Everything, Nothing ...

Showing posts with label Research. Show all posts
Showing posts with label Research. Show all posts

Wednesday, July 09, 2008

MPICH常用网站

MPICH常用网站
2007-05-13 17:32
主要的MPI主页及MPI标准

MPIF主页:http://www.mpi-forum.org
MPI主页:http://www.mcs.anl.gov/mpi
netlib上的MPI站点:http://www.netlib.org/mpi/index.html


MPI的实现

ANL/MSU实现的MPICH主页:http://www-unix.mcs.anl.gov/mpi/mpich
MPICH实现:http://www.mcs.anl/mpi/mpich
MPICH实现:ftp://ftp.mcs.anl.gov/pub/mpi
其它的MPI实现:http://www.mpi.nd.edu/MPI
MPI实现列表:http://www.lsc.nd.edu/MPI2
MSU的MPI项目:http://www.erc.msstate.edu/mpi
LAM的MPI项目:http://www.mpi.nd.edu/lam

关于MPI的文档、讨论与例子

MPI相关材料:http://www-unix.mcs.anl.gov/mpi/tutorial
常见问题:http://www.erc.msstate.edu/mpi/mpi-faq.html
常见问题解答:http://www-unix.mcs.anl.gov/mpi/mpich/faq.html
勘误:http://www.mpi-forum.org/docs
关于MPI的新闻组:
MPI文档:ftp://ftp.mpi-forum.org/pub/docs/
MPI教程:http://www.mcs.anl.gov/mpi/usingmpi
MPI教程:http://www.mcs.anl.gov/mpi/usingmpi2
MPI例子程序:ftp://ftp.mcs.anl.gov/pub/mpi/using/examples
MPI例子程序:ftp://ftp.mcs.anl.gov/pub/mpi/using2/examples
MPI例子程序:http://www-unix.mcs.anl.gov/mpi/tutorial/mpiexmpl/contents.html

其它资源

MSTU Debian FTP:http://torvalds.cs.mtsu.edu/~zach/debian/current/
中科院数学所FTP:ftp://ftp.cc.ac.cn匿名登录
一个科大的FTP:ftp://imt.ustc.edu.cn匿名登录

MS4.0在安腾IA64位服务器上的安装总结

1.操作系统的选择

MS安装指南里 System requirements and recommendations部分说明IA64 Linux可用以下版本
Red Hat Enterprise Linux AS, ES, and WS - 3.0 (updates 4 and 5) and 4.0
SuSE Linux Enterprise Server 9 - SP2 (for standalone execution of CASTEP, Discover, and DMol3 only)
但实际测试发现Itanium 2 Linux中只有以下版本支持JRE1.4.2
Red Hat Enterprise Linux WS 3.0 Gnome2.2.2-metacity 2.4
Red Hat Enterprise Linux ES 3.0 Gnome2.2.2-metacity 2.4
Red Hat Enterprise Linux AS 3.0 Gnome2.2.2-metacity 2.4
Red Hat Enterprise Linux WS 2.1 Gnome-sawfish 1.0
Red Hat Enterprise Linux ES 2.1 Gnome-sawfish 1.0
Red Hat Enterprise Linux AS 2.1 Gnome-sawfish 1.0
SUSE Linux Enterprise Server (SLES) 8 Gnome2-metacity 2.

详见http://castep.bokee.com/viewdiary.180187010.html

如果在其他版本的操作系统上安装MS4.0会出现"Bundled JRE is not binary compatible with host OS/Arch or it is corrupt. Testing bundled JRE failed."的错误。这个问题困扰我好久,网上一直没找到解决方法,在反复多次实验后终于发现是OS版本的问题。

2.在所有节点上安装HPMPI,rpm -ivh /tmp/hpmpi-2.01.00-08.ia64.rpm

3.切换到MS的用户名,安装MS,注意要加cluster参数以用于并行 Install --type cluster

4.安装授权文件

5.配置各节点/etc下hosts入各节点IP和域名,在hosts.equiv文件中加入各节点域名,将hosts.equiv拷贝到用户主目录中并更名为.rhosts

6.修改/etc/exports文件,将MS安装目录共享出去

7.在其他节点上挂载MS安装目录mount -t nfs -o rw,sync,hard,intr ia02:/home/msi/ms /home/msi/ms

8.根据需要配置RSH或SSH

SSH参考帮助里的方法就可以了,在各节点上ssh-keygen -t rsa,然后三个回车,将各节点生成的id_rsa.pub文件合并成一个并改名为authorized_keys2放入各节点的用户主目录下.ssh文件夹中。

配置RSH时可以用rpm -qf `which rlogin` 查看一下目前使用的rlogin是从哪个套件产生的,如果是krb5-workstation就用以下命令rpm -e krb5-workstation 把他卸载 ,不然后出现Trying krb4 rlogin… Connection refused的错误。

9.修改参与计算CPU个数

vi share/data/machines.LINUX

venus:2 //节点名:CPU个数
earth:2
mars:2

vi Gateway/root_default/dsd/conf/gwparams.cfg //修改CPU总数

vi Gateway/root_default/dsd/conf/gw-info.sbd //修改CPU总数
10.配置网关自启动

cp msgateway_control_18888 /etc/rc.d/init.d

/sbin/chkconfig --add msgateway_control_18888

11.IA64上的大量floating-point assist fault 出错提示屏蔽

在用于安装MS的用户登录脚本login script (.profile, .bash_profile, .login, etc.) 里加入下面一行:

prctl --fpemu=silent
以上只是一些注意事项,其他细节参考MS的README_MS_Modeling.htm文件中关于安装卸载部分就可以搞定了,以上是在安腾服务器上安装的步骤,其他架构CPU机器上的安装要简单的多。

Materials-Studio 论坛问答全集

Materials-Studio 论坛问答全集(精选众多论坛讨论贴)
1、问:用MS构造晶体时要先确立空间群,可是那些空间群的代码是啥意思啊,看不懂,我想做的是聚乙烯醇的晶体,嘿嘿,也不知道去哪
可以查到它的空间群
答:A、要做晶体,首先要查询晶体数据,然后利用晶体数据再建立模型。晶体数据来源主要是文献,或者一些数据库,比如CCDC。
你都不知道这个晶体是怎么样的,怎么指定空间群呢?要反过来做事情哦:)
B、我不知道你指示的代码是数字代码还是字母代码,数字代码它对应了字母的代码,而字母的代码它含盖了一些群论的知识(晶系,对
称操作等),如果要具体了解你的物质或者材料属于那一个群,你可以查阅一下相关的手册,当然你要了解一些基本的群论知识.MS自带
了一些材料的晶体结构,你可以查询一下.

2、问:各位高手,我用ms中的castep进行运算。无论cpu是几个核心,它只有一个核心在工作。这个怎么解决呢?
答:请先确认以下几个问题:
1,在什么系统下装,是否装了并行版本。
2,计算时设置参数的地方是否选择了并行。
3,程序运算时,并不是时时刻刻都要用到多个CPU

3、问:我已经成功地安装了MS3.1的Linux版本,
串行的DMol3可以成功运行。
但是运行并行的时候出错。
机器是双Xeon5320(四核)服务器,rsh和rlogin均开启,RHEL4.6系统。
其中hosts.equiv的内容如下:
localhost
ibm-console
machines.LINUX的内容如下:
localhost:8
现在运行RunDMol3.sh时,脚本停在
$MS_INSTALL_ROOT/MPICH/bin/mpirun $nolocal -np $nproc $MS_INSTALL_ROOT/DMol3/bin/dmol3_mpi.exe $rootname
$DMOL3_DATA
这一处,没法执行这一命令
并行运算时,出现以下PIxxxx(x为数字)输出
ibm-console 0 /home/www/MSI/MS3.1/DMol3/bin/dmol3_mpi.exe
localhost 3 /home/www/MSI/MS3.1/DMol3/bin/dmol3_mpi.exe
请问这是什么原因?谢谢!
答:主要是rsh中到ibm-console的没有设置
把/etc/hosts改为
127.0.0.1 localhost.localdomain localhost ibm-console
在后面加个ibm-console
也希望对大家有帮助!

4、问:在最后结果的dos图中,会显示不同电子spd的贡献,我想问的是,
假设MS考虑的原子Mg的电子组态为2p6 3s2,那么最后的dos结果中的s,p是不是就是2p,跟3s的贡献.比如更高能量的3p是否可能出现在
dos中?
如果可能的话,在这种情况下,如何区分2p和3p的贡献,谢谢.
答:A、取决于你的餍势
势里面没有3p电子,DOS怎么会有呢?
自然,你的1p1s也不会出现在你的DOS中。
B、我觉得原子没有d轨道但是计算出来的DOS图上有d轨道的贡献,这可能是电子跃迁所引起的,虽然按照教科书上的d轨道上没有电
子,但是不排除有电子跃迁过去,但及时跃迁也应该只是很少的而且一般也都是这样。
C、我考虑的是dos并非指电子出现的几率,而是能级的密度.3p构成导带能级不可能吗?
没有d电子出现d的贡献可以理解是跃迁过去,那是不是也不否认3p的可能?
既然不可能有3p,怎么会有1s1p?毕竟一开始计算中已经做了近似?
D、第一性原理计算高于费米面的能带本来就不怎么可靠的。
现在,计算的带隙小于实验值,很大的原因就来源于价带计算不准。
p能量高于s不太好理解
这里有成键反键的因素
p的反键是有可能小于s的成键的
另外有时候原子电子组态没有d电子的时候计算结果也出现了d轨道
估计是做轨道投影的问题。vasp就有这个问题,但是可以消除这个错误的。

5、问:那位高手能够介绍一下MS中CASTEP的势能模型的适用范围?
比如说LDA的(PWC,VWN),GGA的(PW91,BP,PBE,BLYP,BOP,VWN-BP,PRBE,HCTH),能够简述一下其大致差别更好
答:LDA局域密度近似(LDA):局域密度近似(LDA)是第一阶梯。它仅仅采用空间点r处的电子密度n(r)来决定那点交换-相关能
密度的形式。交换-相关能密度由密度相同的均匀电子气完全确定。泛函的交换部分就准确的用均匀电子气的微分表达。各种不同的
局域密度近似(LDA)仅仅是相关部分表示方法不同,所有现代应用的局域密度泛函都基于Ceperly和Alder`s在80年代对均匀电子
气总能量的Monte Carlo模拟。
广义梯度近似(GGA):GGA是Jacob阶梯的第二个台阶,将电子密度的梯度也作为一个独立的变量(|∇n(r)|),在描述交换-
相关能方面,梯度引入了非定域性。GGA泛函包含了两个主要的方向:一个称为“无参数”,泛函中新的参数通过已知形式中参数
或在其它准确理论帮助下得到。另外一个就是经验方法,未知参数来自于对实验数据的拟和或通过对原子和分子性质准确的计算。
Perdew,Burke and Emzerhof(PBE)以及Perdew-Wang from 1991(PW91)是无参数的,在量子化学中广泛采用的GGA,比如
Becke,Lee,Parr and Yang(BLYP)是经验性。LYP校正采用了密度的二阶Laplace算符,因此严格上讲属于Jacob阶梯的第三阶,但通常
仍然归类为GGA.

6、[求助]Linux上MS的安装问题
我在RHES4.6上安装MS3.0
但是在castep和dmol下的RunDMol3.sh大小为零,没有东西,
同时在share下也没有bin目录,也找不到ms_setuo.sh脚本
请问是什么原因?
相应的出现这样的错误信息:
/home/sript/MSI/3.0/3rdParty/regxpcom: error while loading shared libraries: libcxa.so.3: cannot open shared object
file: No such file or directory
答:假如你有Materials Studio 3.1的ISO版本以及licene.dat。在linux机器上的用户名为firefox。下面是安装Materials
Studio3.1 linux版本的步骤
1、假定把Materials Studio3.1的iso文件mount到/home/firefox/ms目录,切换到root帐号下,然后用下面的命令如下:
mount -o loop /home/firefox/MS_MODELING.iso /home/firefox/ms
2、文件mount上去后,退出root帐号,回到firefox帐号。然后到/home/firefox/ms/UNIX目录,运行下面的:
./Install
3、开始按提示安装了
1)、指定你要Materials Studio安装到的目录,比如/home/firefox/MaterialsStudio
2)、下面就是指定你所需要安装的模块了,根据自己的需要选择,当然也可以全部选上
3)、然后指定你License_Pack所在的目录,我们假定也是放到/home/firefox/MaterialsStudio/License_Pack目录中,那么先建
立这样的目录:mkdir /home/firefox/MaterialsStudio/License_Pack
接着按提示Enter。开始安装Licens_Pack了。
4)、下面把你所得到的licene文件msilic.lic拷贝到/home/firefox/MaterialsStudio/License_Pack/licenses目录中,并改名为
msilicense.dat。
5)、把/home/firefox/MaterialsStudio/License_Pack/licenses/msilicense.dat完整的文件名按提示输入。这些完成后。就基本
安装好了。下面就是把gateway运行起来。
4、到/home/firefox/MaterialsStudio/Gateway目录,运行下面的命令:
./msgateway_control_18888 start
提示说Gateway start succeeded - running as process xxxx。后,就启动成功了。
5、在ie的地址栏输入,http://127.0.0.1:18888/ ; 则页面会显示该机器上已经安装好的计算模块
基本就安装成功了。(注意之处:不要用root安装它)。

7、问:我想对NaYF4的电子云分布进行模拟计算,请各位高手指点怎么计算??
如果有NaYF4的结构的话,请分享一下.
答:如果是晶体的话,可以找相关的文献看晶格常数和原子坐标,自己构建模型后计算
如果没有文献,可以上无机晶体结构数据库ICSD查询,如果查到,直接导出来
就可以用ms算了
电子云分布算一下电子密度就可以了
可以用dmol和castep
如果要某个轨道的电子云,dmol算homo-lumo上下若干个轨道
castep算能带
最后那visulizer看电子云密度图就ok了

8、问:各位高手请教一下用demolJ计算分子轨道怎么看HUMO,LUMO图,其中的黄色蓝色代表什么啊,谢谢了
答:dmol计算的时候选择properties里面的orbitals
把homo和lumo选上
然后计算,算完以后打开analysis对话框
选择orbitals,就看到你算的homo和lumo了,然后import就在模型中可以看见homo和lumo了
9、求助有MS并行计算经验的朋友解答一下
我虽然在LINUX CLUSTER并行成功了,但是并行的时候CPU计算利用率很不均衡也不稳定
有时候100%有时候0%
而且计算到一定程度的时候会变换运行程序在LINUX中的进程名
换了一次之后CPU使用就开始出现不均衡的情况,有一个CPU 90% ,别的都是2-5%的样子进程处与sleeping状态只有那一个CPU在RUN
另外说一下我每个节点的硬件配置相同,用的是CASTEP模块,K点数是CPU数的整数倍,但是算能带的时候不知道如何调节K点数目,
不均衡不知道是不是算能带的时候导致K点在每个CPU上分布不均造成的
还想问一下 我各个节点都是同一个交换机下的机器,NFS的效率如何评估
答:A、你应该看看内存的使用率,一般如果内存不足,会造成你说的现象
B、所有节点空闲内存都在50%以上啊,发现只有一个节点的CPU占用的厉害 别的不超过10%
C、你选用的几个cpu进行计算,如果是一个,那就是你发生的问题原因,如果不是一个,而是选择几个cpu进行计算,那么应该是节点之间
的通讯问题了,这需要进行机群的的调节了

10、问:CASTEP能算pdos的自旋情况吗
答:可以计算体系的自旋态密度
alpha和beta分别为自旋向上和自旋向下
total spin是总自旋
计算的时候必须把自旋极化-spin polarized 选上才能画自旋态密度

11、问:大家帮忙看看这是什么问题
我再用demol计算动力学的时候,不管怎么设置参数它老出现这个提示
Error: SCF iterations not converged in 50 iterations
Message: SCF not converging. Choose "Use Smearing" on DMol3 SCF panel
or set "Occupation Thermal" in the input file
You may also need to change spin or use symmetry
Resubmit DMol3
Message: DMol3 job failed
Error: DMol3 exiting
请教高手这是什么原因啊?

答:SCF not converging. Choose "Use Smearing" on DMol3 SCF panel
or set "Occupation Thermal" in the input file
无法收敛,加大计算步数。
你错误的原因是体系在50步内不能收敛,你可以增加收敛计算步数,比如增加到100
系统提示你改变"Use Smearing"和"Occupation Thermal" 的设置参数(这两个参数在设置对话框中),还提出建议让你改变自旋和对
称设置来试一试。
不收敛,解决方法一加大圈数,DMol3 calculations-》electronic标签下More选项-》SCF选项,中Max. SCF Cycles加大,最多
1000圈,如果1000圈收敛不了
就用方法二,加大smearing
在SCF标签下把orbital occypancy下面的勾打上,然后设置smearing
设置重小到大逐渐增大,smearing如果很大,虽然能收敛但是最后的构型有问题。
做MS计算,我的感觉是没有必要一步算到位,就是说你的计算精度(能量,力等等)可以一步一步来,或者是smearing慢慢变小,这
样容易收敛些。
12、新手请教几个不专业的问题
1.(1x1)CO中(1x1)是什么意思?
2.态密度、能态密度和电子态密度的区别?
3. K││-resolved是什么意思?
4.如何在半导体上建金属表面?
5.如何翻译?
supercell geometry
the Vienna ab-initio simulation package (VASP)
the generalized gradient approximation
the plane wave expansion
structural relaxations
a tight-binding linear muffintin orbital (TB-LMTO) method
the TB-LMTO method
the principal-layer Green’s function technique
the K││-resolved transmission
6.K││-resolved DOS at the Fermi energy这句话怎么理解啊?
7.用MS能不能计算一个体系的电导?如何去算?
答:1. (1×1)表示的是表面的构型,2×2是(1×1)的四倍,也就是x和y上分别用两倍单位平移定义表面的晶格常数a和b,c轴与表面
方向平行,c的大小取决于原子层和真空层厚度
2.应该是没有区别,可能是称呼不同而已。
3.不清楚什么意思
4. 用ms里面的layer builder构建界面,基底为半导体,半导体上长金属层是通过layer builder构建的
5. supercell geometry——超胞结构
the Vienna ab-initio simulation package (VASP) ——维也纳从头算模拟软件包
the generalized gradient approximation——广义梯度近似
the plane wave expansion——平面波展开
structural relaxations——结构弛豫
a tight-binding linear muffintin orbital (TB-LMTO) method——紧束缚线性丸盒轨道
the TB-LMTO method——紧束缚线性丸盒轨道方法
the principal-layer Green’s function technique——?
the K││-resolved transmission——?
6.不是很清楚,没有上下文
7 先计算出态密度。然后通过由玻尔兹曼方程推导出的电导率公式来算,见固体物理。提供两篇参考文献是做输运系数的计算的:
Georg K.H.Madsen, David J.Singh. Computer Physics Communications 175 (2006 67~71.
Philip B. Aleen ,Warren E.Pickett, Henry Karkauer. Physcical Review B 37 (1988) 2721~2726.
13、请问: 在用CASTEP计算材料空位缺陷时,如何才能实现只取一个空位呢?
答:降低对称性,再DELETE一个原子就可以了
在build下的symtrey有make P1
不是降低对称性,是构建超胞
make supper cell
然后就可以删去一个原子了
正常的话如果晶体是原胞,直接删除一个原子可能造成很大影响
可以建大一点的超晶胞(比如2×2×2超晶胞)然后删掉一个原子,这样比较合理,
毕竟缺陷浓度是很低的。
14、[求助] MS是否可以做原子数在10-20个左右的簇合物的结构模拟
我想得到一些原子数在10-20个左右的一元或二元原子簇合物的低能构型。
MS里的分子模拟功能是否可以模拟出一些能量局部最小化的构型?
如果不行的话,麻烦熟悉这方面的虫子给推荐个软件,最好是免费的。
我的要求很简单,就只要找到低能量构型即可。
手动直接构建实在是太麻烦了,而且说服力很低。
谢谢指教!
答:A、可以,dmol模块可以达到你的要求
B、感谢楼上各位。
不过我琢磨了半天,好像无论是Gaussian还是MS,都是要给定基本定型的构型,然后进行优化后获得更精炼一些的构型。
我想做这样一件事: 已知原子数为 10 的构型,第 11 个原子任意放置,然后程序自动变化第11个原子的位置,甚至变化 11 个原
子的位置 (既第11个原子对其他10原子的构型进行微扰)得到不同能量的局部的低能构型。
自己琢磨的,肯定很多不足,不是否有相关的参考资料。还请赐教,

15、ms discover停电后如何处理
ms discover算了一个星期,还有三天就算完了,结果停电全没了
重开机,启动ms,提示有任务非正常关掉,载入auto recover,job explorer里面还显示未算完的任务,server console打不开,无
法查看是否正在算
查任务管理器,没有找到discover.exe进程,只有msstudio的进程,而且cpu占用高
手动停止job,然后从discover的dynamics模块下选中use restart result,但是结果显示完全从头开始算
没钱买ups,很不走运,两个星期内,这种停电出现两次了
请教大家有什么好的方法,从断电时的数据开始接着算
谢谢
答:如果你在dynamic中trajectory的选项里save选择了full,那么在重启动时,在restart选项里,选中use restart date选项,就
可以了。
16、求助:黄色和蓝色的电子云代表什么问题
请教各位一个问题,在MS的Dmol3下计算能量,在分析里查看HOMO和LUMO轨道的电子云的计算结果时,有两种颜色,默认的是黄色和
蓝色。这种颜色的不同代表什么意思?
还有一般用HOMO和LUMO轨道的电子云说明什么问题啊?我看很多文献说什么(π-π)派-派跃迁就是3.1415926所代表的那个希腊字母
,这里不知道怎么输入。还有(σ-σ)西格玛-西格玛等等,学习化学的人应该能够了解,可以从这上面看出来吗?
答:A、其实两种颜色的电子云不是代表homo和lumo的,只是说这个地方波函数是正号还是负号。因为对于电子云来说,其实是波函
数的平方,这个地方为了区别波函数的正负号,所以引进了两种颜色的电子云
B、在分子进行反应时,最高占据轨道(HOMO)和最低空轨道(LUMO)其有特别重要的作用。
分子轨道对称性守恒原理认为:在反应过程中,当有反应物分子轨道转变成产物的分子轨道时,其对称性始终保持不变,即始终都属
于同一个不可约表示。
如果反应物基态的分子轨道变成的是产物的基态的分子轨道,那么反应就是热允许的。(意即温和加热就可进行)
如果变成的是激发态的分子轨道,就是热禁阻的(意即不可能用温和加热的方法促其进行)。
如果反应物受光激发后由电子占据的分子轨道(激发态轨道)转变成产物的基态分子轨道,这一过程就是光允许的(意即用适当的光
照就可使其进行)。
若转变成产物的更高激发态分子轨道,就是光禁阻的(意即无法用光照促其进行)。

17、请问MS计算能带的问题
由CASTEP或DMol计算得到的能带,要看能带的带系宽度,是怎样看?
是默认费米能级为0吗?那就是只看0附近的宽度吗?
答:A、0两边的宽度,最高占据和最低未占据之间的宽度就是带隙
B、我没有仔细的看过castep的能带,但是看过课题组的人做过报告。
我印象中,对于非金属,castep的费米能级是在价带顶,找出价带顶最大的那个能量和导带底对应的最小的能量之差,就是带隙吧。
至于有的说dos看带隙,这个不大可靠,要看你取了多大的展宽因子,展宽因子大,非金属也变成金属了,没带隙。
另外对于dmol来说,何来什么能带 全部分子能级,不过可能有gap,要从homo-lumo的差值去算gap
C、横轴是布里渊区的高对称点,与晶体的对称性有关
纵轴表示能带的能量。
带隙可以分成直接带隙和间接带隙
直接带隙就是0eV上下最高和最低两个能带在某个高对称点(如G,X、R点等)之间的能量差。
间接带隙就是0eV上能量最低的能带上的最低点的能量值减去0eV下能量最高的能带的最高点的能量值。

18、什么是Overlap populations?
在一篇文章上看见
?Table 2 Bond distances (A? ) and Overlap populations (/E/in brackets)

TiO2 TiO2/Sn6c TiO2/Sn5c TiO2/Sn6-in TiO2/Sn5-in
Ob–Me6c 1.831 (172) 1.963 (220) 1.83 (165) 1.831 (172) 1.832 (171)
Op–Me6c 2.066 (86) 2.107 (32) 1.989 (64) 2.070 (84) 2.069 (83)
Op–Me5c 1.938 (122) 1.912 (114) 1.995 (139) 1.988 (123) 1.933 (120)

O1–Me6c 2.065 (86) 2.165 (86) 2.097 (117) 2.072 (115) 2.086 (119)
O2–Me5c 1.853 (130) 1.854 (119) 2.006 (122) 1.865 (134) 1.890 (112)

O1–Me6-in 1.936 (160) 1.950 (142) 1.983 (163) 2.056 (177) 1.968 (161)

O2–Me5-in 1.968 (83) 1.961 (84) 1.950 (96) 1.976 (86) 2.047 (90)
请问Overlap populations 是什么?如何得到的?
答:A、Overlap populations
重叠布居
CASTEP计算性质时的population选项可以算
B、重叠布居
相当于键级的概念
一般来说,正值表示成共价键,正值越大,共价性越强
0表示纯离子键。
负值表示反键。
主要用于定性分析。不同的赝势、泛函计算的结果可能有差别
19、请教一下
这些数据在那?
您能把您算过的这部分给贴上来吗?
CASTEP计算性质时的population选项这个我已经算了
但是我找不到它在那
谢谢
答:CASTEP 中计算的时候选择properties里面的population analysis下面的
bond population 计算得到的结果除了有atomic population以外还有就是bond population 也可以叫overlap population
在.CST或者.CASTEP文件最后部分。
20、用MS需要什么样的配置?谢谢!
老师打算买一台新的电脑做计算,我不知道一般需要什么样的配置就可以了!希望大家多提供点信息!谢谢了!
答:A、双核 至少奔D 2.8GHz 1G内存
高分子结构 最好准备买服务器!!!!
B、楼上的,没那么夸张吧?
CPU好像用不太多啊,一般的30-60原子有个普通双核也够用了吧?我没感觉出来双核比P2.8快多少啊?
你非要算蛋白质的话,那服务器也肯定不够。。。
内存大些倒是真是,应该需要2G也不嫌多的!
C、我觉得看你做什么体系
你如果做简单的分子体系
用demol可以解决的
谱图电脑夜可以的
如果算晶体
那必须是搞配置的机子
D、关键是cpu,内存,和硬盘
ms计算的结果还是蛮大的,我的硬盘都满了
要速度的话cpu双核可能没有单核快,因为并行效率在windows下好像不是很高
内存大是必须的,2G基本上可以,如果4G更好,反正现在内存也不是太贵
尽量多加点内存把,否则你想算一个稍大的体系,往往发现内存不够,确实够郁闷的。^_^
21、问:模拟晶体表面吸附时,怎么样才能做到只允许垂直于表面方向弛豫,而平行表面的方向不允许呢?是用“Modify--
Constraints”实现的吗?如何实现?
答:就是用“Modify--Constraints”
步骤:第一:选择要加限制的原子
第二:打开constrain对话框
第三:设置固定XY方向不考虑弛豫,实际上就是在x和y方向的笛卡儿或者分数坐标前打勾即可
22、问:与计算速度有关的都有什么因素,我计算的时候,CPU利用率约为百分之五十,刚开始出了几个点,可是后来好几天也没出
一个点,而且job explorer 显示是running,不知道怎么回事?
答:与计算速度有关的因素:cpu速度,内存大小,缓存多少,硬盘和主板接口速度等。
不知道你是用winxp版本还是linux/unix版本
如果是winxp版本,可能cpu是双核/双线程的,你交作业的时候选择单cpu
没有选择用2个cpu,所以cpu利用率大概在50%
23、请问我在CASTEP中计算掺杂后的模型,能用Reflex模块中XRD在计算吗?我直接将CASTEP中的.xsd文件激活后使用了Reflex结果
感觉很不对。
向各位请教!谢谢 我就是建立了一个结构.xsd,然后选了那个模块计算,出来的结果感觉是非常不合理的。你能给我举个例子吗?
答:新建一个文件,导入数据库中的FeS2(pyrite)
在reflex里面选power diffraction
设置xrd谱图的范围,靶的种类等等
点caculate
就ok了
24、问:如何将MS的car文件装换为xyz文件, 坐标比较多。
答:利用MS的 Export 输出为mol 格式的文件
然后利用Hyperchem打开上面的输出文件
另存为xyz文件即可

25、Dmol计算的频率如何看明白啊?
偶用Dmol3算体系的能量,算了其频率,输出文件中一大堆数据看不懂都是什么啊?哪里有详细说明吗?help里好象没有啊。不知道
有这方面的文献吗(中文最好)?请高手们指教啊!
答:性质中频率是算声子的。
最后给出的振动模式 vibrational mode
输出文件主要看outmol文件,文件最后部分有振动频率的数据
如果不是周期体系,且假设体系有N个原子,则给出3N-6个振动频率
其中正值表示稳定振动,负值(虚频)表示不稳定振动,即振动将导致体系能量下降
如果是晶体则有3N个振动频率,其中有3个为刚性振动,频率接近0,表示所有原子沿着相同方向振动。除了振动频率的数值(振模频
率),还有振动的本征矢,每一个频率对应一组本征矢,本征矢表示原子的振动方向(x,y,Z)和幅度。
当发现晶体具有虚频的时候,那么原子按照本征矢方向振动,体系能量下降,会导致结构相变发生。
注意,dmol振动性质的计算并没有考虑电场和原子位移的耦合作用,没有考虑LO和TO分裂,而且对于晶体而言只能算布里渊区G点的
振动频率,而不能计算其他点的振动频率。
完毕
26、问:在计算的过程中CPU利用率变为0,重启以后任务还显示再计算,单CPU 利用就是0,再开始新的任务利用率就不是0 了,这
是为什么?怎么解决阿?谢谢了!
答:虽然软件界面显示在run,但此乃假象也!!cpu为0肯定是没算了。之所以还显示在算,主要原因是你的内存可能不够用了,电
脑不能正确反应实际情况而已。
正常现象,如果交个作业,然后把进程杀了,还显示正常运行,但是运行无法结束
也没有结果。所以看作业是否运行,不单单得看网关的运行状态,还得看cpu的利用率
还有两种可能
第一:就是算完了,网关的内容却没有下载到本地目录。
比如,交一个作业到服务器,算完以后,服务器的cpu利用率变为0,然后
由于网络原因,服务器数据往回传输的时候突然断网,这样可能数据就无法传回来了
由于传送指令发过一次以后,服务器相关的数据目录就发生变化,因此,在网络好的时候选择achieve,可能数据仍是不完整的。
第二:交完作业以后,网络就断了,在这个过程中作业因很多原因死了,cpu利用率为0,此时如果网络好了,在连接服务器,状态不
是完成状态,所以可能一直是runing状态,因此不会有结果。
如果服务器和客户端是一个机器,也可以发生第二种情况。
27、请问:
在CASTEP计算中,k point设置不同,对计算结果有什么影响?
k point设置表示什么?
如设置为4乘4乘4,和设置成9乘9乘6有什么不同,区别在哪?
答:很难一般地回答,只能给出一般建议。注意:一定要检查k网格,首先用较粗糙的网格计算,接下来用精细的网格计算。通过比
较两次的结果,决定选用较粗糙的网格,或是继续进行更精细网格的计算,直到达到收敛。金属体系需要精细的网格,绝缘体使用很
少的k点通常就可以。小单胞需要精细格点,大单胞很可能不需要。因此:单位晶胞内原子数很多(比如40-60个)的绝缘体,可能仅
需要一个(移动后的)k点。另一方面,面心立方的铝可能需要上万个k点以获得好的DOS。对于孤立原子或分子的超晶胞,仅需要在
Gamma点计算。对于表面(层面)的超晶胞计算,仅需要(垂直于表面)z方向上有1个k点。甚至可以增加晶格参数c,这样即使对精
细格点,沿z方向上也只产生一个k点(产生k点后,不要忘记再把c改回)。
k点主要是用来积分计算用的
例如对于连续的波函数,实际的计算是无法采用基于函数的表达式进行计算,而是基于离散数据进行计算,所以需要将空间进行划分
,如果空间中的离散点划分得越细,那么最后通过数值积分得到的结果就越接近体系的真实情况,如果空间的划分越粗糙,计算的结
果越不准确。k点面实际上就是在倒易空间按照a,b,c三个方向进行划分
因此用三个数值表示k点面,第一个数值越大,说明沿着a方向的数值积分点数越多,计算精度越高。在实际计算中,需要进行测试,
一般是逐渐增加k点数,计算能量收敛曲线。
对于表面相关计算,k点面中第三各数值往往取较小,一般为1或2,就是不考虑z轴方向上的周期性。
28、请教:
运行DPD模块,如果运行了1000步,结束后,想继续运行下1000步,怎么办?
是要重新setup中,设定2000嘛?有办法让他在1000步基础上继续运算嘛?
答:Restarting a DPD run
At the end of a DPD run and at specified intervals during the run, DPD saves a restart data file, .Dpd_rst. This
file is downloaded at the end of the run and contains the positions and velocities of all the beads. It enables you
to restart the simulation from part way through or at the end of the previous DPD run.
To restart a simulation
1. Select the mesoscale trajectory document from the run you wish to restart.
2. Choose Modules | DPD | Calculation from the menu bar.
3. Load the parameters from the run you wish to restart, by double-clicking on the associated Settings file
in the Project Explorer.
4. Check Restart.
5. On the Setup tab, choose whether to Reinitialize averages.
6. Change the length of the run.
Note. The time step counter is not reset between simulations. Therefore, the restarted DPD simulation will run for
the Number of steps specified on the Setup tab less the number of steps already completed.
7. If desired, specify new output options, simulation conditions and interaction parameters.
8. Press Run.
Note. The Restart checkbox is enabled only if a restart file is present. However, this file cannot be seen in the
Project Explorer window. You can use the Windows Explorer to verify that the file exists.
Note. When you restart a simulation you can modify some parameters such as the length of the run, output period
options, simulation conditions, interaction parameters, etc. However, others, such as simulation cell size, grid
spacing, cell contents, etc., must remain the same as they were before the restart. These options are disabled when
Restart is checked.
勾选 restart
在点中最后那个.xtd图形文件的条件下,run
你举的例子应该把步数改为 2000
29、求教CASTEP中几何优化问题
1. 几何优化可以在多大范围内改变原子的位置啊?感觉每次优化后的位置调整不是很多啊?
2. 大家在算能带和态密度的时候的时候是用单点能计算给出能带还是几何优化给出能带啊?
另请问WINDOWS版的KEY和LINUX版的KEY能不能混用 求一个MATERIAL STUDIO 4.0的LINUX版本的KEY一个 QQ29988789
十分感谢
答:问题1
几何优化的目的是寻找压力最小的几何结构,原子位置改变不多是因为你建的构形比较合理,比如直接从软件数据库中导入的结构在
0压力下改变很小,如果加个压力就会变化大一些。
问题2
一般都是优化的时候算,如果你想算特定构形的能带或其它性质,就可以用单点能;
30、请问CASTEP计算氧化物磁性,例如CuO的,Modify--------Spin里怎么设置?
Dmol 计算自旋设置时,Multiplicity设置为Auto是系统自动寻找自旋多重度
吗?Multiplicity是设置Auto好,还是设置具体的值好?
答:Multiplicity设置为Auto是系统自动寻找能量最低、最稳定的自旋多重度
31、请教关于MS castep的几个问题
希望各位大侠帮忙,
1 计算表面过程中,计算成功,但是castep结果中提示
------------------------------------------------------------------------ <-- SCF
SCF loop Energy Fermi Energy gain Timer <-- SCF
energy per atom (sec) <-- SCF
------------------------------------------------------------------------ <-- SCF
Initial -1.24983546E 004 5.76278315E 001 55.44 <-- SCF
1 -1.59703259E 004 -1.69750335E 000 2.89330938E 002 197.97 <-- SCF
Warning: There are no empty bands for at least one kpoint and spin; this may slow the convergence and/or lead to an
inaccurate groundstate. If this warning persists, you should consider increasing nextra_bands and/or reducing
smearing_width in the param file. Recommend using nextra_bands of 11 to 24.

2 -1.62762609E 004 -6.23532889E 000 2.54945870E 001 356.62 <-- SCF
Warning: There are no empty bands for at least one kpoint and spin; this may slow the convergence and/or lead to an
inaccurate groundstate. If this warning persists, you should consider increasing nextra_bands and/or reducing
smearing_width in the param file.Recommend using nextra_bands of 11 to 24.

2 要在表面上吸附小分子,如H2 ,或者CO,可以选择不同的位置,比如top, bridge, 等,但是如何确定具体每个原子的position?
直接根据原子半径按照几何知识来计算吗?另外在计算表面吸附total energy的时候,是否选择上"optimized cell" ?
如果不选择,就应该是每个原子的位置保持固定,那么如何知道所确定的吸附原子的位置是对的呢?
3 进行计算时选择上"spin polarized", 对于不同元素,其参数如何设置?"initial pin"和"charge"。

答:A、你把empty band里的空带数目增加试试或者减少smearing 的数值,这两个办法都可以增加精度。
你先把晶胞优化完成后再算吸附,只是注意两次计算所用方法要一致。确定原子的position是要根据能量高低来确定。
B、第一个文标题: 从他提示的关键词来看
NEXTRA_BANDS
This keywords controls the number of extra bands in addition to the number of occupied bands. These extra bands are
necessary for metals or finite temperature insulators. The default value for this parameter is 0.
SMEARING_WIDTH
This keyword determines the width of the Fermi-surface smearing if the system is being treated as a metal.
因为不知道你的具体物质,所以我猜测有两种情况,一是你的结构有问题。二是因为你的结构的特殊性,所以你默认的设置计算是不
可信的,就如提示而言。如果你的结构没问题,你可以按照提示调整一下参数算一算。
.param为后缀名的文件里,有以上两个参数,你打开调整一下,然后计算你调整后的文件,就是点Files,然后Run Files。如果你调
整后没有提示,就说明你的结构需要调整参数。
第二个问题
原子在表面吸附就那么几种位置,你都要计算试一试,找到能量最小的吸附位置。至于吸附原子放在什么位置,你可以通过原子半径
计算,这样计算能加快你的收敛速度。如果不是计算,而是大致放个位置,收敛就和你放的位置有关,如果放的位置很接近收敛位置
,就会很快收敛。因为这这两种方法最终的计算收敛后,他们的位置应该是一样的(误差范围之内)。至于计算表面吸附total
energy的时候,是选择"optimized cell"还是固定表面,我看文献上这两种都有,因为这要和你计算的体系有关,如果体系不是很大
,应该选择optimized cell,这样更近现实验情况。
第三个问题,initial Spin设置,MS提供两种方法,你可以参考。
There are two ways of defining the initial magnetic configuration: either specify the total magnetic moment per unit
cell, which gets uniformly distributed over the space, or provide detailed information on the absolute values and
direction (up or down) of the spins for each atom in the unit cell. The former method can be used for relatively
simple systems where only two solutions are expected (magnetic and non-magnetic). The latter method, which specifies
the spin state of the atoms in the system, is more general and gives much more flexibility. It is possible to set up
ferromagnetic, ferrimagnetic, or antiferromagnetic calculations to get different starting spin arrangements.
不过我记得好像他们公司说过,一般formal spin都可以,除非是计算一些磁性物质,比如Fe、Co这种。目前我都是默认。至于
charge就是系统的电荷,如果不带电就是0,实际是多少就是多少。
自己的一点拙见,仅供参考
C、问题3
MS中的帮助文件中的内容是:
The initial value of the spin moment is quite important for spin-polarized calculations. It should be as close as
possible to the expected value. You can evaluate the spin moment using the high-spin approximation combined with
formal charges. For example, if there are ten Fe3 ions in the system, you assume that each of them has a valence
electronic configuration of d5 (the neutral configuration is s2d6) which corresponds to five unpaired electrons per
Fe ion. The initial spin value for the CASTEP calculation is therefore fifty.
我如果计算Ni, s2d8 应该是2个 unpaired electrons, 所以,设置2*supercell中的Ni个数, 对吗?
D、通常磁性物质需要考虑spin是因为体系中有unpaired electrons。所以spin=unpaired electrons 1,是根据有无未配对电子设置
。你可以几种都试试,以能量低的为准。
E、从帮助的说明来看,应该是设置2*supercell中的Ni个数
另外就是注意你是周期性系统,边上、角上的原子不能算做一个原子,可能有的只是半个会四分之一个。
F、1 这样的提示我在计算掺杂Eu的体系中也见到过,当时也增加了empty bands,另外smear值有人说调大有人说调小,反正我都试了试
,效果都不是很好.我想可能是结构的原因,结构如果不合理的话其他调整不是那么明显.
另外也想补充个问题:我发现跟我没有调整这两个参数的计算结构比较,无论是band structure还是dos结果都有变化,不知道那样才是
合理?
2 我看到一些资料表面吸附时,通常的做法是把结构先optimized cell,然后用优化后的结构参数吸附分子模拟.位置先估计一下,然
后限制下层原子位置,表面的可以弛豫.
3 "spin polarized"的设置我就没怎么接触过了,看lz提供的ms帮助,受教了,谢谢
G、第二个问题:
吸附时每个吸附位都要考虑,而初始位置到底是怎么样的,一方面要看文献有没有类似的计算,另一方面就是经验的问题了。
在晶胞优化的时候才选上"optimized cell",而在计算表面吸附的时候是不能选的。至于固定几层原子,这和你的体系大小有
关,一般来说体系越大就可以多固定几层,体系小就少固定几层,固定原子越多,计算效率就高,花费时间相对就较短,这与你的初
始构型也是相关的,构型好计算就快。如果是在不知道固定几层,就参考文献值,总不会错的,或者固定一半(小体系)或三分之二
(大体系)。
F、总结一下:
问题1
基本上是找到了问题,而且必须解决,通过软件的提示和大家的讨论,可以增加bands或者减小smearing,我会先试试减小smearing
,看看结果如何
问题 2
看来大家大多数认为在吸附的relax过程中不选择optimized cell,原子的位置靠几何计算得到。我先这样试试,但是我还有一点疑
虑,就是如果不选择optimized cell,吸附分子或者原子的位置是不是就保持固定了呢?csfn认为位置也不是保持固定,大家有经验
的可以谈谈。我算完之后其实测量一下就知道到底固定不固定了:)
问题3
争议还比较大,
1 不设置,保持默认,但是默认的initial spin是0,这样不知道对不对?
2 设置spin=unpair electron 1,这个不知道是什么意思?有没有具体文献。
3 按照帮助文件设置,但是要考虑计算中超胞中很多原子是共用。
I、问题2
“但是我还有一点疑虑,就是如果不选择optimized cell,吸附分子或者原子的位置是不是就保持固定了呢?”
这个我应该说可以肯定的,吸附分子或者原子的位置是没有保持固定的,但是你的的单包的体积和形状是改变的,假设你是a,b,c
为10,10,10的边长,优化过后,还是10,10,10。但是你选择“optimized cell”,你会发现变化,可能是9,9,9或者是11, 11
, 11,就是说这个盒子的边长,体积等都在变化了。那么不管你选择还是不选择这个选项,你的原子都是没有固定的。固定的只有
两个情况可以固定,一个是对每个原子采用了固定的设置,这个是一个小技巧,你不去操作是不会自动加上的。第二个是你选择任务
的时候,没有选择结构优化的任务,比如算能量,那么这个时候结构不优化当然原子坐标不变化。应该是比较清楚了吧 这些我是可
以肯定的,不大信的话,你可以做一个小体系测试嘛。
问题1,和问题2,你自己判别了,好像不是什么问题了。总之这个软件是比较智能化了,只要你的结构设置合理,吸附的位置放置合
理,不太离谱,一般来说计算采用它的默认设置都是比较好的。
J、spin=unpairi electron 1,是来自量子化学或者第一性原理计算本身,说明书中不会有介绍。可以参看原子物理学关于角动量
耦合的章节或者读读潘道凯的物质结构。实际上spin=从1到unpairi electron 1的所有可能值,这是因为实际上我们事先不能肯定
它们已经配对。到底最终选择哪个数值需要看能量确定。在CASTEP中考虑spin主要是因为需要处理掺杂、磁性、以及吸附离子等电子
没有完全配对的情况。
K、我讲讲我做化学表面反应计算的步骤吧:
1 晶胞优化;就是搭建你要的基底晶胞并进行优化,这个时候选中optimized cell,这时会给你优化的晶胞参数,以后就不要在去管
了;
2 切面:一般是切低指数表面,优化表面后计算表面能.表面切几层,这要根据实际情况分析,我上面说过,层数越多计算量越大,
同时还要看你固定几层,这个楼上的说得比较详细,只是我看了下文献,好像两个表面的比较少。一般表面能越低,表面越稳定。
3 吸附:把吸附原子或分子放到目标基底上,这个初始结构很重要,给得好则很快收敛,否则收敛很慢甚至不收。初始构型也要多方
面考虑,各个吸附位各种吸附构型还有不同基底等等。
一般最好是先算构型优化,再算能量及DOS等其它性质,计算速度要快一点吧。至于其它参数例如自旋(看你的体系有没有单
电子)等的设置,一般参考文献就好。
L、这些问题已经基本解决,谢谢大家关注!
问题1, 改变empty band的值,按他的要求设置到11-24,就可以解决,但是单纯减小smearing的值似乎解决不了...
问题2 uncheck 'optimized cell"
问题3 好像这个参数并不是影响很大,因为这只是个初始值, 欢迎大家继续套路讨论这个问题,有什么经验的说说吧
M、问题1
我也碰到过这样的问题,一般只要按提示要求改一下空带个数就可以了
问题2
一般是先优化胞,再切表面,加上吸附原子,再优化,第一次优化要选optimize cell,第二次不选,一般要固定最底层的几层原子;
加原子时,如在bridge时,与表面平行的方向上的坐标为表面上两最近邻原子的中点,垂直方向上自己根据实验值或其它的估计,估
计不好就多算一会,top位置就在表面原子的正上方;
问题3
一般算的胞的所有原子的电子个数和是奇数是选spin polarized,偶数时不选;算磁性的时候就看情况了,那两格参数一般不用设置
,如果不是带电体的话

32、怎么可以计算比较准确的扩散激活能?用TS RESEARCH 算的时候所有的基体原子都不能移动,与实际过程不同,得到的激活能准
确嘛?有什么方法可以解决这个问题嘛?
谢谢赐教!
答:A、做表面相关的计算不能优化cell把,如果优化cell的话基本上是算不了了
表面+真空层,晶胞体积太大,算不动的。
一般是先优化晶胞的平衡构型,然后在切表面,
在进行相应计算
B、谢谢各位的建议!可是我想算体扩散,比如在H在Ti中的扩散,先优化含氢的一个构形,再优化另一个构形,可是它要求所有基体
原子不动,只有扩散的原子运动,所以优化的时候就得让基体原子都不动,我不知道这样算的结果是不是合理
C、体扩散应该难度不大,个人认为远离H的原子可以固定,在H附近的原子不能固定,要不然结果可能不具有可信度。如果不好确定H
附近的原子,则建议全部不固定。
D、楼主说的有点像渗碳 渗氮。这是一个扩散穿质的过程,解一个二阶偏微分方程,解的准确性关键在一些系数的选取,像扩散系数
D。不知道第一性原理能不能算这种很具体的系数。
33、请教DMol3 Calculation 出现的问题
我是新手,DMol3 Calculation 点Run后出现一个对话框
Failed to create job for the server: DMol3.
No default location defined for server DMol3.
Use Server Console to specify a gateway.

答:A、网关没有启动,或者被防火墙拦截了。
如果是winxp系统,把系统防火墙关闭,如果还有杀毒软件的防火墙,
第一次运行,要允许Apache.exe等访问网络
还有一个可能就是你安装的只是客户端,而没有安装服务器段
如果全部都安装的话,第一次运行你可能没有让网关启动。
B、根据我的经验如果在安装的时候没出什么错误的话是不用重装的,这可能是serve console里面出错了,你试试把以前的sever删
掉,再新建一个,选服务器(如果是自己的机子就输入你的计算机名),在这之前确保网络是通的,防火墙关掉(WINDOWS自带的也
关掉),装完之后test一下。如果这样不行的话,你可以在cmd命令行中输入:ping IP地址,看一下网络是否是通的;ping 服务器
名字,看一下名称解析是否有问题。出现这种问题往往是由于网络的问题,比如在运行任务的时候,修复一下IP地址,马上会出现问
题。奇怪的是在服务器是自己的电脑的情况下,在新建SERVE 的时候也必须关掉防火墙。
C、可如果有防火墙的话,到防火墙里面开启相关项如Apache.exe,alg,lmgrd,perl,msi等,看看行不
34、请问能在castep中得到内聚能吗?怎样得到,请高手指教,谢谢
答:1。如果是结合能,castep和dmol都可以算
dmol能直接给出结果,在outmol文件中。
castep得自己算。
算法:晶体的总能量(final energy)-(N1×晶胞中原子1个数原子+N2×晶胞中原子2个数原子+。。。。),就是原子成键后导致
体系的能量下降。
原子的能量在.castep文件开始算赝势的地方找。

35、请教MS中切面几何优化和原子之间的配位
1.我用MS-中DMol3切面做几何优化,假如主族氧化物A2O3,切三层,五层,七层,分子式分别为A8O16,A16O28,A24O40,五层优化始终
不收敛,设置自旋激化,和无自旋激化都试了就是不收敛,其他都是默认值。请问应该怎样设置?为什么?
2.体相O是四配位的,A是六配位的,切面时第一层是O第二层是A出现两种情况,一种O是两配位,A是六配位,另一种O是两配位,A有
四配位和五配位,做吸附气体小分子哪种切面好,为什么?
谢谢指教

答:A、你看一下你的输出结果,看一看他的收敛程度
如果没有收敛,那就改变一下结构试一试
如果一直是往收敛的方向走,那就增大一下SCI cycles 试一试
也许在多进行几步SCI 就可以收敛
第二问题,这两种情况你都要计算
进行比较,找出最有利的吸附构型,具体那个面容易吸附
积要看你的吸附体,也要看你的吸附质
不同的物质不一样
一般情况下,容易发生在晶面间距较大的面
B、SCI设置为1000轮了,就是每次几何优化第九轮时自洽迭代不收敛,一会儿变小,一会儿变大,不收敛.Smearing为0.005或0.0
1都试了不收敛.还可以调什么参数?就是想计算表面能看选五层还是七层就可以,选层方面有其他方法确定吗?
选100面晶面间距是较大,吸附小分子CO,NO,NO2,N2O,H2
CH4等应该选哪种面?
C、如果这样,五层每次都是第九步不收敛,我觉得可能有一下几方面原因,(自己的拙见,仅供参考):
一、再检查一下你的结构是不是有问题,七层都可以收敛,说明你的计算机是满足要求的。不会能因为内存不足等原因。建议你先用
半经验优化一下,然后再用Dmol优化,这样l优化后的结构Dmol再优化会收敛快一些,也比较容易收敛。
二、循环次数你已经增大了,如果是你说的一会变大,一会变小,那就是结构在震荡,也有可能说明这个表面结构就是不稳定的。
如果不行,只能调节收敛精度了。
至于选层,没有什么具体要求,一般都是3-6层,这样要你自己计算,一般都是把基层都优化以下,看看表面结构影响到几层。比如
,你有优化的3层,4层结构,发现第三层、第四层对表面影响很小,计算到第三层就可以了。
至于吸附,选什么面。一般情况下首先晶面间距较大的面。但也要具体问题具体分析。比如对CO,一方面看表面是C吸附,还是O吸附
,还要看是金属吸附还是O吸附,
他们会选择最有利于吸附的晶面。比如是金属吸附C,那就会选择含金属原子比较多的面。
D、跟smearing 设置有关,你可以试一下
smearing从小逐渐增大,就收敛了。不过smearing太大,最后优化的构型可能会有问题。DMOL优化不收敛,一般得设置smearing,你
可以看outmole文件,里面如果不收敛的话会有体系你修改什么参数的。
好运

36、请教用MS计算半导体氧化物的能带时怎样设置路径?
用MS计算半导体氧化物的能带时,More按扭里有K点Path给出有G(0 0 0),F(0 0.5 0),Q(0 0.5 0.5),Z(0 0 0.5),B(0.5 0 0).其自动
的路径是G-F-Q-Z-G,但是得到的能带很乱,请高手指点怎样设置路径能够好些?
答:A、实际能带结构应该是4维图像,而从对称性来考虑,我们关心的仅仅是一些高对称点之间的能量分布,所以一般的结构都是给
出的2维图像。再者,通常意义上能带结构中k-point的数值只有相对的意义。
能带关系就是E_k关系,所以如果在CASTEP中把能带图export成数据格式的话,看到的两列数据分别是k和能量。k对应的数值
如何理解,这个好像不同软件包做法并不一样。CASTEP中将所选的所有bandline也就是kpoint-path的长度定义为一,不同区域的宽
度(kpoint-path)对应两个高对称点的相对距离。
如果你使用的是Castep自己默认的,那么他确认的都是高对称k-point,得出的图像E和K-point应该是对应的。你可以自己设
置,但是这样计算,你的出的DOS是不可用的。
你检查一下你的模型是不是有问题
最好能够把你的图传上,这样好分析原因
B、最后的方法就是所有的点都两两相邻,
比如系统有G,X,M,R四个点
路径如下G-X-M-G-R-X-M-R所有的点都两两相邻
这样反应能带的信息更全一些。

37、请教MS里面的图怎样复制到WORD里面来?
我在MS计算出的图中点复制,然后在WORD 里面粘贴出来的 是一些数据,请问改如何操作呢?谢谢
答:1、你可以利用这些数据在Origin里作图,这种方法你可以调解坐标轴,坐标。然后把Origin里的图拷贝到Word里。
2、利用MS的File菜单下的Export,将你要的图形输出为图形,好像只能输出为bmp格式,然后插入到你的Word里
3、就是利用一些抓图软件,我常用的是UltraSnap,通过抓图软件将你要的图形粘贴到Word里

38、请教
主族金属氧化物切面做吸附小分子要不要考虑自旋极化?
不吝指教
答:A、这个不好说
要看你计算什么物质
自旋极化对他们有没有影响,如果有影响必须考虑
如果没有影响,或影响很小,可以忽略

B、假如A2B3,是主族的,切面三层分子式为A6B12,五层A12B21,七层是A18B30,都不是A2B3的整数倍,考虑自旋极化,五层七层不收敛,是
不是不用自旋极化?
还有要是不用自旋极化,如果做吸附NO,NO 分子有单电子那要不要考虑自旋激化?
谢谢指点!
C、不收敛和用不用自旋极化应该没有关系,不收敛应该检查Cut-off energy、smear、 Band数、体系是否合理等问题
39、请问各位大侠,在很多文章中都可以看到电荷密度图
这是怎么来的?
如何在CASTEP中实现电荷密度图示????
答:A、电荷密度图在ms里面是很容易实现的,首先你在选择计算任务的时候,找到电荷密度的单词,然后选中进行计算。
算完以后,你在分析的时候,选择分析电荷密度,就可以看到电荷密度图示。既可以调节2d也可以看3d的图。
如果你是vasp算的电荷密度图,那么就要程序转换成ms认识的格式,才能导入到ms里面看了,一样是2d和3d都能显示。如果你用
lev00处理,当然也行,那就是origin直接画图,一般可能是等高图多。
B、view--->toolbars--->volume visualization, 可以调出creat slices,这样就可以选择2D 的electron density,然后选中这个
slice,在properties中修改 slice crystal direction(修改所取面的晶向),修改slice fractional postion(修改你所取的面的确
定位置),然后从工具栏选择color map, 即可得到想要的面的电荷密度
40、关于价态的问题
我算了一下TiO2的结构
但是它给出的结果如下:
Species Ion s p d f Total Charge (e)
==============================================================
O 1 1.87 4.79 0.00 0.00 6.65 -0.65
O 2 1.87 4.79 0.00 0.00 6.65 -0.65
O 3 1.86 4.78 0.00 0.00 6.65 -0.65
O 4 1.86 4.78 0.00 0.00 6.65 -0.65
O 5 1.84 4.80 0.00 0.00 6.64 -0.64
O 6 1.84 4.80 0.00 0.00 6.64 -0.64
O 7 1.84 4.80 0.00 0.00 6.64 -0.64
O 8 1.84 4.80 0.00 0.00 6.65 -0.65
O 9 1.84 4.78 0.00 0.00 6.62 -0.62
O 10 1.84 4.78 0.00 0.00 6.62 -0.62
O 11 1.84 4.78 0.00 0.00 6.62 -0.62
O 12 1.84 4.78 0.00 0.00 6.62 -0.62
Ti 1 2.24 6.25 2.23 0.00 10.72 1.28
Ti 2 2.24 6.25 2.23 0.00 10.72 1.28
Ti 3 2.21 6.24 2.24 0.00 10.69 1.31
Ti 4 2.21 6.24 2.24 0.00 10.69 1.31
Ti 5 2.25 6.26 2.18 0.00 10.69 1.31
Ti 6 2.25 6.25 2.19 0.00 10.68 1.32
Ti 7 2.25 6.26 2.18 0.00 10.69 1.31
Ti 8 2.26 6.27 2.17 0.00 10.70 1.30
理论上说:Ti应该是+4价,O是-2价,为什么才是-0。6和1。3多呢?
希望高手指教
答:A、你说的是一个分子式 的价态,而计算的结果是一个晶胞上每个原子均匀分得的电子量是晶胞能量最低的优化结果,从计算结
果可以看到即使是同一个原子价态也不同,这是由于各个原子在空间分布不同。
B、离子键或者共价键之类的吧,共用电子的程度不同.
另外也顺便有个疑问:改变某原子的电荷在计算中有影响吗?
C、这个是否与你定义的原子的半径有一定的关系,如果你定义的O的原子半径大,则所带的电量就大,反之亦然。不知道MS里面是不
是这样,不过VASP里面是这样的
D、前面说的是设置问题
关于你说的计算结果,理解是这样的:
以O1为例子
Species Ion s p d f Total Charge (e)
======================
O 1 1.87 4.79 0.00 0.00 6.65 -0.65
计算以前O1的电子结构是 2s2 2p4, Total =6(e )
计算后O1的结构变为2s1.872p4.79,Total =6.65(e )
-0.65 表明优化以后,O1得到0.65(e )
E、如果考虑的是纯离子,当然就是+4和-2了。
但是由于Ti和O之间形成共价键,价带不是完全有O原子轨道构成
而是有O原子和Ti原子轨道贡献构成。这部分贡献可以通过PDOS可以分析出来。
CASTEP中PDOS面积表示电子数。
也就是由于形成共价键,导致价带有Ti原子的贡献,当电子在价带上填充时
填充在价带中Ti原子所贡献的那部分态上的电子划归为Ti原子的电子,这样
由于价带不是纯O原子的轨道贡献,当然不可能得到2个电子,所以也就不可能是-2价。
如果Ti和O是纯离子,也就是Ti的原子轨道在价带上对O轨道的贡献可以忽略,这时候
才可能打到+4和-2价。
由于Ti-O之间是共价键,所以O和Ti不可能为-2和+4价。

如何阅读文献

如何阅读文献
1。由点到面。选工作实践中的疑点,热点,由一个小枝节,检索较全的文献,一般近期的20 篇左右已经相当多了。之所以不必在意3年以前的,是因为知识更新非常快,且网上能查到的多为近几年的全文。学习别人是怎么发现解决问题的。知道目前对这个问题的共同看法,和分歧。然后,扩展开,根据兴趣和研究的目的,知道,在研究的领域:谁的文章被引用的次数多,谁的文章最多最新最有启发性。去图书馆找他的文章看全文。逐步扩展自己的视野,构建个人的专业知识结构和看法。
2。由杂到精。有了一定的知识基础以后,对于繁杂的文献,要有个人的判断。追踪某个专题、某个专家的研究进展,比较对于同一专题的论点的发展,掌握其新的方法或新结论,或注意作者观点的改变,探究其原因。培养个人的学术修养。对于高质量高水平的期刊,定期浏览,从面上了解学术进展和热点,根据个人的兴趣和工作进展,逐篇仔细阅读新作。
3。好记性不如烂笔头。无论是工作中的点滴发现,思想火花,都应该写下来。我和王忠诚院士、顾玉东院士的接触中,发现他们都有记卡片的习惯。病例随访、文献观点,等等。到写作文章时,都是现成的材料。现在有了电脑,但是写文献综述是一个完善知识结构的好方法。随时记下论点,个人心得,会有事半功倍的成绩。无论写在纸上,还是记载在电脑内,都应该有一个记事簿,并且经常整理。
4。对于下载的文献,要以其内容建立以专题杂志按时间先后的专门分类。哪些需要仔细阅读并保存,哪些用处不大,待删除,哪些需要阅读却尚未阅读。以后想到时,还能及时找到。
5。天天学习。文献天天有。如果只作为一个收藏家,就失去了研究的意义。下载的目的是学习。通过阅读,掌握专业领域的方法和知识。只要坚持学习,就会积累起自己的知识架构。水到渠成,游刃有余。
之二:
对于初次进入一个领域的新手,必须阅读大量的文献,才能把握本领域的动态和方向。
记得一个留洋的研究生说,起初导师让他读大量的文献,而且每天都规定了数量,好像是100篇吧?由于刚刚接触这一领域,对许多问题还没有什么概念,读起来十分吃力,许多内容也读不懂。请教导师,却被告知只要每天把数量读够就行了。后来随着阅读量的增加,终于最后融汇贯通,也理解了导师的方法。
所以,我觉得对新手而言,应当重视阅读文献的数量,积累多了,自然就由量变发展为质变了。
而且,每个作者的研究方法多少有所区别,读得多了,渐渐就会比较出研究方法的优点和缺点,对自己今后的研究大有裨益。
其实,由于现在科技进步很快,即使是自己从事的领域,也有很多新技术、新观点不停的出现,所以,即使是个“老手”,如果懒于更新自己的知识,也会很快落后。
之三
在文献多如牛毛的今天,其感觉犹如日益增多的帖子,在不想漏掉好东西的前提下
1 把握动态,每天入园后先浏览一遍自上次登陆以来的新帖子,感兴趣的再看看,必要时保存或收藏,这大体相当与杂志的目录,以此保证不会漏掉新的东西
2 回过头来重点看看感兴趣版块的精华帖、高人气帖、加密帖等,这相当于感兴趣砖头杂志的综述、评论、或是cutting edge(JI)之类的
3 好东西不仅仅只在感兴趣版块,其他版块也有一些,在空的时候看看其他版块的精华帖、高人气帖、加密帖等,这相当于相关砖头杂志的综述、评论、或是cutting edge(JI)之类的
4 用搜索功能查找特定关键词的帖子,大体相当于定题检索
5 随着对领域的熟悉,特别是对大师级人物的熟悉,有时可看看这些作者的帖子
6 在信息的今天,没谁敢说已经超一流,不需再接受新的东西;更没有人能够熟悉所有的领域
之四
邹承鲁院士写他是如何读文献的:
无论题目从何而来,都必需紧密追踪当前有关科学领域发展的动向。从研究生时代开始,在导师 教导下,以周围同学为榜样,我就养成了每周必定去图书馆浏览最新期刊的习惯,几十年如一日,雷打不动。如果确实有事,下周必定补上。我当时有一个小记录册,登录所有对本专业重要的刊物,每期读过后,一定做记录,决不遗漏一期,直至今日。现在可以在网上阅读所有重要刊物的目录和摘要,这就更容易做到了。掌握文献、对文献进行综合,以批判的眼光评价文献,并从中提取出有用的和正确的信息以指导今后的研究是一个能独立工作的科学工作者必备的能力。
阅读文献以追踪当前发展动态时,务须切记发挥自己判断力,不可盲从,即使是知名科学家和教科书有时也会有错误。古人说得好:“ 尽信书不如无书”。在追踪当前发展的重要方向时切记,你看到的问题别人也同样会看到,越是重要的问题竞争必然越是剧烈,在研究条件不如人时,如果没有创新的研究思想,独到的研究方案是不可能超越他人得到成功的。虽然国际上也有对于某些重要课题一哄而起的情况,但在我国似乎特别严重。缺乏自己的创新思想而片面一哄而起追求热点,是一条必然失败的路线,最多只能是为别人成果锦上添花,或做一些小修小补的工作而已。关键在于自己的创新思想。创新思想来自何处,虽然灵机一动产生了重要的创新思想,在科学史上确实有所记载,但这毕竟是比较罕见的,而远远更为常见的是天才出于勤奋,创新出于积累,积累可以是个人积累,也可以是本人所在单位的长期积累。这就是前面提到的旺火炉原理,也是诺贝尔奖经常出在少数几个单位的原因。只有勤奋努力才能不断有优秀工作的积累,才可能在工作中逐渐产生真正创新的,别人无法剽窃的创新思想,才有可能在重大问题上取得突破。而在一个炉火熊熊的旺火炉中,不断会有优秀工作的积累,优秀人才的产生,并且创新思想和人才的不断相互作用,相互启发,相互激励,就会不断创造出新的突破性成果。
之五:
MIT人工智能(AI)实验室的教授和学生总结的如何读论文,还不错!
阅读论文是需要练习的技能。不可能完整地阅读所有的论文。
阅读论文可分为三个阶段:
第一阶段是看论文中是否有感兴趣的东西。AI论文含有摘要,其中可能有内容的介绍,但是也有可能没有或者总结得不好,因此需要你跳读,这看一点那看一点,了解作者究竟做了些什么。内容目录(the table of contents)、结论部分(conclusion)和简介(introduction)是三个重点。如果这些方法都不行,就只好顺序快速浏览了。一旦搞清楚了论文的大概和创新点,就可以决定是否需要进行第二阶段了。
在第二阶段,要找出论文真正具有内容的部分。很多15页的论文可以重写为一页左右的篇幅;因此需要你寻找那些真正激动人心的地方,这经常隐藏于某个地方。论文作者从其工作中所发现的感兴趣的地方,未必是你感兴趣的,反之亦然。
最后,如果觉得该论文确实有价值,返回去通篇精读。  读论文时要牢记一个问题,“我应该如何利用该论文?”“真的像作者宣称的那样么?”“如果……会发生什么?”。理解论文得到了什么结论并不等同于理解了该论文。理解论文,就要了解论文的目的,作者所作的选择(很多都是隐含的),假设和形式化是否可行,论文指出了怎样的方向,论文所涉及领域都有哪些问题,作者的研究中持续出现的难点模式是什么,论文所表达的策略观点是什么,诸如此类。
之六
1.多数文章看摘要,少数文章看全文
掌握了一点查全文的技巧,往往会以搞到全文为乐,以至于没有时间看文章的内容,更不屑于看摘要。真正有用的全文并不多,过分追求全文是浪费,不可走极端。当然只看摘要也是不对的。
2.集中时间看文献
看过总会遗忘。看文献的时间越分散,浪费时间越多。集中时间看更容易联系起来,形成整体印象。
3.做好记录和标记
复印或打印的文献,直接用笔标记或批注。pdf 或html 格式的文献,可以用编辑器标亮或改变文字颜色。这是避免时间浪费的又一重要手段。否则等于没看。
4.准备引用的文章要亲自看过。
转引造成的以讹传讹不胜枚举。
5.注意文章的参考价值。
刊物的影响因子、文章的被引次数能反映文章的参考价值。但要注意引用这篇文章的其它文章是如何评价这篇文章的:支持还是反对,补充还是纠错

Materials Studio在cluster上的安装指南

Materials Studio在cluster上的安装指南
ID:205 分类:Materials Studio/安装与编译 日期:2005-8-18

关键词:MS, Materials Studio, cluster, install, 安装

推荐使用 Rocks (http://www.rocksclusters.org/),安装和管理比较方便

一、安装Rocks 4.0 (详见 rocks-usersguide-4.0.0.pdf)


二、装好之后Enabling RSH on frontend and all Compute Nodes (详见 rocks-usersguide-4.0.0.pdf)
# cp /home/install/rocks-dist/lan/arch/build/graphs/default/base-rsh.xml \
/home/install/site-profiles/4.0.0/graphs/default/
Where arch is your architecture ("i386", "x86_64" or "ia64").

edit /home/install/site-profiles/4.0.0/graphs/default/base-rsh.xml and change the
following:


to:

xinetd
rsh

(也就是去掉这段中的第一行和最后一行)

To apply your customized configuration scripts to compute nodes, rebuild the distribution:
# cd /home/install
# rocks-dist dist
Then, reinstall your compute nodes.

最重要的是别忘了在frontend上也要启动rsh
具体方法是到安装包中找到
rsh-0.17-25.1.i386.rpm
rsh-server-0.17-25.1.i386.rpm
并安装, 然后启动
#/sbin/chkconfig --level 345 rsh on
#/sbin/chkconfig --level 345 rlogin on
#/sbin/chkconfig --level 345 rexec on
#service xinetd restart


三、安装MS Modeling (参考zixia量化版上qchem写的指南)
1. 获得MS Modeling 3.1的安装光盘或镜象文件
2. 用普通用户执行/mnt/cdrom/UNIX/Install --type cluster
3. 输入要安装的目标全路径,如/home/public/MSI/MaterialsStudio
4. 选择要安装的模块,如17全安装
5. 选择License管理软件的安装目标目录,输入两个回车即可
6. 选择是否启动Gateway服务,一般选是
7. 选择安装license文件,选择3不装
8. 复制你的license文件到/home/public/MSI/License_Pack/licenses (目录要与安装
License管理软件的目录相一致)文件名必须改成msilicense.dat


至此就可以跑串行的程序了,要并行还必须加把劲



四、copy MaterialsStudio\hosts.equiv to /etc/hosts.equiv或者自己新建一个
内容如下:
localhost
localhost.localdomain
compute-0-0
compute-0-1
compute-0-2
.
.
.
compute-0-6
compute-0-7
.
.
.

在安装用户根目录下建一个.rhosts文件, 内容跟hosts.equiv一样



五、修改 MaterialsStudio/MPICH/share/machines.LINUX 文件,把要用的计算节点写上去,格式如下:
compute-0-0:2
compute-0-1:2
compute-0-2:2
.
.
.
compute-0-6:2
compute-0-7:2
.
.
.
冒号后为每个节点的cpu


六、如果用的是csh, 在安装用户根目录下修改.cshrc(bash用户修改.bashrc), 内容如下:
eval `/home/msi/MaterialsStudio/Licensing/Setup/lic_setup.sh -s csh`
eval `/home/msi/MaterialsStudio/share/bin/ms_setup.sh -s csh`

setenv TMPDIR /home/msi/tmp
setenv DMOL3_DATA /home/msi/MaterialsStudio/Data/Resources/Quantum/DMol3

setenv DMOL_TMP /home/msi/tmp
setenv MESODYN_HOME /home/msi/tmp
setenv PATH ${PATH}:/home/msi/MaterialsStudio/DMol3/bin

并在安装用户根目录下建一个tmp临时文件夹



七、 修改MaterialsStudio/Gateway/root_default/dsd/conf下两个文件gw-info.sbd gwparams.cfg中的total cpu
否则在客户机上的job_control中只能看到安装节点上的cpu



八、远程客户机用Server Console工具
选中左边的Server Gateways,右键新建server gateway
输入前述主机的IP地址
建好的server gateway,可以在上面点右键运行测试作业,以测试正常与否。

记得先用小任务测试一切是否正常。

...

enjoying it and good luck!

Sunday, June 29, 2008

Molecular Engineering of Low Friction and Biocompatible Surfaces

a good proposal abstract

molecular dynamics simulation of biotribology

one of the papers of the PI


Molecular simulation studies of nanoscale friction between phosphorylcholine self-assembled monolayer surfaces: Correlation between surface hydration and friction
J. Chem. Phys. 127, 084708 (2007); DOI:10.1063/1.2759910

http://link.aip.org/link/?JCPSA6/127/084708/1
http://scitation.aip.org/getpdf/servlet/GetPDFServlet?filetype=pdf&id=JCPSA6000127000008084708000001&idtype=cvips&prog=normal


http://www.nsf.gov/awardsearch/showAward.do?AwardNumber=0758358

Award Abstract #0758358
Molecular Engineering of Low Friction and Biocompatible Surfaces


NSF Org: CMMI
Division of Civil, Mechanical, and Manufacturing Innovation


Initial Amendment Date: June 3, 2008

Latest Amendment Date: June 3, 2008

Award Number: 0758358

Award Instrument: Standard Grant

Program Manager: Clark V. Cooper
CMMI Division of Civil, Mechanical, and Manufacturing Innovation
ENG Directorate for Engineering


Start Date: June 1, 2008

Expires: May 31, 2011 (Estimated)

Awarded Amount to Date: $250000

Investigator(s): Shaoyi Jiang sjiang@u.washington.edu(Principal Investigator)


Sponsor: University of Washington
1100 NE 45th St, Suite 300
SEATTLE, WA 98105 206/543-4043


NSF Program(s): NANO NON-SOLIC SCI & ENG AWD,
MATERIALS DESIGN & SURFACE ENG

Field Application(s): 0106000 Materials Research

Program Reference Code(s): MANU,9146,1633,1444,024E

Program Element Code(s): 7237,1633

ABSTRACT

Lubrication is critical to the success of total joint replacement. Despite significant progress in the understanding of molecular tribology for chemical systems, there is still a lack of a fundamental understanding of the lubrication mechanisms for joints and other biological systems. Currently, the most common solution to arthritis is total joint replacement. However, the biocompatibility of implanted materials poses a great challenge. Poly(ethylene glycol) (PEG) and zwitterionic-based materials are two commonly used biocompatible materials. It is hypothesized that zwitterionic-based coatings will have very low friction. In this work, friction between two surfaces covered by zwitterionic polymer brushes will be studied using both non-equilibrium molecular dynamics simulations and chemical force microscopy experiments. The objectives of this work are to gain insights into the origin of the friction between zwitterionic coatings at the molecular level, to establish the relationship between nano-scale friction and surface hydration, and to compare the performance of three unique zwitterionic coatings. PEG will also be studied for comparison. The success of this work will provide a fundamental understanding of the interfaces encountered in bionanotribology and will guide the design of new biocompatible materials for applications in artificial joints and BioMEMS/NEMS. The outcome of this work is to identify and design coatings with both low friction and high compatibility. Graduate and undergraduate students will be involved in this project, particularly those from underrepresented groups. As a part of the ongoing undergraduate curriculum reform, the PI is teaching a new course on ?Biomolecular Interfaces? and will teach a new undergraduate thermodynamics course which will integrate classical thermodynamics, statistical mechanics and molecular simulation. Results from the proposed work will contribute significantly to these two new courses. The knowledge will also be disseminated through several other courses that the PI gives lectures to every year.




Please report errors in award information by writing to: awardsearch@nsf.gov.

Friday, June 27, 2008

科研人员如何利用社交网络

科研人员如何利用社交网络



社交网络在年青人中间很流行——尤以Facebook, myspace有名。科研人员也有一些比较专业的交流网络社区,国外有2collab, 国内有“生物谷”、“丁香园”等等。

科研人员是怎么利用和看待这些新兴的web2.0的网络社区呢?是聊天、看新闻、分享图片、会朋友,还是从事与工作有关的事?

2collab最近做了一个调查。结果发现:科研人员毕竟是科研人员(科学家),还是很严肃的,他们利用社交网络主要的目的,还是与工作有关。

这是一篇英文文章,挺长,不翻译了,转贴下面供大家参考。

2collab Survey Reveals that Scientists and Researchers are “All Business” with Social Applications

Social media will have major influence on key aspects of research within five years

AMSTERDAM – June 9, 2008 – 2collab (www.2collab.com), the research collaboration platform from Elsevier, the world's leading publisher of science, technology and medical (STM) information, announced today the results of a survey, asking researchers about the role of social media in their professional lives. The survey, which yielded over 1,800 responses, revealed that scientists are using blogs, wikis, and social networking and bookmarking applications primarily for professional reasons. Results show that these social media applications have provided scientists and researchers with additional resources to help them collaborate, connect, share and discover information.

2collab surveyed science, medical and technical information professionals working in academia and government institutions to establish exactly what influence new web applications are having on the way scientific research is conducted. Over 50% of respondents see web-based social applications playing a key role in shaping the future of research. The largest influence will be on critical analysis and evaluation of research data, professional networking and collaboration, dissemination of research output, career development, as well as grant application and funding.

Results show that many researchers believe social applications will have a major influence on the future of research. One respondent, an Environmental Science researcher based in Spain commented, “Social media and electronic journals will be the future of scientific information dissemination. Current scientific journals must not disappear but the business model will change.”

Comments from survey respondents identified several issues need to be addressed before mass acceptance by the research community is possible – namely the need for specialist tools, higher security, and validation of users. However, these concerns were not seen as insurmountable obstacles, and many anticipated tremendous potential for social media. “Existing social networks are mostly used for casual social interaction between young people. In order to be more relevant to academia, networks with professional credibility and accountability will need to develop,” writes another respondent, a Canada-based associate professor of Biochemistry, Genetics & Molecular Biology.

“While it is clear that scientists and researchers will continue to use traditional sources for information discovery, the survey indicates that social media applications will provide additional indicators of quality and discovery,” states Brant Emery, development manager for 2collab. “In an era where information travels fastest digitally, online applications will offer these professionals what one researcher stated as a “source of power.” Creating these online scientific communities gives everyone a chance to offer their voices and participate in research, thus increasing the flow of communication, access to knowledge and helping accelerate scientific discovery.”

本文引用地址:http://www.sciencenet.cn/blog/user_content.aspx?id=30431

Sunday, June 01, 2008

海外职称汉译小议

由于各国职称体系的差别,海外的职称很难准确地翻译成汉语,因此使从海外回国谋职的学者和国内的用人单位感到困惑。光含有“Fellow”这个词的几个职称,就令人头大。比方说,Research Fellow,直译作“研究员”,本也不错,但长期以来国内“研究员”是科研单位的最高级技术职称,和海外的Research Fellow并不相应。再比如,在国外一些高科技公司有一种职称叫Fellow,这是最高技术职称,在Principal Engineer之上的。为说明其定位,引用某公司关于Fellow的说明如下:

Fellows represent the highest level of technical standing in the company. They possess a track record of outstanding contribution to product development. Fellows have introduced new technology of wide import for the company's products and can define areas of strategic importance. They can be assigned as mentors to other staff in the organization.

这类职称在国内如何对号入座?如果翻译成“总工程师”,那比Fellow低一级的Principal Engineer又翻译成什么?还有Postdoctoral Fellow,相当于国内的“博士后”,按说没什么疑问,居然也有些问题。

据说有一位从澳大利亚回国的学者,简历上写有“国家级博士后”字样,引起一些议论,主要是说“国家级”三字“可笑”,有自吹之嫌。我也认为 “国家级”的译法不妥,但也知道它并非全然凭空臆造。

澳大利亚的博士后有一类和美加等国及中国的相似,由课题负责人聘用,工资来自课题负责人的科研经费,目前基本起点工资为六万澳元左右。还有一类博士后叫做Australian Research Council Postdoctoral Fellow。这个Australian Research Council 简称 ARC,官方的定义是:   

The Australian Research Council (ARC) is a statutory authority within the Australian Government's Innovation, Industry, Science and Research portfolio. The ARC advises the Government on research matters and manages the National Competitive Grants Program, a significant component of Australia's investment in research and development.

看起来有点相当于中国的国家自然科学基金委员会,不同的是,中国国家自然科学基金委员会并不直接审批博士后资格(听说是这样,不知是否有误)。

澳洲的ARC博士后和普通的博士后在职称和待遇上没有太大区别,但其申请难度大得多,通常要比较优秀的博士毕业生才能得到,因为:

(1) ARC博士后不是向大学和教授申请,而要向ARC提交申请,大学作为提供研究条件的单位,还必须同时向ARC提供有关研究环境的资料。
  
(2) ARC博士后的申请要独立地写课题计划,由ARC指定的专家(通常3到4人)评议。

(3) ARC博士后有严格的名额限制。我知道90年代初的情况,以92何93年为例,当时全澳 大利亚一年里总共只有50个ARC博士后名额(包括不同的领域)。上面提到的“国家级博士后”的海归学者,是属2004年一批的,所以有人查了2004年的记录,ARC当年一共也才批准了约100名ARC博士后。

可见,那位海归学者在简历上自称“国家级”,未必是有意误导,或许是想表明他(她)是直接申请到国家科学基金的那类ARC博士后,而又找不到其它更确切的名称吧。

都知道译诗难,原来译职称更难。诗不能直译还可意译,职称既不能直译更不可意译。我个人的建议是:不译。诸位以为如何?

本文引用地址:http://www.sciencenet.cn/blog/user_content.aspx?id=18602

欲识真理,勿轻信之

(一)

高分子材料的“热流变简单性(Thermorheological simplicity)”给John Douglass Ferry教授(1912-2003)带来了荣誉,却曾给他的学生Donald J. Plazek带来了烦恼。

说起来,那是是上世纪50年代中期的事了。

话题要从高分子材料的粘弹性说起。粘弹性的一个重要表现是应力松弛,即当试样持续保持一定形变时,维持这一形变所需要的力会虽时间的延长而逐渐减小。也就是说,如果你把恒温下材料的应力松弛模量对时间作图,你会得到一条随时间衰减的模量-时间曲线。

材料的松弛模量不仅是时间的函数,而且受温度的影响十分显著。温度标志热平衡状态时分子及其内部原子的微观动能的平均值。随著微观平均动能的增加,原子克服化学键内旋转阻碍势垒的机率就增加,分子的柔性也随之增加,分子的自由体积也增大,分子间的缠绕比较松动,因此应力松弛过程加快。

所谓热流变简单性,是指有这样一类高分子聚合物,例如聚异丁烯及其它非晶态的热塑性塑料之类材料,温度的影响等效于把模量-时间曲线(时间轴为对数坐标)沿时间轴平移(有时要略作小的纵向移动修正)而曲线形状保持不变。这意味著,可以把在不同温度下作的只覆盖几个时间数量级的模量-时间曲线族水平移动叠合成一条覆盖十几个时间数量级的主曲线。所要求的平移量是温度的函数,称为“时-温平移因子”。

简单地说,升高温度与延长观察时间效果相同。

这就是著名的“时温等效原理”,或称“时温迭加原理”。时-温平移因子的计算公式很简单,在温度接近玻璃化温度和高于玻璃化温度以上100度的范围里,可用流行的WLF方程来计算。方程的名称来自三个发现者的姓:Williams,Landel和Ferry。WLF方程是一个经验方程,但Ferry根据自由体积的概念进行了阐释。在上述温度范围之外,计算时-温平移因子可用另外的关系式,例如Arrhenius方程。

Ferry在热流变简单性的研究领域作出了令人瞩目的贡献,而他的学生Plazek,却遇到难以启齿的尴尬,因为Plazek在实验室里测量了一些数据,和时温等效原理不相符。他相信是自己的实验搞错了,便偷偷地把这些“坏结果”藏在抽屉里,怀著小孩子作了错事的心情,不敢告诉Ferry教授。

(二)

四十年以后,有一篇论文发表在1996年第40卷Journal of Rheology上,它的标题赫然醒目:Oh, Thermorheological Simplicity Wherefore Art Thou?(《哦,热流变简单性,祢在哪里?》),作者不是别人,正是Donald J. Plazek。

在这篇文章里,Plazek公布了许多不符合时温等效原理的实验数据,并指出,热流变简单性要求高分子链中不同的运动单元的松驰具有相同的温度倚赖性,否则随著温度的变化,响应曲线除了改变位置外,还将改变形状,不可能简化到一条主曲线,他把这种现象称为“热流变复杂性(Thermorheological complexity) ”。他的数据表明,甚至连被认为最符合热流变简单性的聚异丁烯,也具有热流变复杂性。

热流变复杂性深化和丰富了热流变简单性的理论。使我们了解了,热流变简单性理论不是普遍适用的。时温迭加的失效,也不是实验的失败,而是提供了更多的信息,可以帮助我们判别和区分分子运动单元不同的温度倚赖机理。

为了表彰在流变学领域作出杰出贡献的研究者,国际流变学会自1948年起设立一年一度的Bingham奖。

对热流变简单性作出的重大贡献的J. D. Ferry荣获1953年Bingham奖。

对热流变简单性提出质疑的Donald J. Plazek荣获1995年Bingham奖。

(三)

你从一个口袋里随机地模出一只白玻璃球,第二次又模出一只白玻璃球,第三次,以至第十次,还是白玻璃球。于是你得出“口袋里都是白玻璃球”的猜想。如果后来模出个黑玻璃球来,这个猜想便不成立了。你可能有了新的猜想:“口袋里都是玻璃球。”可是,新猜想还是有可能被一个木头球给否定掉。“好吧,”你说,“口袋里都是球!”对不对?仍然要再检验,谁也不敢担保不会模出一只猫来。

口袋里的东西是有限的,所以早晚有真相大白的一天,可是自然界的事物是层出不穷的,这种检验也就没完没了。

显然,从有限数量的事实中不可能合法地证实一个全称命题为真。英国科学哲学家Karl Popper甚至得出一个惊人的结论,在有限数量的证据下,任何理论被证实的概率都等于零。他认为,一个理论之所以被认为是科学的,不是因为有多少证据支持它,而是因为它预先规定了这样一些事实,这些事实一旦被发现,该理论可被证伪。

基于此,他提出了如下科学理论发展逻辑的四段模式:

1) 从问题开始(the old problem);
2) 提出试探性的理论(formation of tentative theories);
3) 尝试通过包括判决性的讨论和实验来排除理论的错误(attempts of elimination through critical discussion, including experimental testing);
4) 由上述对理论的判决性讨论和实验所引出的新问题(the new problems that arise from the critical discussion of our theory).
然后,上述这个过程又重复进行。

Karl Popper的模式是有争议的,它似乎并不符合许多学科的实际。例如,这一模式彻底排斥了归纳法,但至少在我们知道的某些领域里,归纳法在提出试探性的理论这一阶段是功不可没的,前面说过,WLF方程最早便是从实验数据归纳出来的,而它的理论推导则是后来的事。

但有一点Karl Popper是对的,那就是,对一种理论的真正检验,是让该理论担受被证伪的风险,而不是寻找对该理论有利的证据,更不应该掩盖对该理论不利的证据。Plazek的故事也告诉了我们这个道理。

谈诗可以没有科学,但谈科学不可无诗。Plazek的论文的结尾引用了德国诗人Novalis的名句作为座右铭,原文如下:

To become acquainted
with a truth we must
first have disbelieved it and
disputed against it.

这是说,为了认识真理,首先我们必须不轻信它,并且要辩驳它。

本文引用地址:http://www.sciencenet.cn/blog/user_content.aspx?id=22364

神学·流变学·爱因斯坦

http://www.sciencenet.cn/blog/user_content.aspx?id=16719

一、“神学”?

友人来访,在书房里闲聊。

“嗳,什么是工程神学?”他突然莫名其妙地问了一句。

工程神学?我被他问糊涂了,老实回答不晓得。

他的目光越过我的肩膀,注视著我身后的什么地方。我回过头去,看看身后书架,恍然大悟。

“你是说这个吧?”我从书架上取下一本Roger Tanner 写的“Engineering Rheology”,递给他说,“这是工程流变学呀。”

流变学(Rheology)和神学(Theology)只差一个英文字母。

我安慰他说,很多人都曾搞错的。有一个叫Markus Reiner[1] 的学者,是流变学的奠基人之一,就说过他经常会收到一些信件,信封上写的收信人的地址是“神学实验室”;他甚至还收到邀请信,请他参加国际神学研讨会。

二、“流变学”命名记

Rheology这个词是怎么来的?这就要从1927年谈起了。当时,这个Reiner先生是巴勒斯坦的土木工程师,应美国化学家Eugene C. Bingham 的邀请到美国和Bingham 合作一项研究,在工作中,他们发现双方有那么多的共同感兴趣的问题,不约而同地感觉到,应该有一个介于力学、化学和工程科学之间的交叉学科。

“你,一个土木工程师,而我,一个化学家,我们合作解决共同的问题。随着胶体化学的迅速发展,这种情况会越来越普遍的。”Bingham说,“要建立一个物理分支来处理这类问题才好。”

Reiner说,有一个分支叫做连续介质力学。

“不行,”Bingham说,“那会吓跑化学家。”

他们请教了一位语言学家。凡是语言学家,都是言必称希腊的,这位语言学家创造了“Rheology”这个新词,就是来源于古希腊哲学家赫拉克利特(公元前540~公元 前475年)的名言:“万物皆流。”“Rheo”在希腊语里是流动的意思。

1929年4月29日,来自不同领域的一批杰出的科学家和工程师聚集在美国美国俄亥俄州首府哥伦布市,成立了流变学会,把流变学正式定义为研究材料的流动和形变的科学。这一天,就算是流变学这个混血儿呱呱坠地的生日了。

三、群山在上帝面前流动──流变学看世界

早在十七世纪,虎克已经阐述了弹性固体应力和应变的线性关系,牛顿已经阐述了黏性流体和应变率的线性关系。不过,我们与其把它们看作定律,不如把它们分别看作某一类材料的定义。定律必须是客观的,定义可以是主观的。也就是说,不管符合所描述的规律的固体或流体是否存在,都不妨碍我们定义它。

符合上述定义的固体和流体,被分别称为虎克固体和牛顿流体。流变学不掠人之美,通常不研究这两种材料,而研究“非牛顿流体”。你到厨房去,打破一个生鸡蛋,观察蛋清的流动行为,就会发现蛋清既有些粘性,又带些弹性,这是一类非牛顿流体,又叫黏弹性流体。工业中各类悬浮液、塑料熔体、橡胶、油漆、涂料等,生物流体中的关节润滑液、细胞液、血液等,都是非牛顿流体。在自然界,可以说,非牛顿流体普遍存在,虎克固体和牛顿流体仅在一定条件下才存在。

从流变学的观点来看,固体和流体的区分已失去其绝对性,流变学以“万物皆流”作为自己的标志就说明了这一点。许多流变学教科书和杂志都引用了旧约圣经记载的女先知底波拉(Deborah)的一句歌词:“群山在上帝面前流动……”(中文和合本和英文新国际本均写作“震动”,某些版本写作“熔融”) 这不表明流变学真的和神学有什么联系,而是借助这句歌词来阐述这样一个思想:山在上帝面前流动,而不是在人面前流动,就是说,如果你的观察时间足够长,或者说,如果材料的松弛时间相对于你的观察时间来说非常短,那么你将观察到材料的流动。反之,你看到的材料则呈现“风雨不动安如山”的固体状态。

换句话说,流变学里的材料是“因时而异”的。这里的“时”有两个,一个是材料本身的特征时间,另一个是观察流动变形过程的特征时间,这两个时间的比值,叫做“底波拉数”(底波拉,就是圣经里那位女先知)。前面提到的黏弹性流体,究竟是更多地表现为弹性,还是表现为黏性,就看在具体过程中底波拉数的大小了。

四、从微观到宏观

在传统的固体力学和流体力学中,本构关系已经被定义好了,不成问题了。然而,本构方程却是流变学的中心问题。

研究材料本构关系有两种方法──连续介质力学方法和微观结构模型(或称分子模型)方法。连续介质力学的方法只能以“唯象”的方式反映宏观尺度上所观测到的物理现象,但复杂流体系统中,宏观流动特性是由流体本身的微观结构所决定,这些微观结构可以是最细节的分子结构,也可以是结晶、悬浮物、不互溶的混合物以及填充物等不同层次的特征结构。微观结构模型(或称分子模型)的方法成为近年来流变学领域的发展方向。

微观结构模型方法的基本思路是,首先构建一个代表微观结构的力学模型,再通过对微结构的力学分析,写出运动方程,并与构型分布方程结合写出扩散方程,解扩散方程得到构型概率分布函数,然后通过统计方法得到流体的平均应力,这样,原则上可以得到表达宏观应力应变关系的本构方程。

微观结构模型方法的最早的尝试,可追溯到二十世纪初爱因斯坦关于糖水的研究。

五、爱因斯坦

1905年被称为爱因斯坦的“奇迹年”。在这一年里他不仅写了两篇引起时空观念革命的狭义相对论的论文,一篇后来让他获得诺贝尔奖的光量子以及光电效应的论文,还写了一篇有关糖水的博士论文和另一篇与此相关的分析布朗运动的论文[2] 。

那篇糖水博士论文,题目是《分子大小的新测定》。这项研究有深刻的科学史背景。当时,物质是否由分子和原子组成,还处在争议之中,原子和分子论只不过是一个看起来说得通的假定,并未被科学界完全接受。爱因斯坦为了帮助明确微观结构的概念,就研究如何从假设的微观结构推论出可测的宏观物理性质。如果理论预测的结果和实验直接测量的结果能够符合,那么假设的微观结构就有更多的真实性。

流体的一个重要的宏观物理性质是其黏滞系数,或称黏度。向水里加糖,黏度会增加,这个现象给爱因斯坦一个启发,是否可以利用经典流体动力学,计算出溶质分子的大小和数目对溶液黏度的影响呢?他的基本假设是,分散在水中的糖分子体积远大于水分子的体积,因此糖分子可以简化为悬浮在水中的刚性球形颗粒来处理,水则可以被近似为连续介质。同时,还假设这些颗粒的惯性可以忽略不计,它们彼此没有干扰等等。通过流体动力学的分析,爱因斯坦得到了溶液黏度对溶质分子大小和每单位溶液体积中溶质分子数目的依赖关系。应用这个结果,再辅以一个扩散方程,便可以从已知的糖溶液和水溶剂的黏度推算出糖分子的大小和Avogadro常数来。这个扩散方程,应用了关于球体在粘性介质中运动阻力的Stokes定律,现代的流变学著作通常称之为Stokes-Einstein方程。

第一个方程的推导是论文中技巧最复杂的部分,可是推导中出了点错,使得最终结果有一个系数不正确。爱因斯坦自己没察觉,苏黎世大学的评阅人也没发现,于是爱因斯坦先生如愿地成了爱因斯坦博士。

这个有错误的公式发表在1906年第19卷的Annualen der Physik上,刚开始并没有引起广泛的关注。1909年,爱因斯坦向法国物理学家Jean Baptiste Perrin建议把这个方法用于悬浮液问题的研究。后来,Perrin的学生Jacques Bancelin 写信给爱因斯坦,说自己的实验结果并不完全符合爱因斯坦的理论推算。这意味著,以下几种可能发生的错误,至少有一种发生了:

(1) 刚性球形颗粒悬浮液模型不足以描述糖溶液真实的微观结构。
(2) 从微观结构模型到宏观材料特性的推导过程存在逻辑(数学)错误。
(3) 实验有错误,或者实验的条件超出了理论的适用范围。

爱因斯坦首先自己重新检查了推导过程,但没有发现问题,于是又请自己的前学生、当时的合作者Ludwig Hopf 帮助再检查。结果Hopf发现了爱因斯坦原文中速度梯度计算中的一个错误。爱因斯坦在1911年第34卷Annualen der Physik上发表了仅有两页纸纠错误文章,并用纠正后的公式重新计算了Avogadro常数,得到6.56×1023 /mol,这个值很接近目前公认的6.02×1023 /mol。

怎么?爱因斯坦也会出错?当然会,因为他是科学家,不是上帝。科学本来就是沿著“实际问题──理论模型──实验检验──修改或推翻模型──再检验”这么一条路径走过来的。

六、简评

爱因斯坦博士论文审稿人Alfred Kleiner评论说,这篇论文“做了流体动力学中最困难的论证和计算。”

爱因斯坦的朋友和传记作者、物理学家Abraham Pais[3] 在“Subtle Is the Lord: The Science And the Life of Albert Einstein”一书中评价爱因斯坦的这篇论文说:“不难发现,这篇研究颗粒悬浮液总体流变性质的论文,有不同凡响的广泛的实际应用。”

爱因斯坦的另一个传记作家Walter Isaacson[4] 说,爱因斯坦这篇论文,是迄今为止他的著作中被引用最多的一篇。

对流变学来说,爱因斯坦的主要贡献是打开一条从微观结构模型出发建立本构方程的思路,至于爱因斯坦的模型本身,应用范围是很有限的,因为:

(1) 它预言,如果一些刚性球形微粒悬浮在液体中,那么,液体黏度增加的比率正好等於小球体积浓度的2.5倍。可是,除了浓度小于0.02的稀悬浮液以外,这个预言总是和实验不一致。这是因为,浓度稍大些以后,“悬浮微粒不影响彼此的运动”的假设就不成立了。

(2) 从刚性小球悬浮液模型推导出的宏观性质是各向同性的,而流变学所关心的许多材料却是各向异性的。例如短纤维悬浮液,必须考虑柱形的或椭球形的悬浮颗粒模型[5] 。更复杂些,象高分子溶液这种黏弹性流体,最简单也要构建一个由弹簧连接两球在粘性介质中运动的力学模型[6]。

有一个事实是很有些哲学意义的,这就是,对黏弹性流体,从微观结构模型得到的方程和用连续介质方法所得到的方程形式形同。这让人相信,自然规律是客观存在的,这些“方程”本来就隐藏在大自然的背后,科学家可以通过不同的途径把它们找出来,而不是造出来。

参考文献

[1] M. Reiner, The Deborah number, Physics Today, 62,Jan. 1964.
[2] Albert Einstein, Investigations on the Theory of the Brownian Movement, Ed. R. Furth, Transl. A.D. Cowper, Dover, N.Y., 1956.
[3] Abraham Pais , Subtle Is the Lord: The Science and the Life of Albert Einstein, Oxford Univ. Press, N.Y. , 2005.
[4] Walter Isaacson, Einstein: His Life and Universe, Simon & Schuster, N.Y. 2007.
[5] N. Phan-Thien and R. Zheng , “Macroscopic modelling of the evolution of fibre orientation during flow”, in Flow-induced Alignment in Composite materials, Chapter 3. Eds. T.D. Papathanasiou and D.C. Guell, Woodhead, Cambridge England, 1997.
[6] R.I. Tanner, Engineering Rheology, 2nd Edition, Oxford Univ. Press, N.Y. , 2000.


本文引用地址:http://www.sciencenet.cn/blog/user_content.aspx?id=16719

Monday, May 19, 2008

常用纤维术语

常用纤维术语

DTY:假捻变形丝称为DTY(D raw Tex-tu red Y a rn),也称弹力丝。

DTY网络丝:网络丝是指丝条在网络喷嘴中,经喷射气流作用,单丝互相缠结而呈周期性网络点的长丝。网络加工多用于POY、FDY和DTY的加工,网络技术与DTY技术结合制造的低弹网络丝,既有变形丝的蓬松性和良好的弹性,又有许多周期性和网络点,提高了长丝的紧密度,省去了纺织加工的若干工序,并能改善丝束通过喷水织机的能力。

POY与FDY:高速纺丝的纺丝速度为3000~6000m/m in,纺丝速度4000m/m in以下的卷绕丝具有较高的取向度,为预取向丝,通称POY(pre-o r ien ted ya rn)。若在纺丝过程中引入拉伸作用,可获得具有高取向度和中等结晶度的卷绕丝,为全拉伸丝,通称FDY(fu lly d raw ya rn)。

DT:拉伸加捻丝称为DT(D raw Tw is t)。以POY为原丝经牵伸加捻机,拉伸为主并给予少量的捻度,即可得到DT。 100D/36F,150D/36F,50D/18F等,这些是纤维规格的表示方法。斜线上方的数据表示纤维的纤度,D是纤维的纤度单位“旦”,即在标准状态下,以9000米长纤维的克重表示,如100克重即为100旦(100D);斜线下方的数据表示纺丝时使用喷丝板的孔数,也表示该规格的丝具有的单丝根数,例如36F,意思是纺丝时使用的喷丝板有36孔,即该纤维有36根单丝。

大有光,半消光,全消光:为了消除纤维的光泽,采用在熔体中加入二氧化钛(T iO 2)以消减纤维的光泽。如果在熔体中不加T iO 2为有光丝(或大有光丝),加入0.3%为半消光丝,大于0.3%为全消光丝。

50D/18F铁:为50旦18孔,铁管卷装。75D/36F纸:为75旦36孔,纸管卷装。150D/36F阳离子:为150旦36孔,经阳离子改善染色性能。

210D/72F肥瘦丝:为210旦72孔的竹节丝。“肥瘦丝”为非标纺织词汇,一般理解为“竹节丝”,即周期性的一段粗一段细的丝。

POY:预取向丝,全称:PRE-ORIENTED YARN 或者PARTIALLY ORIENTED YARN。 指经高速纺丝获得的取向度在未取向丝和拉伸丝之间的未完全拉伸的化纤长丝。与未拉伸丝相比,它具有一定程度的取向,稳定性好,常常用做拉伸假捻变形丝(DTY)的专用丝。(一般不用于织造)

DTY:拉伸变形丝,全称:DRAW TEXTURED YARN。是利用POY做原丝,进行拉伸和假捻变形加工制成。往往有一定的弹性及收缩性。(一般分有网络和无网络丝,网络的意思是网络节点)

FDY:全拉身丝。全称:FULL DRAW YARN。采用纺丝拉伸进一步制得的合成纤维长丝。纤维已经充分拉伸,可以直接用于纺织加工。(一般就叫它长丝)

Wednesday, April 30, 2008

11 Reasons Why Manuscripts are Rejected

11 Reasons Why Manuscripts are Rejected

San Francisco Edit

Manuscripts submitted for peer review publication may be rejected for a
number of different reasons, most of which are avoidable.

It should be noted that the reasons for accepting manuscripts are not the
mirror image of the reasons for rejecting manuscripts. The main reasons for
accepting manuscripts are: their contribution and relevance to the field,
excellence of writing, and quality of the study design.

Many journals expect reviewers to assess the scientific merits and validity
of research in submitted manuscripts; however, reviewers can become critical
of manuscripts containing numerous language errors, which are difficult to
eliminate without careful editing. Scientific writing demands both good
science and well written manuscripts.

Following are the principal reasons why manuscripts are rejected. They are
all equally important because reviewers tend to focus on different issues
depending on their individual concerns and the journal's requirements.

1. Poor experimental design and/or inadequate investigation. An inadequate
sample size, a biased sample, a non-unique concept, and scientific flaws in
the study are common faults.

2. Failure to conform to the targeted journal. This is a common mistake.
The focus of the manuscript is not within the scope of the journal and/or
the guidelines of the targeted journal are not followed. This can easily be
avoided by reading the targeted journal and reviewing the author guidelines.

3. Poor English grammar, style, and syntax. Though poor writing may not
result in outright rejection of a manuscript, it may well influence the
reviewer's and editor's overall impression of the manuscript. It has been
shown that a well written manuscript has a better chance of being accepted.

4. Insufficient problem statement. It is important to clearly define and
appropriately frame the study's question.

5. Methods not described in detail. Details are insufficient to repeat the
results. The study design, apparatus used, and procedures followed must be
made clear. In some cases it might be better to put too much information
into the methods section rather than to put too little; information deemed
unnecessary can always be removed prior to publication.

6. Overinterpretation of results. Some reviewers have indicated that a
clear and ''honest'' approach to the interpretation of the results is likely
to increase the chances of a manuscript being accepted. Identify possible
biases and confounding variables, both during the design phase of the study
and the interpretation of the results. Describe experimental results
concisely.

7. Inappropriate or incomplete statistics. Using inappropriate statistical
methods and overstating the implications of the results is a common error.
Use an appropriate test and do not make the statistics too complicated.
Quantify and present findings with appropriate indicators of measurement
error or uncertainty (such as confidence intervals).

8. Unsatisfactory or confusing presentation of data in tables or figures.
The tables or figures do not conform in style and quantity to the journal's
guidelines and are cluttered with numbers. Make tables and graphs easy to
read. Some editors may start by looking quickly at the tables, graphs, and
figures to determine if the manuscript is worth considering.

9. Conclusions not supported by data. Make sure your conclusions are not
overstated, are supported, and answer the study's questions. Be sure to
provide alternative explanations, and do not simply restate the results.

10. Incomplete, inaccurate, or outdated review of the literature. Be sure
to conduct a complete literature search and only list references relevant to
the study. The reviewers of your manuscript will be experts in the field
and will be aware of all the pertinent research conducted.

11. Author unwilling to revise the manuscript to address reviewer's
suggestions. This can easily be resolved. Taking the reviewers'
suggestions into account when revising your manuscript will nearly always
result in a better manuscript. If the editor indicates willingness to
evaluate a revision, it means the manuscript may be publishable if the
reviewers' concerns could be addressed satisfactorily.

我们的学术论文引用率为什么低? 存在六大问题

我们的学术论文引用率为什么低? 存在六大问题

     2008-03-06 来源: 人民日报

  编者按:随着我国科技事业的发展,我国学术论文的数量和质量都在不断上升。但
是以引用率为指标衡量,我国学术论文总体水平与国际水平相比还有相当差距。本版刊
登的两篇文章,就如何看待学术论文引用率低、怎样提高引用率等问题进行探讨。

  学术论文引用率为什么低?

  ● 学术论文存在六个问题:引言笼统、不介绍同行工作、参考文献太少、缺乏研
究过程、缺乏原创、不介绍已有成果

  ● 学术论文引用率低的四大原因:浮躁风、传统文化的影响、科学历史、科学技
术上不成熟

  ● 提高学术论文质量,要从提高引用率开始

  改革开放30年来,中国科技界发生了翻天覆地的变化,其中一个比较有特色的、人
们长年累月关注的事情,就是每年向全世界发布的SCI(科学引文索引)统计结果当中
,我国的学术论文总数是多少,论文增长多少篇,被引用多少次。从整体上看,我国的
论文总数在增长,引用率也在提高。但人们在欢欣鼓舞的同时,也会发现:国内的论文
引用率仍然不够,跟我们的大国身份很不相称,和科技投入不相称;投入的变化十分巨
大,引用的变化非常缓慢。
  学术论文质量存在哪些问题?

  我认为我们的学术论文存在六个问题:

  第一个问题是引言过于笼统,起不到引言的作用。对研究背景、内容、工作的意义
,没有实质性描述。第二个问题是整个学术论文没有对同行工作的介绍,把我们的学术
论文一放到国际背景下,看起来有点像小学生的论文。第三个问题是作者手上没有有分
量的文献,参考文献量太少,有的只有两三篇,有的根本没有参考文献。第四个问题是
在写作方面缺乏研究过程的重要环节的完整信息,多数人不会结合讨论来解释他得到的
结果。科学技术、理工农医的论文的重要一环是解析实验结果,而我们的一些论文往往
以结论代替讨论,没有讨论,人们就不知道它有什么重要性和创新性。第五个问题是原
创性的论文没有第一手的研究结果,自己研究分析一通,得出的是别人的结果。第六个
问题是不介绍同类工作已有的结果,不提人家研究到什么程度了,只说自己是世界先进
水平,这是违背科学道德的。

  解决这些问题,要从很多方面入手。我想,如果我们能从正面提倡引用别人的论文
,就能解决我们这六个问题中的好几个问题。如果他能够认真地引用论文,前三个问题
就解决了,第四个问题也可以部分的解决了,因为作者就会明白什么叫讨论,什么叫结
果,什么叫结论。作者手上没有好论文,就会形成低水平论文指导低水平论文的局面。

  从有近代科技以来,以中国人命名的定律、方法、结果极少。比如,我们有爱因斯
坦方程,但很少听到张三方程,李四方程。我们感到很遗憾,我们自己也有责任,有主
观的因素,也有客观的因素。不管怎么说,我们已经做过的工作不引用,使得我们丧失
了很多话语权。

  为什么中国人自己的论文引用率低?

  我们的学术论文引用率低,就像粮食烂在仓库里,没有人发现。提高学术论文质量
,要从提高引用率开始。我们希望引用率能不断提高,不但要引用国外的论文,还要引
用中国的论文,不但要引用自己的论文,还要引用同行的论文,要引用有分量的论文,
要引用被公认的论文。

  我认为,学术论文引用率低有四个原因:

  第一就是浮躁风。有的人写论文的目的只是为了评职称,评教授、博导,是为了拿
毕业证,拿学位证书,只要有论文就行了,发表是目的,引用不引用无所谓,不管天下
还有谁跟他做一样的工作。还有人认为,引用了别人的观点,自己不就算不上创新了嘛
?为了戴上创新的帽子,宁愿掩耳盗铃。确有这样的事,我不引用,我不犯错误,但我
引用了我就评不上一等奖,为什么?不是原创性了!这是对原创性的片面认识。有了青
藏高原,才有喜马拉雅山,才有珠穆朗玛峰。没有这三个层次,就没有珠穆朗玛峰。可
是,我们许多论文给人的感觉就是,看不出他的基础工作,同行的工作,不知道他的“
青藏高原”在哪,“喜马拉雅山”在哪,但是你已经知道他是“珠穆朗玛峰”了,是世
界领先了,这就是浮躁风。

  第二个原因就是传统文化的影响,自我谦虚,同时又文人相轻。谦虚使自己原来做
得很好的工作不引用。一个科学家从事工作十几年,他发表论文起码有二三十篇,他自
己写文章从来不引用自己以前的文章,他认为那是丢脸的,要被别人说是自引,不知从
什么时候起,人们认为自引不好。科学技术的连续性、发展性要求不管是谁的,是对的
就可以引用。文人相轻使得不屑于引用别人的论文。

  第三个原因就是科学历史造成了我们引用少。什么叫科学历史?因为现代科技对中
国人来讲是个外来文化,都是舶来品,因此自己的东西没有什么可以引用的。没有引证
的需要,没有形成传统。而在西方发达国家,引用是一个传统,是科技界的道德,第一
道德就是要引用,公开说张三干的就是比我好,这就是西方传统。

  第四个原因还要归因于我们在科学技术上不成熟。我们在科学技术上稳定的发展科
技才30年,这30年对我们来讲很宝贵。但是,我们要实事求是的说,人家已经发展了
200年、300年。科学技术上的不成熟导致了我们缺乏引用,但我们不能等到成熟了才行
,我们要跨越式的发展。

  所以,我倡导提高学术论文引用率,一个是写论文的专家要多引用论文,包括别人
的论文和自己的论文;二是从事科技管理的专家要鼓励从各方面采取措施提高引用率,
把这两件事做好,我相信中国学术论文的质量会逐渐提高。(作者为中国科协书记处书
记冯长根)

  评价论文不以中西划线

  引用中国人论文并不丢份

  提高科技论文引用率,就要提倡中国的科研人员引用中国人自己发表的科技论文。
当然,无论是谁发表的(中国人的、外国人的),只要是好的,都要引用。但目前的现
状是,相当一批博士生导师对学生说,你们不要看中文的文献,你们就看外国的东西。
他的用意可能是好的,你不看外国文献怎么把握世界科技发展的前沿。另外有的学生懒
,外文不过关,博导就逼着他们读原文文献。这样下去,如果学生只看外文文献,不看
中国文献,怎么会引用?目前中国人的优秀论文不是很多,但也不是没有。因此,要提
倡引用中国人自己发表的科技论文,像提倡使用国货那样,是很有必要的。像钱学森这
样的科学大师,都很注重引用中国人的论文成果。

  提高科技论文引用率,科技学术期刊应扮演很重要的角色。如果科技学术期刊要求
严格,凡来稿的参考文献不符合要求的就退回去,就会促进文献引用。现在很多期刊靠
版面费生存,不想让参考文献占太多版面,有些编辑部甚至限制参考文献的数量。前些
年有些更野蛮的做法,明明投稿人有参考文献,编辑部发稿时加一个括号,说“参考文
献略”。

  期刊的影响因子大小和参考文献的平均篇数是直接相关的,而我们现在的整体情况
是,平均每篇论文的参考文献比国际上低得多。在本来就少的参考文献中,引用中国人
优秀论文的就更少。

  提高科技论文引用率,解决好“获取”(access)问题也是很重要的一个方面,那
就是我们的论文容易不容易被检索到、被阅读到。读到了还不一定引用,读不到更谈不
上引用。有一个例子,中科院物理所有一个《中国物理快报》,前几年他们拿到中国科
协、中国自然科学基金委的期刊补助是怎么用的呢?是用来印期刊抽印本,而有的杂志
拿到期刊补助就给编辑部人员发了奖金。国外大多数科学家并不是势利眼,看到你的文
章不错就会引用。由于《中国物理快报》将抽印引本直接寄到国外同行科学家手上,对
提高该刊引用率起到了明显的作用。现在他们不一定继续这么做了,因为网络已经很发
达。

  还有一个例子,20世纪90年代我在驻美使馆科技处工作期间,一个留美博士生给我
打电话,她说:“我的导师要编生命科学的百科全书,其中一个词条涉及牛胰岛素的合
成,我对导师说牛胰岛素是中国人最早合成的,导师说不知道。怎么办?”我和中科院
上海生物化学所联系,找到了当初发表人工合成牛胰岛素成果的杂志,把复印件传真过
去,那位导师就承认了,在词条中加了一句话:中国人首先合成牛胰岛素。如果access
的问题解决得好,人家自然会认账,就不需要留学生为中国争夺这个优先权而发生那么
多麻烦了。(中国科学技术信息研究所总工程师、研究员武夷山)

Saturday, March 01, 2008

我作RESEARCH的几点体会

http://www.mitbbs.com/article1/Economics/31145874_0_1.html

完全个人观点,不见得试用于所有人。

第一不要太迷恋TECHNICAL的东西。从自己和周围一些朋友的例子,我发现很多中国学
生对TECHNICAL的东西特别感兴趣。什么新的东西都想学,不管究竟对自己的RESEARCH
有没有帮助。所以很多中国学生的通病是,前两年上基础课的时候都很优秀,经常是全
A。但到了三四年级时,进入RESEARCH很慢。很多人的毕业论文都是匆忙上阵,甚至迟
迟不能毕业。当然我并不是说TECHNICAL的东西不重要。博士课前两年是积累一些
GENERAL TECHNICAL SKILLS很重要。剩下的时间主要集中在开始RESEARCH和学习与自己
RESEARCH相关的TECHNICAL SKILLS。

第二要在三年级就开始选定一到两个具体的TOPIC。我觉得从什么报纸,杂志或网站上
找灵感都是瞎掰。如果自己不是JOHN NASH那样的奇才,老老实实从文献里找灵感。看
看自己的导师在作什么,看看自己领域的大牛(如自己领域TOP JOURNAL的EDITOR)在作
什么。作研究都是大牛挖坑,小牛灌水。对于JUNIOR RESEARCHERS,我感觉最有效的研
究方法就是紧跟大牛挖坑的步伐。认准一个题目后,静下心好好清理一下LITERATURE的
脉络。不要朝三暮四,这山望着那山高,整天试图去追逐所谓的HOT TOPIC。

整理文献时先作一个LITERATURE TREE:谁提出了这个TOPIC,LITERATURE里都从什么方
向研究了这个题目,它们的关系是什么。对每个分支只包括最重要的1到两个文章。把
大牛和中牛抓住就行了,剩下的小鱼小虾就算了。参阅一些LITERATURE REVIEW的文章
,如果有的话。然后对经典的文章,如提出TOPIC的文章要进行复制。每个公式,每个
结果都试图从新作一边。如果有问题可以向原作者请教。我感觉他们大部份还是很NICE
的。复制的工作量其实很大,但也可以学到很多东西。很多东西不靠亲自作是无法体会
的。看看自己的导师和领域的大牛最近的研究在LITERATURE中的定位。复制自己感兴趣
的文章。

有了这些准备工作之后就可以进行自己的RESEARCH了。有没有什么办法把导师或大牛的
文章与其他研究联系起来,有没有什么EMPIRICAL STUDY可以做来支持导师和大牛的论
点,这些都是不错的研究方向。另外,一般文章中都会提出一些自己的不足和未来的研
究方向。这些都是不错的研究题材。我这里一直强调要跟大牛走。说白了,学术界其实
就是被这些人控制着。大家都说中国有学霸,学阀。其实什么地方都一样。论文就是观
点之争,不拉党结派怎么在江湖上混啊。对刚出道的人,最好还是厚道点。多作点
CONSTRUCTIVE的工作,少批评。我前面给的都是这方面的例子。江湖险恶,只是很多人
在作学生的时候不知道罢了。比如说JOURNAL EDITOR的权力其实很大。他们很清楚什么
人持有什么观点。他们喜欢的东西就发给持相同观点的审稿人,不喜欢的就发给不同观
点的审稿人。结果可想而知。以前的导师不久前让我审一篇稿子。作者声称发现数据支
持我导师一篇很重要的JPE文章。看了看文章的统计方法没什么大问题,我当然举双手
双脚赞成发表了。

当然也可以指出大牛文章的不足。但是不能为批评而批评。能找出改进方法解决问题的
文章才是好文章。比如可以这样写,我复制了A的模型,发现有什么地方和数据不符,
原因是什么。我对原模型作了一些改动后,现在模型和数据MATCH了。写作过程一定要
注意,多POSITIVE,少NEGATIVE。如在指出A模型和数据的冲突时,不要过多强调这种
冲突可能引起的问题。而应该强调解决问题后带来的好处。这样其实已经是对原作者的
间接批评了。如果A是领域大牛或作EDITOR,这篇文章十有八九能通过。

想跟上大牛步伐,看JOURNAL ARTICLES是不够的。这些JOURNAL上的东西至少是3-4年前
的东西了。要多留意WORKING PAPER。列出自己领域20大牛人。经常看看他们在干什么
,和自己目前的RESEARCH有什么联系。参加一些WORKING PAPER的EMAILING LIST,比如
NBER等。常看看TOP10 DEPARTMENT本领域的SEMINAR LIST。我想再强调一次,这样作的
目的并不是要去追逐本领域的HOT TOPIC,而是看看别人的研究和自己目前的RESEARCH
有什么联系,能不能给自己点灵感。至少在自己CAREER的前3-4年把RESEARCH INTEREST
限定在两个方向,每个方向整出2篇像样的文章,再考虑换方向。少而精要比泛泛涉猎
几个不同方向更容易引起别人注意。

就先谈这么多,以后有时间了再补充。

另外,要专心作RESEARCH就要减少在MITBBS的灌水时间。每天减少一小时灌水时间,一
年能多读多少篇论文啊。没有BBS北美广大WSN能平均提前1年毕业。:-)

Thursday, February 28, 2008

Materials research and the ‘energy crisis’

http://dx.doi.org/10.1016/S1369-7021(08)70035-X

ScienceDirect - Materials Today Materials research and the ‘energy crisis’

Volume 11, Issue 3, March 2008, Page 64

Opinion

Materials research and the ‘energy crisis’

David Cahena,
aWeizmann Institute of Science, Israel

Available online 15 February 2008.



Forget about ‘the’ solution. Instead, we need to work toward a strong, sustainable mosaic of many solutions.






I will try to outline my credo for why and how, from the basic science point of view, we should tackle energy-related materials research issues.

In the title, I put energy crisis in inverted commas. Why? Because for at least the next few hundred years, we do not lack a relatively cheap source of energy – we have coal (and coal liquefaction and gasification are known processes).

Unfortunately, because most coal is rather dirty, this leads to assured environmental and highly likely climate problems. Even those that doubt the latter cannot ignore the former. Visit some of the world's emerging industrial areas and you can quote from the 1960s Tom Lehrer song Pollution, “Don’t drink the water and don’t breathe the air,” and add in land pollution, although it does not rhyme. Thus, we find the true driving forces for weaning us off fossil fuels and for developing alternative, sustainable energy resources (ASER).

And there is another strong materials reason: it is an utter waste to just burn oil, as it is such a valuable natural resource. The long-term interest of oil-producing countries is to use oil as a natural materials resource. Indeed, legendary Saudi oil minister Yamani's quote, “the stone age did not end because of a lack of stones,” says it all.

Before giving my view on the roles for basic science in developing ASER, I should stress the importance of energy conservation to reduce the amount of fossil fuel needed to get the same amount and types of work done. There are materials-related issues such as improving insulation, reducing friction between moving parts, improving materials for natural lighting, and recycling materials. To these we can add improving waste heat use, reducing waste energy, and increasing the efficiency of current power-generation options (including renewable ones such as solar water heaters) with less pollution. Admittedly, these are not the glamor topics that may please your favorite journal's editor, as much as the following ones, but they are critical in the short term.




© The New Yorker Collection 1998 Frank Cotham from cartoonbank.com. All Rights Reserved.


Past experience teaches us that results of breakthroughs in basic science today will only start to be felt after some 15–20 years. Well, one may say, that is fine. But here is the catch. Basic scientific research in ASER decreased so much after the drop in oil prices in the early-1980s that we have a very narrow base of relevant fundamental science on which to build new technologies. That is why we need to kick start long-term support for the sorely needed basic research now.

Now let me stick my neck out: if I were a research program manager, I’d support:

1. Exploratory, blue-sky basic research per se as the best proven way to stumble on new ideas† ;

2. Optics – cheap ways to use larger parts of the solar spectrum in quantum conversion systems such as photo-(bio)chemical and photovoltaic devices;

3. Heterogeneous catalysis for reduction reactions‡ to bring us closer to the holy grail of efficient cheap artificial photosynthesis and find replacements for noble metal catalysts; and

4. Plant science.

Clearly, especially (2) and (3) present major challenges for materials research.

I am aware that my own field of photovoltaic materials is not mentioned specifically, although it definitely is in (2). Instead, I have tried to give a more holistic picture for a purpose. ASER suffered and suffers still from too many claims of the solution and we, the researchers, are the culprits. Even if such claims bring publicity, in the end they harm the whole area. It is my opinion that there is not one solution. Instead, we need to work toward a strong, sustainable mosaic of many solutions, which, as a whole, will provide the solution.


† We should, though, make sure that researchers are aware of the issues, i.e. Louis Pasteur's famous dictum, “Chance favors the prepared mind”, applies! [cf. Nat. Mater. (2008) 7, 93.]
‡ Basic catalysis research today is directed mostly towards oxidation – the oil industry's interest, as their basic starting materials are reduced carbon. Furthermore, it focuses on homogeneous catalysis, which provides only a small fraction of today's industrial catalysts.

The 21st century engineering graduate

http://dx.doi.org/10.1016/S1369-7021(08)70002-6

ScienceDirect - Materials Today The 21st century engineering graduate

Volume 11, Issue 3, March 2008, Page 6

Comment

The 21st century engineering graduate

Ruth Grahama, , EnVision Project Director
aImperial College London, UK

Available online 15 February 2008.



There is a growing appreciation in the academic community of the need for change in engineering undergraduate education.






Move to change the way we educate engineers comes in response to a number of factors, such as growing demands from industry for graduates with a broader skill base, increasing international competition in engineering education, and the motivations and experiences of prospective students.

The EnVision project at Imperial College London, UK, aims to transform undergraduate engineering education – who we teach, what we teach, and how we teach. This is clearly an ambitious goal – it is a huge challenge to transform undergraduate education and related support activities across nine engineering departments of 3000 undergraduates.

The initial phase of the project, following its establishment in March 2005, required the answering of a simple question: “What, if anything, do we need to change?” Surveys were conducted of over 2500 of our stakeholders – students, alumni, academics, industry, and professional bodies – alongside a study of the educational developments at the best institutions for engineering education across the world. These studies revealed a remarkable degree of consensus over what changes were required, particularly in the description of the ‘ideal’ engineering graduate, as well as a number of more unexpected findings. I was particularly interested in the extent to which students, graduates, and industry employers saw sustainability – the ability to take a leading role in designing solutions to local, national, and global challenges affecting society – as an important theme in the skill set of the engineering graduate of the future.

After drawing together and assessing this information, the themes of EnVision were identified:

1. Improve and sustain our ability to recruit the most able students;

2. Improve the motivation and engineering aspirations of undergraduates;

3. Ensure that our graduates possess the skills, knowledge, and attitudes necessary to become internal leaders in engineering in both industry and academia;

4. Enhance the faculty's educational provision, through the transformation and development of both curricular and noncurricular activities;

5. Significantly improve the physical learning environment and facilities in the faculty; and

6. Improve the environment for support, reward, and celebration of excellent teaching.

We have since been working with students, academics, and other key groups to determine exactly how these changes should be designed and implemented. With the project now in its implementation phase, a large number of activities are already underway.

One critical element of the project is to effect a cultural change by which teaching excellence is promoted, rewarded, and celebrated. Last June, Imperial's Faculty of Engineering presented the Inaugural Awards for Teaching Excellence in Engineering Education. The awards recognize and reward individuals and small teams renowned for the excellence of their teaching. Three awards were presented, each of value at €13 500, and this has now been established as an annual event.

Another strong theme in EnVision is the establishment and support of undergraduate projects that broaden the personal and professional leadership and communications skills of students. The projects encourage them to apply their theoretical knowledge to complex real-life engineering situations. This helps to motivate and inspire them to a career of lifelong learning in engineering. One example is Racing Green, which brings students together from across the engineering faculty to design, build, and race a zero-emission electric hybrid fuel cell racing cart. Students from six engineering departments are working together, using cutting-edge technology to develop a race vehicle ready for its first time trial this year.

A major driving force of EnVision is to shift the focus of engineering education towards ‘learning by doing’. This allows students to deepen their theoretical understanding and develop their professional skills by applying their engineering knowledge in real-world situations. To reflect this shift in the way engineering is taught at Imperial, we are also developing new teaching and learning spaces, which symbolize and inspire the imagination and creativity evident in the very best achievements of engineering technology and practice.

External communications and network building are also key elements of the project. In September 2007, we organized a high-level strategy forum to explore the issue of how to equip the engineering graduate of the future to take a leading role in designing solutions to local, national, and global challenges affecting society and the world around us. This event responded directly to calls from many of our stakeholders for an increased acknowledgment of sustainability in undergraduate engineering education. It brought together decision makers from the academic community, government, professional bodies, nongovernmental organizations (NGOs), industry, and student groups and was a resounding success.

Although the Envision project is still in an early stage of implementation, it already seems to be making a significant impact. I can see that the positive changes we are starting to make at Imperial through this program are producing some significant waves, and I believe we have a real opportunity to set a new benchmark for engineering education worldwide.

Duplicity in publishing papers

http://dx.doi.org/10.1016/S1369-7021(08)70001-4

ScienceDirect - Materials Today Duplicity in publishing papers

Volume 11, Issue 3, March 2008, Page 1

Editorial

Duplicity in publishing papers

Jonathan Wood, Editor, Materials Today

Available online 15 February 2008.



Tools that compare papers for similarities could reduce plagiarism and duplicate publications.






How big a problem is plagiarism? We might suspect that a fair amount of dubious publishing practices go on, such as publishing the same results more than once, submitting a paper to more than one journal, or reproducing others' work without acknowledgement. It seems to have become more prevalent as pressures to publish increase.

But how do we begin to identify incidences of plagiarism? It is surely not enough to go by anecdotal reports, or rely on researchers to highlight cases as they come across them by chance in the literature. Well, it's probably not that difficult. Computer-based tools that search documents for similarities have been around for a while and are widely used to identify cheating in school and college exams. Now Harold Garner and Mounir Errami at the University of Texas Southwestern Medical Center have developed a computer code that checks multiple documents for duplication of key words and compares word order and proximity. Applying their eTBLAST program to >7 million abstracts in Medline, 70 000 papers came back as being highly similar [Nature (2008) 451, 397].

Now there are a number of circumstances where significant duplication is to be expected and is wholly ethical. These include updates to clinical trials, conference papers, and corrections to papers. But Garner and Errami are determined in their pursuit of unethical practices. They have placed the 70 000 potential duplicate papers on a publically available database, Déjà vu (http://spore.swmed.edu/dejavu), and have begun to check them manually (quite a task!). Already, one suspected case of plagiarism has reportedly resulted in an investigation by a journal. “We can identify near-duplicate publications using our search engine,” says Garner. “But neither the computer nor we can make judgment calls as to whether an article is plagiarized or otherwise unethical. That task must be left to human reviewers, such as university ethics committees and journal editors.”

That does point to one problem: whose responsibility is it to chase cases of unethical practice, journals or universities? The simplicity of instigating such computational tools probably points toward the journals in identifying suspect papers (and many publishers are jointly investigating the best way to proceed), but subsequent steps are more open to debate.

I am convinced that it would be a very worthwhile step. Caution needs to be applied with suspect papers and common standards about how much new work is needed for a new paper need to be agreed upon, but clear-cut cases should be discovered in this way. But it does lead me to wonder what else is possible? Can figures be checked for manipulation or duplication? This is where scientific fraud and misrepresentation of data most often happen.

The top ten advances in materials science

http://www.materialstoday.com/archive/2008/11-01/top10.html

The top ten advances in materials science

What are the defining discoveries, moments of inspiration, or shifts in understanding that have shaped the dynamic field of materials science we know today? Here’s what we think are the most significant.

December 19, 2007

Jonathan Wood
Editor, Materials Today
E-mail: j.wood@elsevier.com

Read this article in pdf format

The ending of one year and the beginning of the next is a strange time. It is very human to mark the passing of time, remembering what has been done before looking forward to what’s to come. As the New Year arrives, whether you prefer peaceful reflection or joyous celebration, awards or resolutions, one thing is clear from any newspaper or magazine. It is, above all, a time to draw up lists. Who are we to disagree?

We’ve assembled a list of the top ten advances in materials science over the last 50 years. We thought long and hard. We sought the advice of our editorial advisory panel and asked leaders in the field to add their own contributions. We hope the results are interesting and thought-provoking.

In making the final selection, we have tried to focus on the advances that have either changed our lives or are in the process of changing them. This is arguable, of course. Should an advance alter all our daily lives, or does fundamentally changing the research arena count? What about discoveries that can be clearly attributed to a certain date and investigator, or those developments that have come about incrementally through the efforts of many? Where does materials science stop and electronics, physics, or chemistry begin? And how do you assess the value of things like plastic bags? Undeniably they are a boon for carrying shopping but now also an item of scorn for energy and waste reasons.

Instead of ruling any of these out, we’ve tried to come up with a balanced selection. In doing so, we hope to start some debate about the discoveries that most mark out today’s materials science. Let us know what we’ve missed. If you’re incredulous that organic electronics or high-temperature superconductors aren’t in the top ten, tell us why. Should Kevlar, Post-it notes, float glass, or F1 racing tires be in the list? What will define the next 50 years of materials science?

If you believe materials scientists are unsung heroes, that our work goes unnoticed and unheralded, here is your ammunition. With our time limit of 50 years, the list is of immediate relevance. It is about how materials science is affecting our world today, now.

1. International Technology Roadmap for Semiconductors


Semiconductor research is guided by the ITRS. (Courtesy of SEMATECH.)
OK, so it’s not a research discovery, solely a way of organizing research priorities and planning R&D. But the International Technology Roadmap for Semiconductors (ITRS) is a remarkable achievement (see The history of the ITRS). It sets out goals for innovation, technology needs, and measures for progress that all can sign up to in the fiercely competitive microelectronics industry.

A mixture of science, technology, and economics, it’s hard to see how the ITRS could do better in driving forward advances in this area, whether it’s in materials, characterization, fabrication, or device design. And it is an appropriate first choice in this list. Not only is electronics absolutely critical to our modern world, progress in semiconductor processing and advances in materials science have gone hand-in-hand for the last 50 years.

Let’s just hope the International Panel on Climate Change enjoys similar success in driving innovation and reaching agreed goals.

2. Scanning probe microscopes

The invention of the scanning tunneling microscope (STM) by Heinrich Rohrer and Gerd Binnig at IBM’s Zurich Research Laboratory was deservedly awarded the Nobel Prize for Physics in 1986.


Rohrer (left) and Binnig (right) with a first-generation scanning tunneling microscope. (Courtesy of IBM Zurich Research Laboratory.)
Not only is this a new microscopy technique – remarkable enough in itself – but it provides a way to probe the local properties of a sample directly with nanometer resolution. Quickly followed by the atomic force microscope (AFM), this new access to the nanoscale world (see Making sense of the nanoworld), arguably brought about the current ubiquity of nanotechnology. The invention immeasurably increased our abilities at this scale.

3. Giant magnetoresistive effect

The 2007 Nobel Prize for Physics went jointly to Albert Fert of Université Paris-Sud, France, and Peter Grünberg of Forschungszentrum Jülich, Germany, for independently discovering the giant magnetoresistance (GMR) effect in 1988. So it is no surprise to see this advance on our list.

GMR describes the large change in electrical resistance seen in stacked layers of magnetic and nonmagnetic materials when an external magnetic field is altered. Thanks largely to the subsequent work of Stuart Parkin and coworkers at IBM Research, the phenomenon has been put to great effect in the read heads in hard disk drives. These devices are able to read out the information stored magnetically on a hard disk through changes in electrical current.

The high sensitivity of GMR read heads to tiny magnetic fields means that the magnetic bits on the hard disk can be greatly reduced in size. The phenomenal expansion in our ability to store data that we continue to witness today can be traced back to this discovery.

4. Semiconductor lasers and LEDs

The development of semiconductor lasers and light-emitting diodes (LEDs) in 1962 is a great materials science story (seeThe III-V laser and LED after 45 years). They are now the basis of telecommunications, CD and DVD players, laser printers, barcode readers, you name it. The advent of solid-state lighting is also likely to make a significant contribution to reducing our energy usage.

5. National Nanotechnology Initiative

Bill Clinton gets some of the credit for the fifth materials science development on our list. He was the US president who announced the establishment of the National Nanotechnology Initiative (NNI) in 2000, a US federal, multi-agency research program in nanoscale science and technology.

The NNI has had an immense impact. It cemented the importance and promise of a nascent, emerging field, establishing it immediately as the most exciting area in the whole of the physical sciences. Nanotechnology simultaneously gained an identity, a vision, and a remarkable level of funding through the initiative. It also established a method of funding interdisciplinary science in such a way that the rest of the world would have to try to match.

Mihail C. Roco of the National Science Foundation was one of those who was involved in the initial NNI vision setting and national organizational efforts. “During 1997 to 1999, I worked with an initially small group including Stan Williams, Paul Alivisatos, James Murday, Dick Siegel, and Evelyn Hu,” recalls Roco. “We envisioned a ‘new industrial revolution’ powered by systematic control of matter at the nanoscale. With this vision, we built a national coalition involving academia, industry, and a group of agencies that become the nucleus of the NNI, launched in 2000.”

The NNI now involves 26 independent agencies and has an estimated budget of ~$1.5 billion in 2008. It has been the largest single investor in nanotechnology research in the world, providing over $7 billion in the last seven years. Now 65 countries have national research focus projects on nanotechnology, while industry nanotechnology R&D has exceeded that of governments worldwide. The global nano-related R&D budget was in excess of $12 billion in 2007.

On behalf of the interagency group, Roco proposed the NNI on March 11, 1999 at the White House Office of Science and Technology Policy (OSTP). The fear of many was that there was little chance of nanotechnology becoming a national priority program. Surely it would be perceived as being of interest just to a small group of researchers? Instead, by defining nanotechnology as a broad platform for scientific advancement, education, medicine, and the economy, the NNI was approved with a budget of $489 million in 2001. “The NNI was prepared with the same rigor as a science project,” says Roco.

6. Carbon fiber reinforced plastics


Carbon fiber-reinforced plastics were at the heart of this bike built by Lotus Engineering for the 1992 Barcelona Olympics. It helped Chris Boardman win gold. (Courtesy of Lotus.)
The last 50 years have seen advanced composites take off – quite literally, in that many applications of these light but strong materials have been in aviation and aerospace. But modern composite materials have touched just about all industries, including transport, packaging, civil engineering, and sport. They can be found in Formula 1 cars, armor, and wind turbine rotor blades.

Leading the charge are carbon fiber reinforced plastics or, more properly, continous carbon fiber organic-matrix composites. These materials bond extremely stiff, high-strength carbon fibers into a polymer matrix to give a combined material that is also exceptionally tough and light in weight.

The early 1960s saw the development of carbon fibers produced from rayon, polyacrylonitrile, and pitch-based precursors. The long, oriented aromatic molecular chains give the fibers exceptional strength and stiffness. This was a real gain over the amorphous glass fibers used previously in composite materials.

The development of carbon fibers, together with advances in design, modeling, and manufacturing, has given rise to composite materials with controlled, specific properties. “Rather than an engineer using a constant set of material characteristics, organic-matrix composites and the associated manufacturing methodology now enables the engineer to design the material for a specific application,” says Richard A. Vaia of the Air Force Research Laboratory. “The manufacturing science has opened up new frontiers, effectively moving component design down to materials design.” The spectacular gain in performance has seen the increasing use of these materials despite the cost and increased difficulty in design, shaping, and recycling, such that the new Boeing 787 uses composites extensively in its wings and fuselage.

7. Materials for Li ion batteries

It is hard to remember how we coped before laptops and cellular phones came along. This revolution would not have been possible without a transition from rechargeable batteries using aqueous electrolytes, where H+ is the working ion, to the much higher energy densities of Li ion batteries.

Li ion batteries required the development of novel electrode materials that satisfy a number of considerations. In particular, the cathode needs a lightweight framework structure with free volume in between to allow a large amount of Li ions to be inserted and extracted reversibly with high mobility.

The process of materials design and discovery involved a mixture of clever chemical and electrochemical intuition, rational assessment of the technical requirements, and substantial experimental effort, and is dominated by the work of John B. Goodenough and colleagues at the University of Oxford in the 1980s. They came up with the cathode material LiCoO2 that Sony combined with a carbon anode in 1991 to give us the batteries that make possible the portable devices we know today. Work continues to develop cathode materials without the toxic Co and with three-dimensional framework structures like LiFeO4 for environmentally benign, high-energy density batteries.

8. Carbon nanotubes


Viewgraph showing a single- or double-walled CNT published in 1976. (Reprinted with permission from Oberlin, A., et al., J. Cryst. Growth (1976) 32, 335. © 1976 Elsevier.)
Although a discovery normally attributed to Sumio Iijima of NEC, Japan in 1991, the observation of nanotubes of carbon had actually been made on previous occasions (see A journey on the nanotube). However, following on from the excitement of the discovery of C60 buckyballs in 1985 – a new form of carbon – Iijima’s observations of new fullerene tubes aroused great interest immediately.


Today, the remarkable, unique, and phenomenally promising properties of these nanoscale carbon structures have placed them right among the hottest topics of materials science. So why are they only at number eight in this list? Well, there still remains much to sort out in their synthesis, purification, large-scale production, and assembly into devices. And there’s also the very frustrating inability to manufacture uniform samples of nanotubes with the same properties.

9. Soft lithography

The ability to fabricate functional structures and working devices in different materials is central to the production of microelectronic devices, data-storage systems, and many other products. This process is almost exclusively carried out by highly specialized, complex, and very expensive photolithography equipment confined to the controlled environments of cleanrooms. How valuable, then, is the introduction of an alternative?

Soft lithography makes use of the simple, ancient concept of using a stamp to produce patterns again and again. It can be used on many different substrates, be they flat, curved, or flexible. What’s more, soft lithography is cheap, offers nanoscale resolution, and can be applied to new areas in biotechnology and medicine.

The initial technique of microcontact printing (µCP) was developed in 1993 at the lab of George Whitesides at Harvard University. “Microcontact printing has revolutionized many aspects of materials research,” says Byron Gates of Simon Fraser University, Canada. “Molecules are transferred to a substrate using an elastomeric stamp. This poly(dimethylsiloxane) or PDMS stamp conforms to the substrate, unlike hard masks used in previous lithography techniques.” In this way, molecules can be printed over large areas in well-defined patterns with features just 30 nm in size. As well as the transfer of small organic molecules, µCP has been adapted to print solid materials directly, extending its capabilities into nanofabrication. Since 1993, µCP has expanded into a suite of printing, molding, and embossing methods known as soft lithography. All of them use an elastomeric stamp to reproduce a pattern from a master template over and over again.

“All these techniques share one thing: the use of organic materials and polymers – ‘soft matter’ in the language of physicists,” says Younan Xia of the University of Washington in St. Louis. “Soft lithography offers an attractive route to microscale structures and systems needed for applications in biotechnology, and most of them exceed the traditional scope defined by classic photolithography.”

10. Metamaterials


The metamaterial structure of an invisibility cloak that hides objects from microwave radiation. (Credit: David Schurig, Duke University.)
The beginning of the new millennium brought great excitement when it was conclusively demonstrated that a material with a negative refractive index could exist. Light, or at least microwaves, would bend the ‘wrong way’ on entering this material, according to a standard understanding of Snell’s law of refraction. This ended a long-standing argument over Veselago’s prediction in the 1960s that materials simultaneously having a negative permeability and a negative permittivity would have a negative refractive index. At the same time, it opened up a perplexing new optical world full of counterintuitive results that can be explained using 19th century classical electromagnetism.

But the surprising optical properties don’t arise from the material’s composition as its structure. The first metamaterial was a composite of metal wires and split rings assembled on a lattice of printed circuit boards. It was an example of a metamaterial – an artificial structure of repeated micro-sized elements designed for specific properties.

“Metamaterials derive their properties as much from their internal structure as from their chemical composition,” explains John Pendry of Imperial College London, UK. “Adding structure to chemistry as an ingredient greatly increases the range of properties that we can access. There is a new realization that metamaterials can give access to properties not found in nature.”

Crucially, if the structure of the material is much smaller than the light’s wavelength, then an overall permittivity and permeability of the material can still be used with Maxwell’s equations to describe the electric and magnetic response of the material. Thin wire structures can generate a negative electrical response at gigahertz frequencies, while split-ring structures generate a negative magnetic response. These structures were combined for the first time in 2000 by David Smith, Willie Padilla, and Shelly Schultz at the University of California, San Diego to make a negatively refracting material. “Now many people are going through a process of feverish invention as new possibilities are explored, pushing the concept up in frequency towards the visible and also downwards, even to create novel dc responses,” says Pendry.

“Theorists too have been inspired,” adds Pendry, who pointed out that a negative refractive index could be used to construct a ‘perfect lens’. Such lenses would have a resolution unlimited by fundamental physics of the design, and only limited by quality of manufacture. “A new approach to subwavelength imaging now rides on the back of the metamaterial concept,” he says. Several suggestions for invisibility cloaks to hide objects from electromagnetic radiation have also been made. All of these proposals imply the use of metamaterials to realize their designs.

“The first applications [of metamaterials] will be simple improvements of existing products,” Pendry expects. “For example, lightweight lenses for radar waves have been manufactured using metamaterials. Then entirely novel applications will follow, probably developed by the research students of today’s metamaterials researchers.”

The history of the ITRS
The ITRS provides a guideline for research and development for integrated circuit technology needs within a 15-year horizon. Updated annually, the ITRS evolved from a series of workshops and assessments conducted by industry leaders in the late 1980s to ascertain precompetitive critical needs. The first national technology roadmap efforts began in 1992 and in 1993 the first Semiconductor Technology Roadmap effort was sponsored by the Semiconductor Industry Association, supported by the Semiconductor Research Corporation, and edited and produced by SEMATECH.

In 1994, the roadmap was updated by a team of over 400 technologists and renamed the National Technology Roadmap for Semiconductors (NTRS). In 1997, the NTRS began to emphasize the challenges, technology requirements, and potential solutions for each roadmap topic. The NTRS was reviewed for the first time in 1998 by an international team that included technologists from Europe, Japan, Korea, and Taiwan. The first ITRS was produced in 1999, the first ever international industry roadmap of its kind.

The ITRS is based on the consensus of a substantial team. More than 1200 participants were involved from industry, national laboratories, and academia in 2005 and 2006. As the manufacturing of semiconductors becomes more challenging, the ITRS teams are expanding the role of roadmapping into new topics with the potential of guiding the industry beyond complementary metal-oxide-semiconductor systems. The new 2007 edition will have 18 chapters and over 1000 pages, it is estimated.

Linda Wilson, ITRS managing editor, SEMATECH, and Alain Diebold, College of Nanoscale Science and Engineering, University at Albany, State University of New York



Making sense of the nanoworld
The fabrication of the first STM in March 1981 in IBM’s Zurich Research Laboratory made it possible for the first time to produce real-space images of electrically conductive surfaces with subnanometer spatial resolution. The development of the AFM in 1986 at IBM Almaden Research Center and Stanford University permitted explorations to be extended to electrically insulating and biological materials.

These two inventions have opened doors into the nanoscale world, and ultimately to nanotechnology. Looking at individual nano-entities such as single molecules, how they react to an external stimuli, how they move and dance on a surface, and how they recognize and talk to each other is no longer science fiction. Moreover, these nanotools allow the manipulation of individual nano-objects and enable scientists to gain a quantitative insight into their physical and chemical properties. Thus they have become crucial in optimizing the performance of nanodevices.

The ultimate impact of these tools will surely cover a huge range of disciplines, including materials science, (opto)electronics, medicine, catalysis, and they will offer new solutions to key problems such as energy and the environment.

In the end, SPM techniques are all about the five senses. Sight is achieved by gently touching surfaces. Hearing: the acoustic response of the tip allows detailed insights into the mechanical properties of surfaces. The same tips, once functionalized with well-defined groups, can identify functional groups through molecular recognition, thus they can finally smell and taste the new and thrilling perfume and flavor of the nanoworld.

Paolo Samorì, ISIS-ULP/CNRS, Strasbourg, France and ISOF-CNR, Bologna, Italy



The III-V laser and LED after 45 years
A significant fraction of the Earth’s population has, by now, seen an LED. But few are aware it is not a conventional light source, rather an electronic source related to the transistor.

As John Bardeen’s (one of the inventors of the transistor) first student and then colleague for 40 years, I heard him explain many times that it was not known until the transistor that a current could create a nonequilibrium electron-hole population in a semiconductor. Subsequently, electron-hole recombination could re-establish equilibrium, delivering light.

As we studied recombination for transistor reasons, we were on the path to the laser and LED, especially when we moved to the direct-gap III-V compounds. Studying GaAs for tunnel diodes in 1960–62, I was not happy with its 1.4 eV (infrared) bandgap. I learned how to shift GaAs towards GaP, to GaAs1–xPx and red light wavelengths.

In 1962, a small number of us realized that the GaAs p-n junction might serve as the basis of a laser. But I wanted to work not in the infrared, but with GaAs1–xPx in the visible region where the eye sees. I knew enough about lasers to know I needed a cavity to help my red p-n junctions become lasers.

My astute colleague at General Electric (GE), Bob Hall, was one step ahead of me. He made GaAs diodes with Fabry-Perot resonator edges, with the crystal itself the cavity – very clever! He preferred polishing to make his diode cavities and I preferred cleaving (not so easy).

Then, one early fall day, Hall’s boss called me to tell me that Hall was running a laser, and would I please give up cleaving! I devised at once a simple method to polish my diode Fabry-Perot cavities, and immediately had red III-V alloy lasers and LEDs.

With Hall’s infrared GaAs lasers and incoherent emitters and my visible, red GaAs1–xPx lasers and LEDs, GE announced the availability of these devices for sale late in 1962. The red LED was practical from the beginning, and only got better and cheaper over time.

Now, after 45 years of work by many people, the high-brightness, high-performance LED promises to take over lighting. The scale and variety of what is happening is surprising, totally unbelievable. Since we are talking about an ‘ultimate lamp’, this work won’t stop, will only grow and, of necessity, become cheaper. This will make the universal use of the LED possible – appearing everywhere in lighting and decorating!

Nick Holonyak, Jr., University of Illinois at Urbana-Champaign



A journey on the nanotube
Sumio Iijima reported the observation of multiwalled carbon nanotubes (CNTs) in 1991 [Nature 354, 56]. Then in 1993, two independent groups, Iijima and Ichihashi [Nature 363, 603] and Bethune et al. [Nature 363, 605] reported the growth of single-walled CNTs in the same issue of Nature.

The impact of these papers on the scientific community has been tremendous, perhaps leading to the birth of nanoscience and nanotechnology.

However, the first direct observation of multiwalled CNTs was recorded in 1952 by Radushkevich and Lukyanovich [Zurn. Fisic. Chim. (1952) 26, 88], while an image of a single- or possibly double-walled CNT was published in 1976 by Oberlin et al. [J. Cryst. Growth (1976) 32, 335].

Aside from the controversy surrounding their discovery, the tremendous mechanical, electrical, and thermal properties of CNTs combined with a low density promise to revolutionize materials science.

Applications are appearing in integrated nanoelectromechanical systems working in the gigahertz frequency band, exquisitely sensitive mechanical sensors, ultrasharp scanning probe microscopy tips, nanosized drug delivery vehicles, and so on. Moreover, using CNTs as fiber reinforcements could lead to innovative new composite materials.

Even if miniaturization tends to be the focus for CNTs, in mechanics there is also the opposite trend because the human scale is the meter. CNTs are strong and stiff mainly because they are small and thus nearly defect-free – their best attribute. Thus, controlling and minimizing defects while scaling up CNT structures would be a real breakthrough.

For example, a macroscopic cable having the same strength-to-density ratio as a single, defect-free nanoscopic CNT would allow us to build fantastic structures such as a terrestrial space elevator. Here, a cable attached to the planet’s surface could carry payloads into space.

Alternatively, if CNT materials that mimic the hairs on the feet of spiders and geckos could be scaled up, a Spiderman suit for clinging to walls would be within the reach of all of us. There is also plenty of room at the top.

Nicola Pugno, Politecnico di Torino, Italy