫外辐射实验、水环境实验和温度实验。
结果超疏水铜表面经过10 d的户外实验后，其接触角由初始状态的158.5°降至131.1°，表明该表面的超疏水性能已失效。
经过2次组合循环实验（每次循环的时间为12 h）、20 d紫外辐射实验及30 d水环境实验后，该表面的接触角分别降至130.3°、124.5°、131.7°，表明该表面均已失去超疏水性；经过40 d高温实验后，表面的超疏水性开始失效。
XPS谱图表明，在超疏水性失效后该表面不存在硫元素，即正十二硫醇已经脱离表面。
结论超疏水铜表面的硫醇分子脱落是超疏水性失效的根本原因。
紫外辐射、水和高温是导致超疏水铜表面超疏水性失效的主要因素。
其中，紫外辐射或水对超疏水性的破坏速度比高温快。
相较于单一因素（紫外辐射、水或高温），三者的协同作用更加速了硫醇分子从表面的脱落，导致超疏水性失效的速度更快。
关键词：超疏水铜表面；化学刻蚀法；硫醇改性；环境耐候性；超疏水性失效；失效机理中图分类号：TG17；TB34 文献标识码：A 文章编号：1001-3660(2023)11-0326-09DOI：10.16490/ki.issn.1001-3660.2023.11.027Weather Resistance and Failure Mechanism of SuperhydrophobicCopper Surface Modified by MercaptansDU Wen-qian1, WANG De-hui1,2, YU Hua-li1, LI Luo-hui-zi1, LUO Jing1, DENG Xu1*(1. Institute of Fundamental and Frontier Sciences, University of Electronic Science and Technology of China,Chengdu 610054, China; 2. Key Laboratory of Icing and Anti/De-icing, China Aerodynamics Research andDevelopment Center, Sichuan Mianyang 621000, China)ABSTRACT: The superhydrophobic copper surface has a very broad application prospect in the fields of self-cleaning, anti-frost, anti-corrosion, reduction of resistance and residue of fluid in copper pipes, etc. However, the weather resistance of the superhydrophobic copper surface in the real application environment is not clear, and there is still short of systematic research.Therefore, the work aims to explore the failure factors of mercaptan modified superhydrophobic copper surface in outdoor environment and the mechanism of superhydrophobic failure. The nanostructure was constructed on the copper surface by收稿日期：2022-11-15；修订日期：2023-02-08Received：2022-11-15；Revised：2023-02-08基金项目：中国空气动力研究与发展中心结冰与防除冰重点实验室开放课题（IADL20210411）Fund：Key Laboratory of Icing and Anti/De-icing of CARDC (IADL20210411)引文格式：杜文倩, 王德辉, 余华丽, 等. 硫醇改性超疏水铜表面的耐候性及失效机理[J]. 表面技术, 2023, 52(11): 326-334.DU Wen-qian, WANG De-hui, YU Hua-li, et al. Weather Resistance and Failure Mechanism of Superhydrophobic Copper Surface Modified by Mercaptans[J]. Surface Technology, 2023, 52(11): 326-334.*通信作者（Corresponding author）第52卷第11期杜文倩，等：硫醇改性超疏水铜表面的耐候性及失效机理·327·chemical etching, and then the surface was modified with 1-dodecanethiol to obtain a superhydrophobic copper surface (SCS).The SCS was placed outdoors to study its weather resistance. The mechanism of failure was explored by four kinds of experiments simulating outdoor conditions. Simulation experiments included combinatorial cycle experiment (cyclic conditions including UV radiation, rain and condensation), UV radiation experiment, water environment experiment and temperature experiment ("high temperature" in this work referred to the high temperature in the outdoor environment). The mechanism of superhydrophobicity failure of SCS was verified by the environmental weather resistance of hydrophobic copper surface (HCS) without rough microstructure. Scanning electron microscope (SEM), contact angle measurement and X-ray photoelectron spectrometer (XPS) were used to characterize the morphology, wettability and chemical composition of SCS, respectively. After10 days of outdoor experiments, the water contact angle of SCS was reduced from the initial 158.5° to 131.1°. According to theSEM diagram and Energy Dispersive Spectrometer (EDS) analysis, impurities containing metal/non-metal inorganic substances and organic matter were absorbed on the surface. The experimental results showed that the superhydrophobic performance of SCS failed and the self-cleaning performance was also lost. The water contact angle of SCS decreased to 130.3°, 124.5° and 131.7° after 2 combinatorial cycle experiments (12 hours of each cycle), 20 days of UV radiation experiment, and 30 days of water environment experiment respectively, indicating that SCS lost its superhydrophobic property. After 20 days of high temperature experiment, SCS still maintained superhydrophobic property, but the superhydrophobic property began to fail after40 days of high temperature experiment. XPS spectrum indicated that sulfur element was no longer present on SCS after thefailure of superhydrophobic properties, i.e. 1-dodecanethiol was removed from SCS. The shedding of mercaptan molecules on the SCS was the root cause of superhydrophobic failure.UV radiation, water and high temperature were the main factors leading to the failure of superhydrophobicity of SCS. The superhydrophobic property of SCS was destroyed faster by UV radiation or water than by high temperature. Compared with the effect of a single factor (UV radiation, water or high temperature), the synergistic effect of the three accelerated the shedding of mercaptan molecules from SCS, thereby accelerating the rate of superhydrophobic failure.By studying the weather resistance and failure mechanism of HCS, it is found that the wettability change trend of HCS is consistent with that of SCS, and the hydrophobicity failure of HCS without rough microstructure is also caused by the shedding of mercaptan molecules on the surface. The results verify that the mercaptan molecule detachment from the surface results in the superhydrophobic failure of SCS.KEY WORDS: superhydrophobic copper surface; chemical etching method; mercaptan modification; environmental weather resistance; superhydrophobic failure; failure mechanism超疏水表面在自清洁[1-6]、防腐[7-8]、防结冰[9-10]、防雾[11-12]、集水[13]、减阻[14]、液滴操控[15]及生物医学[16]等领域展现出广阔的应用前景，引起了研究人员的广泛关注。
Ultralow thermal conductivity of b -Cu 2Se by atomic fluidityand structure distortionHyoungchul Kim,a ,b ,⇑Sedat Ballikaya,c ,d Hang Chi,c Jae-Pyung Ahn,e Kiyong Ahn,b Ctirad Uher c andMassoud Kaviany aaDepartment of Mechanical Engineering,University of Michigan,Ann Arbor,MI 48109,USAbHigh-Temperature Energy Materials Research Center,Korea Institute of Science and Technology,Seoul 136-791,Republic of KoreacDepartment of Physics,University of Michigan,Ann Arbor,MI 48109,USA dDepartment of Physics,University of Istanbul,Vezneciler,Istanbul 34134,TurkeyeAdvanced Analysis Center,Korea Institute of Science and Technology,Seoul 136-791,Republic of KoreaReceived 5September 2014;revised 2December 2014;accepted 5December 2014Abstract—We demonstrate a prototype thermal evolution path for liquid thermal conductivity in solids.Thermal evolution of b -Cu 2Se shows large interstitial displacement of constituent atoms marked by glass-like transitions and an asymptotic liquid thermal ing ab initio molecular dynamics (AIMD),we identify these transitions,and confirm them with in situ transmission electron microscopy and electron energy loss spectros-copy.The thermal disorder of the Cu +ions forms homopolar Cu–Cu bonds under a rigid Se framework,and at yet higher temperatures the Se framework undergoes thermal distortion.The non-equilibrium AIMD prediction of lattice thermal conductivity shows significant suppression of the phonon transport,in agreement with experiments and molecular behavior.Ó2014Acta Materialia Inc.Published by Elsevier Ltd.All rights reserved.Keywords:Structure evolution;Thermal conductivity;Thermophysical property;Thermoelectrics;Phonon engineering1.IntroductionThermal structure evolution is ubiquitous in solids,including expansion,phase transformation and melting;as well as amorphous and crystalline morphologies,transitional superionics,macromolecules and quasi-crystals exist.These have attracted interest due to the possibility of being able to tune properties via structural anomalies.In thermoelectrics,structure evolution approaching liquid behavior has led to the “phonon-glass electron-crystal ”paradigm.Copper selenide,a chalcogenide,is a mixed conductor (electronic–ionic)that exhibits a phase transition [1–6].It undergoes a a –b transition (at T tr ;a Àb $410K)from the low-temperature monoclinic a -phase to the high-tempera-ture b -phase (face-centered cubic (fcc)structure,Fm 3m ;b -Cu 2Se)[1–6].Its critical electron and phonon scatterings during the second-order phase transition have beenexplored [7],however,its relevant thermoelectric (TE)properties suffer from structural instability and a narrow temperature domain.The b -phase also offers a high TE figure-of-merit over unity,due to abnormal Cu migration reducing lattice thermal conductivity (j L )[6,7].Fig.1a and b show the b -Cu 2Se structure consisting of eight cations (Cu +)and four anions (Se 2À)and several near-neighbor interstitial sites [3,5,8].Since the crystallinity of fcc b -Cu 2Se is maintained for T >T tr ;a Àb and the thermophysical properties are con-trolled by the Se framework,no significant property changes are expected.However,measurements show dra-matic changes,surprisingly without any distinct phase tran-sitions.The b -Cu 2Se liquid-like behavior includes reduced specific heat (c v )approaching the theoretical limit for liquid (i.e.c v ¼2k B per atom,where k B is the Boltzmann con-stant)and a colossal linear thermal expansion coefficient (a l )of 10.7Â10À5K À1above 800K [5].Another anomaly has been associated with the significant scattering of lattice phonons,i.e.glassy behavior close to the amorphous limit [5,6,9,10],as desired for “phonon-glass electron-crystals ”of superior TE materials [11].So far,clarifying the related physics and predicting the phonon transport have not been successful due to the theoretical and experimental (especially at high temperature)challenges./10.1016/j.actamat.2014.12.0081359-6462/Ó2014Acta Materialia Inc.Published by Elsevier Ltd.All rights reserved.⇑Correspondingauthor at:High-Temperature Energy MaterialsResearch Center,Korea Institute of Science and Technology,Seoul 136-791,Republic of Korea;e-mail:hyoungchul@kist.re.krAvailable online at ScienceDirectActa Materialia 86(2015)247–253/locate/actamatWe report the heterogeneous structure evolution of b-Cu2Se at elevated temperatures,which explains the anoma-lous thermophysical properties and significant suppression of phonon transport,close to a liquid-state behavior,with-out any phase ing the density-functional theory (DFT)including ab initio molecular dynamics(AIMD) simulations,we predict the temperature-dependent lattice dynamics and thermal conductivity changes due to the atomicfluidity and homopolar Cu–Cu bonds.We use in situ transmission electron microscopy(TEM)and elec-tron energy loss spectroscopy(EELS)techniques to test our theoretical predictions.2.Methodology2.1.Sample preparationCu2Se was synthesized via a solid-state reaction method by mixing Cu(99.9999%shot)and Se(99.999%powder),in their respective stoichiometric ratios,in a glovebox.The charge was then loaded into a carbon-coated quartz tube and sealed under a vacuum of10À3torr.The quartz tube was placed in a furnace and heated to1423K at a rate of 1K minÀ1with a subsequent soak for12h to ensure thor-ough mixing of the constituents.The temperature of the furnace was then decreased to1073K at a rate of 4K minÀ1and held for10days to complete the reaction. The furnace wasfinally turned offand allowed to cool to room temperature.The resulting ingot was ground into a fine powder in a glovebox and loaded into a cylindrical graphite die for spark plasma sintering under a dynamic vacuum.The powders werefirst cold pressed under a pres-sure of50MPa,and then the pressure was reduced to 10MPa.The samples were heated under this reduced load at a rate of50K minÀ1to afinal temperature of823K at which point the uniaxial pressure was increased to50MPa, with a subsequent isothermal/isobaric hold.The ingot was cut with a diamond-blade saw into disks for thermal diffu-sivity measurements and into rectangular bars for transport characterization.2.2.Property measurementsThe low-and high-temperature AC Hall effect measure-ments shown in Fig.2a were performed to determine the carrier type,and concentration of the charge carriers and their mobility.The Hall coefficient(R H)measurements were performed on a Hall bar with typical thickness of1mm in a Quantum Design Magnetic Property Measurement System (2–300K)using silver epoxy contacts and/or in a custom-ized air-bore Oxford Instruments superconducting magnet (300–850K)using pressure contacts.The signal was recorded using an AC resistance bridge(LR700,Linear Research),magneticfields ofÆ1T were applied in order to eliminate probe misalignment.All high-temperature transport properties in Fig.2b and c were observed under the dynamicflow of argon in the temperature range300–900K.The electrical conductivity(r e)was measured using a home-built apparatus with a standard four-probe poten-tiometric configuration,and the results are shown in Fig.2b.We have repeated the measurements for each sample at least twice and in each of these have repeated a heating–cooling cycle.After thefirst measurement,the sample was removed and all surfaces polished to make a new contact with new thermocouples.As shown in Fig.2b,the thermal hysteresis becomes negligible and there were no material losses.However,Brown et al.[12]report that under normal operating conditions material loss does occur(mostly due to vaporization of Se)and innovative methods should be developed to avoid this.The total ther-mal conductivity(j)was calculated from j¼D q c p,where D is the thermal diffusivity,q is the bulk density and c p is the specific heat capacity at constant pressure.Thermal dif-fusivity was measured using a Flashline5000laser-based apparatus(Anter Corporation).Specific heat was measured by a differential scanning calorimeter(DSC404Pegasus, Netzsch),and density at room temperature was obtained by Archimedes’method,indicating a value corresponding to98%of the theoretical density.Linear thermal expansion was measured using a horizontal pushrod dilatometer(DIL 402E,Netzsch).The sample was heated with5K minÀ1to 1143K under argon atmosphere.For analysis of structural and electronic properties,a transmission electron micro-scope(Tecnai F20,FEI)operating at200kV and equipped with a CCD imaging system(GIF866,Gatan)was used. Electron energy loss spectra were acquired with a disper-sion of0.05eV/channel,a2mm GIF entrance aperture and parallel illumination.2.3.First-principles calculationsWe investigated the high-temperature electronic struc-ture and the lattice dynamics of b-Cu2Se using the DFT method(including equilibrium and non-equilibrium AIMD)implemented in the Vienna Ab initio Simulation Package(VASP)[13].The PBE parameterization of the GGA for the exchange-correlated functional[14]and the PAW method for modeling core electrons(energy cutoff=355.2eV)[15,16]were used.High-temperature lattice dynamics were investigated by equilibrium AIMD.These simulations were performed on supercells consisting of 324atoms(3Â3Â3conventional cells).To consider the thermal expansion in the high-temperature simulations, we prepared supercells with the lattice parameter (a=5.864A˚at T=453K)[5,6]using our measured linear(a)(b)Fig.1.Crystal structure of b-Cu2Se.The b-Cu2Se crystal structure(blue and green are Cu and Se atoms).(b)Multiple atomic sites forCu+and Se2Àions,with designations.The Se atoms form a rigidframework(fcc site,4a)and Cu+ions readily drift among the differentsites(octahedral,4b;tetrahedral,8c;trigonal site,32f).(For interpre-tation of the references to color infigure legend,the reader isreferred to the web version of this article.)248H.Kim et al./Acta Materialia86(2015)247–253thermal expansion coefficient.The Brillouin zone is sampled at the gamma point.After constant-temperature simulations with the Nos e thermostat for1ps(0.5fs time step)reaching equilibrium,we collect atomic trajectories for20ps(1fs time step).To predict the lattice thermal conductivity of b-Cu2Se, we use non-equilibrium ab initio molecular dynamics (NEAIMD).The VASP code was modified to perform the NEAIMD-energy exchange[17,18]as reported in the literature[19,20].The non-equilibrium method is one of the computational techniques used to predict the lattice thermal conductivity.It is derived from Fourier’s law:the time rate of heat transfer is proportional to the negative gradient in the temperature and to the area[21,22].The lattice thermal conductivity using NEAIMD simulations is computed as the ratio of an applied heatflux to the resulting temperature gradient[19,20],i.e.j L¼À½QðtÞ=A ðd T=d zÞÀ1;ð1Þwhere the overbar indicates time average and Q(t)is the heatflow rate.The heatflux is imposed by dividing the simulation cell into sections of equal width,and exchanging kinetic energy between hot and cold sections.The temper-ature gradient along the z axis is computed from the mean temperature of adjacent sections.Simulations were performed on supercells consisting of192(4Â2Â2con-ventional cells),288(6Â2Â2)and384(8Â2Â2)atoms. Structure preparations were the same as the equilibrium AIMD simulations.We used the constant-temperature sim-ulations with the Nos e thermostat for1ps(0.5fs time step) and,after reaching equilibrium,non-equilibrium calcula-tions were performed for20ps(1fs time step).Because the exchange of kinetic energy results in non-Newtonian dynamics in the hot and cold sections,only the linear portion of the temperature gradient is considered in calculating the lattice thermal conductivity.3.Results and discussion3.1.Anomalies in thermophysical propertiesFrom the various measurements shown in Fig.2a–c,we observe an abrupt change in the transport properties resulting from the distinct structural transition between the a-and b-phase at T tr;aÀb$410K.For the T>T tr;aÀb regime,the b-Cu2Se has a stable thermophysical structure due to the rigid Se framework and the negligible anomalies in high-temperature transport properties.However,several anomalies in the thermophysical properties of the b-Cu2Se (without any phase transition)are observed from the3.Thermophysical properties of b-Cu2Se as a function temperature.(a)Measured lattice parameter and linear expansion coefficient.(b)Measured density and the predicted modulus.(c)Measured specific heat capacity at constant predicted specific heat capacity at constant volume and the parameter.The predicted glass transition temperatures,T g;also shown.Shaded areas show temperature regimes.of the measured b-Cu2Se properties respect to temperature.(a)Hall coefficient,(b)electrical conductivity(includingtotal thermal conductivity.The conductivity measurements where during heating and cooling of the measurements were repeated twice,asH.Kim et al./Acta Materialia86(2015)247–253249detailed experimental and computational analyses.Fig.3 presents the temperature dependence of the thermophysical properties of b-Cu2Se.The low-temperature linear thermal expansion coefficient is nearly constant and consistent with the reported value,2:3Â10À5KÀ1for T$773K[5],while the high-temperature results have two distinct peaks.Con-structing tangent lines to the lattice parameter(a)curve, there are two distinct temperatures,$800and$1000K, where the tangents intersect.Such transition temperatures in the thermal expansion are commonly observed in the glass transitions of amorphous materials.While the glass transition theory generally applies to amorphous phases, the glass transition temperature(T g)signifies the glass to rubber transition.Noting the b-Cu2Se crystallinity(fcc for T>T tr;aÀb)up to the melting point(T m$1380K),here we suggest its glass transition,i.e.thermal structure hetero-geneity,at two distinct temperatures.Based on the mea-sured thermal expansion coefficient and the specific heat capacity at constant pressure,we use a lattice parameterrelation(a¼a þa a l D T,where a is the reference lattice parameter of a sample at T ),the bulk density(q¼m=a3, where m is sample mass),specific heat capacity at constant volume[c v¼c pÀcð3a lÞT=q,where c is constant],bulk modulus½B¼qðc pÀc vÞ=ð3a lÞ2T ,and the Gru¨neisen parameterðc G¼3a l B=q c vÞ,as shown in Fig.3a–c.Above the Debye temperature($292K[10]),B is obtained from the thermodynamic properties and the quasi-harmonic model,and its product with the volumetric thermal expan-sion is constant[23].These thermophysical properties undergo noticeable changes at two distinct temperatures, without notable phase changes.Below T¼800K,denoted as the“first glass transition point”,B;c v and c G are nearly constant.For800<T<1000K,B and c v decrease,and the c G increases with temperature.Above T¼1000K,the “second transition point”,most properties reach a plateau. These three distinct regimes explain the evolution of struc-tural heterogeneity,and the associated significant phonon scatterings are discussed later.3.2.Heterogeneous structure evolutionTo elucidate this thermal structure evolution leading to heterogeneous glass transitions,we use lattice dynamics analyses based on the DFT to reveal the unique vibrational behavior of b-Cu2Se as a function of temperature(see calculation details in Section2).Fig.4shows the tempera-ture-dependent atomic trajectories of Cu and Se atoms in b-Cu2Se obtained from the equilibrium AIMD simulations. Unique lattice dynamics features reported in the literature [5–7,9,10]are successfully reproduced:the Cu atoms are highly disordered,while the Se atoms remain approxi-mately in their rigid framework.Based on these AIMD results,the temperature-dependent bond length(r)of vari-ous atoms in b-Cu2Se atoms are calculated(Fig.5a).For the Cu atoms,multiple interstitial sites are accessed,and their bond lengths are much reduced(at T=500K, r Cu–Cu is2.94and2.60A˚for ideal symmetry and distorted structure,respectively)and decreases with temperature(at T=1100K,r Cu–Cu is2.56A˚).This feature is also observed in another Cu-related bond length,r Cu–Se,while r Se–Se fol-lows the pure thermally expanded symmetrical structure. The atomic trajectories of Cu atoms become random and some migration among sites occurs,but for the Se atoms below900K this is very limited(because of the larger ion size and lack of interstitial sites).Suchfindings are consis-tent with the qualitative understanding of b-Cu2Se lattice dynamics.Also,we expect the formation of homopolar Cu–Cu bonds with high-temperature structural distortions because the predicted r Cu–Cu is surprisingly in agreement with that of Cu metal(2.56A˚at500K[24],dash-dotted line in Fig.5a).This supports the hypothesis we stated above,and will be verified with TEM-EELS studies in the following section.Quantitatively,we also analyze the structural tran-sitions of b-Cu2Se using the structure factor S(q),i.e.SðqÞ¼X N u:c:n¼1fnexpði qÁr nÞ;ð2Þwhere q is the wave vector,r is the position vector and f is the atomic scattering factor,calculated from equilibrium AIMD(Fig.5b).As expected,the S(q)of a perfect crystal is represented as a series of sharp peaks at the designated wave vectors,while the thermally disordered structure gives broad and shifted(to lower q)peaks.The peaks associated with Cu–Cu and Cu–Se bonds are observed at3A˚À1,while Se–Se correlations give$1.9A˚À1.The second sharp diffrac-tion peak(at3A˚À1)in S CuÀCuðqÞand S CuÀSeðqÞis reduced to unity close to its asymptotic limit.These changes suggest significant alteration of the spatial periodicity of Cu-related bonds in the b-Cu2Se structure.We suggest marking this Cu thermal disorder,as T g;I$800K was observed in the expansion measurement and becomes more pronounced for T>700K(Cu–Cu panel in Fig.5b).The Se–Se corre-lations among the nearest neighbors persist in S SeÀSe(q),and Se–Se framework persists up to1000K.Similarly,wefind a close relation between the Se atom thermal disorder and mark T g;II$1000K,by interpreting the peak transitions in Fig.5b(Se–Se panel).Since the logarithm of the intensity of the diffraction peaks is proportional to the Debye–Wal-ler factor and decreases linearly with temperature[25],we use the squared modulus of S(q)to analyze this transition [26]:j S0iðqÞj2¼1fiN2X Nn¼1ficosðqÁr nÞ"#2þX Nn¼1fisinðqÁr nÞ"#28<:9=;:ð3ÞThis structural amplitude is strictly unity only at0K and has a clear linearity in a single-phase system,while it is Fig.4.Atomistic trajectories of b-Cu2Se sampled with AIMD atomic trajectories for3ps,for four different temperatures:(a)500,(b)700, and(d)1100K.250H.Kim et al./Acta Materialia86(2015)247–253close to zero for the liquid phase and its linearity is deflected at the phase transition temperature.The thermalevolution of j S0i ðqÞj2confirms the heterogeneous phase tran-sition of b-Cu2Se(for Cu-and Se-related bonds,T g$800 and1000K),as observed in Fig.5c.3.3.Chemical and electronic state analysisThe presence of these thermal structural heterogeneities was verified by analysis of the chemical and electronic states by(i)in situ TEM with sample-temperature control and(ii)EELS.Before TEM-EELS characterization,the sample was a bulk slab,and its crystal orientation was iden-tified by electron backscatter diffraction(EBSD).The f111g cubic plane,a layered structure owing to alternating segregation of Cu and Se atoms,is most relevant for the TEM-EELS analyses of b-Cu2Se(see Fig.6a).Thefinal EBSD mapping results shown in Fig.6b provide the crystal orientations of all the grains.The circled blue region corre-sponds to the normal direction parallel to f111g with an error of12.5°.This region was milled with a focused ion beam to prepare samples for in situ TEM-EELS and Fig.6c shows the images at two elevated temperatures. Other than carbon deposited on the ultra-thin sample edges,no significant physical change is observed,which is required to reduce uncertainties.As we expected,Cu and its compounds show distinct EELS spectra[7,27,28].Our EELS characterizations detect changes in the L2and L3core-loss edges of b-Cu2Se,as shown in Fig.6d.(i)As temperature increases,the EELS spectra above980eV show diminishing decay and reach a plateau for T P773K.This abnormal decay can be associ-ated with the multiple scattering of the high-temperature structure,while single scattering occurs in the low-temper-ature structure.(ii)Broader L3and L2peaks(or larger shoulders)are observed,while the L3/L2magnitude ratio becomes smaller,with increase in temperature.(iii)The magnitude of the Cu0peak(i.e.the right edges of the L3 and L2shadow region)becomes stronger as temperature paring these thermal evolutions with theThermal evolution of b-Cu2Se structure.(Cu–Cu,Cu–Se and Se–Se)in disordered simulations.The vertical arrows demonstrate Temperature-dependent partial and total The temperature dependence of the amplitude approximation between two glass transition projected quantities are also shown.The trend for each.The shaded areas shows temperature(b)experimental results on b-Cu2Se.(a)The f111g cubic plane(parallel to the sample normal)of individual grains.The rollingimages taken at300and773K.(d)Temperature dependenceof each ionic state of Cu atom are shaded with dashedreader is referred to the web version of this article.)H.Kim et al./Acta Materialia86(2015)247–253251conventional EELS observations of different Cu ionic states [27,28],we draw close analogies between them.Thefine structure of the low-temperature structure and the Cu ions (e.g.Cu+or Cu2+)[7,27,28]result in relatively sharp edges for L3and L2,while the weak(or broad)L3and L2ioniza-tion edges appear in the high-temperature structure and Cu metal[27,28].The changes in Cu ionic states result in sim-ilar variations in the white-line ratio L3/L2.From these analogies we conclude that the ionic states of Cu in b-Cu2Se depend on temperature and that T$773K(similar to T g;I) is the threshold temperature for Cu disorder,i.e.the pres-ence of Cu0(Cu–Cu direct bond)for T>T g;I,and Cu+ for T<T g;I.ttice thermal conductivityThe homopolar Cu–Cu bonds,two structural transitions and the related anomalous thermophysical properties of b-Cu2Se result in significant suppression of lattice thermal conductivity j L,reaching the limit of liquid thermal con-ductivity(j liquid).As noted in Section2,we use NEAIMD simulations and verify this observed suppression of lattice thermal putational details are presented in ing three different simulation cell lengths l,we verify the expected size effect and extrapolate the lat-tice thermal conductivity for an infinite structure with the linear extrapolation of their reciprocal relation, 1=j L¼1=j L;1þc=l,where c is a constant[19].Fig.7a shows the l dependence of the calculated j L and their extrapolation to very large l,for several temperatures. Due to the extensive computation time for NEAIMD sim-ulations,we use limited cell volumes.The uncertainties associated with the linear extrapolation of the limited data sets are shown by error bars in Fig.7b.The variations of the predicted lattice thermal conduc-tivity,as a function of temperature,are shown in Fig.7b, and are in favorable agreement with experiments[5,6,9]. The predicted lattice thermal conductivity is constant for T<T g;I,and decreases for T>T g;I,while the bulk, homogeneous crystal is expected to be dominated by the interphonon scattering and follow the TÀ1dependence [32].The minimum(amorphous-solid)thermal conductivityj min based on the phonon mean-free-path(k p)of one-half of its phonon wavelength[29]is also shown in Fig.7b,with its decomposition.The NEAIMD j L for T<T g;I is in gen-eral agreement with the total j min,indicating significant Cu+disorder and homopolar Cu–Cu bond formation noted in the above.Despite the stable Se framework,the interstitial random distribution of Cu+ions and reduction in bond length,reaching the interatomic separation of metallic Cu,result in large suppression of phonon trans-port,approaching j min.Applying the Bridgman theory for j liquid based on liquid analogy to the lattice[30,31,21], another lower limit to phonon transport is j liquid.As shown in Fig.7b,the j liquid model predicts a value slightly lower than j min for T<T g;I and close to j min;L for T>T g;II.We consider this as an additional phonon suppression due to the atomicfluidity of b-Cu2Se beyond T g;I.This is consis-tent with the principle of sound waves influids having a longitudinal component only(not supporting shear stress). The NEAIMD results approach j min;L and j liquid and sup-ports thisfluidity limit.The suggested two glass transitions of b-Cu2Se,correlate well with the anomalous changes in lattice thermal conductivity toward these limits.4.ConclusionsIn summary,the temperature-dependent lattice dynam-ics and the measured thermal expansion of b-Cu2Se show large interstitial displacement of the Cu+ions and thepres-ence of two distinct glass transition temperatures associated with the two elements.These heterogeneous evolutions result in significant suppression of the phonon transport (i.e.temperature-independent lattice thermal conductivity approaching j liquid).These results provide new strategy and direction for achieving superior thermal insulators, thermal data storage devices and TE materials. AcknowledgementsThe work at the University of Michigan is supported as part of the Center for Solar and Thermal Energy Conversion at the Uni-versity of Michigan,an Energy Frontier Research Center funded Variation of the inverse of the predicted j L with respectvariations with respect to temperature,and comparisonliquid thermal conductivity[21,30,31],and the Slack measurements and predictions in Fig.3and the literaturek B u f/l2f,where u f is the speed of soundðB=qÞ252H.Kim et al./Acta Materialia86(2015)247–253by the US Department of Energy,Office of Science,Office of Basic Energy Sciences under Award Number DE-SC0000957.Work at the Korea Institute of Science and Technology(KIST)was sup-ported by the institutional research program of the KIST (2E24691).S.B.kindly acknowledgesfinancial support by Scien-tific Research Projects Coordination Unit of Istanbul University with project number of43041and43053and would also like to thank the Scientific and Technological Research Council of Tur-key(TUBITAK)forfinancial support.We are thankful to Dr. Dong-Ik Kim of KIST for the EBSD analysis.References[1]W.Borchert,Gitterumwandlungen im system Cu2-x Se,Z.Kristallogr.106(1945)5–24.[2]A.Stevels, F.Jellinek,Phase transformations in copperchalcogenides:I.The copper–selenium system,Recl.Trav.Chim.90(1971)273–283.[3]R.Heyding,R.Murry,The crystal structure of Cu1.8Se,Cu3Se2,a-and c-CuSe,CuSe2,and CuSe2II,Can.J.Chem.54 (1976)841–848.[4]S.Danilkin,M.Avdeev,T.Sakuma,R.Macquart,C.Ling,Neutron diffraction study of diffuse scattering in Cu2-d Se superionic compounds,pd.509(2011)5460–5465.[5]H.Liu,X.Shi,F.Xu,L.Zhang,W.Zhang,L.Chen,et al.,Copper ion liquid-like thermoelectrics,Nat.Mater.11(2012) 422–425.[6]B.Yu,W.Liu,S.Chen,H.Wang,H.Wang,G.Chen,et al.,Thermoelectric properties of copper selenide with ordered selenium layer and disordered copper layer,Nano Energy1 (2012)472–478.[7]H.Liu,X.Yuan,L.Ping,X.Shi,F.Xu,Y.He,et al.,Ultrahigh thermoelectric performance by electron and pho-non critical scattering in Cu2Se1-x I x,Adv.Mater.25(2013) 6607–6612.[8]M.Oliveria,R.McMullan, B.Wuensch,Single crystalneutron diffraction analysis of the cation distribution in the high-temperature phases a-Cu2-x S,a-Cu2-x Se,and a-Ag2Se, Solid State Ion.28–30(1988)1332–1337.[9]S.Ballikaya,H.Chi,J.Salvador,C.Uher,Thermoelectricproperties of Ag-doped Cu2Se and Cu2Te,J.Mater.Chem.A 1(2013)12478–12484.[10]Y.He,T.Day,T.Zhang,H.Liu,X.Shi,L.Chen,et al.,Highthermoelectric performance in non-toxic earth-abundant copper sulfide,Adv.Mater.26(2014)3974–3978.[11]G.Slack,New materials and performance limits for thermo-electric cooling,CRC Handbook of Thermoelectrics,CRC Press,Boca Raton,FL,1995.[12]D.Brown,T.Day,T.Caillat,G.Snyder,Chemical stabilityof(Ag,Cu)2Se:a historical overview,J.Electron.Mater.42 (2013)2014–2019.[13]G.Kresse,J.Furthmu¨ller,Efficient iterative schemes forab initio total-energy calculations using a plane-wave basis set, Phys.Rev.B54(1996)11169–11186.[14]J.Perdew,K.Burke,M.Ernzerhof,Generalized gradient approx-imation made simple,Phys.Rev.Lett.77(1996)3865–3868. [15]P.Blo¨chl,Projector augmented-wave method,Phys.Rev.B50(1994)17953–17979.[16]G.Kresse,D.Joubert,From ultrasoft pseudopotentials to theprojector augmented-wave method,Phys.Rev.B59(1999) 1758–1775.[17]F.Mu¨ller-Plathe,A simple nonequilibrium molecular dynam-ics method for calculating the thermal conductivity,J.Chem.Phys.106(1997)6082–6085.[18]P.Jund,R.Jullien,Molecular-dynamics calculation of thethermal conductivity of vitreous silica,Phys.Rev.B59(1999) 13707–13711.[19]S.Stackhouse,L.Stixrude,B.Karki,Thermal conductivity ofpericlase(MgO)fromfirst principles,Phys.Rev.Lett.104 (2010)208501–1–4.[20]D.Wang,L.Tang,M.Long,Z.Shuai,Anisotropic thermaltransport in organic molecular crystals from nonequilibrium molecular dynamics simulations,J.Phys.Chem.C115(2011) 5940–5946.[21]M.Kaviany,Heat Transfer Physics,Cambridge UniversityPress,New York,2008.[22]M.Kaviany,Essentials of Heat Transfer,Cambridge Uni-versity Press,New York,2011.[23]Anderson O,Isaak D.Mineral Physics and Crystallography:A Handbook of Physical Constants.Washington DC:AGU;1995.[24]M.Straumanis,L.Yu,Lattice parameters,densities,expan-sion coefficients,and perfection of structure of Cu and of Cu-In a phase,Acta Cryst.A25(1969)676–682.[25]C.Kittel,Introduction to Solid State Physics,Wiley,NewYork,NY,1971.[26]ers,S.Lemehov,M.Hou,M.Verwerft,Comparisonof interatomic potentials for UO2.Part II:Molecular dynamics simulations,J.Nucl.Mater.376(2008)66–77. [27]C.Ahn,O.Krivanek,EELS Atlas,Gatan Inc.,Warrendale,PA,1983.[28]F.Hofer,P.Golob,New examples for near-edgefinestructures in electron energy loss spectroscopy,Ultramicros-copy21(1987)379–384.[29]D.Cahill,S.Watson,R.Pohl,Lower limit to the thermalconductivity of disordered crystals,Phys.Rev.B46(1992) 6131–6140.[30]P.Bridgman,The Physics of High Pressures,Bell,London,1931.[31]J.Horrocks,E.McLaughlin,Thermal conductivity of simplemolecules in the condensed state,Trans.Faraday Soc.56 (1960)206–212.[32]G.Slack,The thermal conductivity of nonmetallic solids,Solid State Physics,Academic Press,New York,1979.H.Kim et al./Acta Materialia86(2015)247–253253。
