AVerticalKelvinTestStructureforMeasuringtheTrueSpecificContactResistivity

IEEE ELECTRON DEVICE LETTERS, VOL, EDL-7, NO. 4, APRIL 1986

电气与电子工程师协会电子设备投稿栏&＃xff0c;第四卷&＃xff0c;电子器件7&＃xff0c;1986年4月

A Vertical Kelvin Test Structure for Measuring the True Specific Contact Resistivity

一个测量真实的特定的接触电阻率的垂直温度测试结构

TAN FU LEI, MEMBER, IEEE, LEN-YI LEU, AND CHUNG LEN LEE

谭弗雷&＃xff0c;成员&＃xff0c;IEEE,

Abstract-A vertical kelvin test structure, which can be used to measure the true specific contact resistivity of a metallization system, is proposed and studied. For this test structure, the driving current flows"vertically," thus the sheet resistance and current crowding effects are eliminated and measurement on the ture specific contact resistivity becomes possible. Experimental works show that this test structure gives a more linear relation between resistance and contact area than the conventional six-terminal test structure.

摘要-垂直温度测试结构能够被用来测量真实的特定的金属化系统的接触电阻率&＃xff0c;垂直温度测试结构被计划和学习。对于这个测试结构来说&＃xff0c;驱动电流垂直地流过&＃xff0c;从而片状的阻力和电流催逼效应被消除和在真实的特定的可接触电阻变得可能。实验性的工作展示了在电阻和接触面之间这个测试结构比传统的6-终端测试结构给出了更多的线性关系。

In this letter, a "vertical" kelvin test structure to measure the "true" specific contact resistivity is presented. For this test structure, the driving current flows "vertically" from the metal contact pad toward the contacted substrate. This eliminates the current crowding effects which are inherent in the horizontal type of test structures, and makes the determination of the "true" specific contact resistivity possible. This test structure can be incorporated with the six-terminal test structure [4] to compare the measured results.

在这个栏目里&＃xff0c;测量“真实的”特定的接触电阻率的“垂直的”温度测试结构被推出。对于这个测试结构&＃xff0c;驱动电流垂直地从金属制的接触衬垫流到接触基片。消除在这个测试结构的卧式中的电流聚集效应&＃xff0c;造就了真实的特定的接触电阻可能性。这个测试结构能够和6-终端测试结构[4]合并来比较测量结果。

II. TEST STUCTURE

    The cross section of the test structure is shown in Fig. 1(a) along with its top view in Fig. 1(b). The driving current I is forced from pad 1 toward the substrate and the voltage V is sensed between pads 2 and 3 along the implanted bar. The vertical current flow is restricted by an isolation p-n junction. For this structure, it can be seen that the current flows vertically and only through the contact window, hence the current distribution in the contact resion is uniform provided that the contact region is metallurgically uniform. The contact resistance Rc is thus directly measured with the value of V/I, where Ac is the contact area through which the current passes. No lateral current crowding effect, which is mentioned in [6] and [7], is expected. Also, the sheet resistance effect on determining the value of pc become minimum for this test structure since it is not involved at all in determining pc.

2. 测试结构

    这个测试结构的横截面和它的顶视图图1(b)被展示在图1(a)。驱动电流I强迫衬垫1趋向基片&＃xff0c;电压V被产生在沿着嵌入阀的衬垫2和3之间。垂直电流被一个绝缘PN结限制。对于这个测试结构&＃xff0c;它能被看到电流仅仅通过接触窗口垂直地流过&＃xff0c;因此如果这个接触区是金属制统一的在接触区的电流分配是统一的。接触电阻Rc可以使用V/I的值直接测量&＃xff0c;Ac是电流通过的接触面。没有侧面电流聚集效应&＃xff0c;侧面电流聚集效应将在第六和第七部分提及。同样的&＃xff0c;虽然表面电阻不能对pc产生根本影响&＃xff0c;对于这个测试结构来说表面电阻对pc流变成最小值有一定影响。

Fig. 1. (a) The cross-sectional view of the proposed vertical Kelvin test
structure for measuring the true specific contact resistivity. (b) The top view
of the structure. In the figure, the horizontal structure for the six-terminal
measurement is also incorporated.

图1&＃xff08;a&＃xff09;测量真实特定接触电阻率的垂直温度测试结构的横截面图

图1&＃xff08;b&＃xff09;测量真实特定接触电阻率的垂直温度测试结构的顶视图。在这个图里&＃xff0c;6-终端测量的水平结构也被加入进来

    Manuscript received September 18, 1985; revised December 2, 1985. This work was supported by Electronic Research and Service Organization (ERSO), ITRI of Taiwan, Republic of China.

    The authors are with the Institute of Electronics, National Chiao Tung University, Hsin Chu, Taiwan, Republic of China.

    IEEE Log Nmber 8608010.

手稿在1985年9月18日接收&＃xff0c;1985年12月2日校订。这项工作被电子开发和服务组织&＃xff08;ERSO&＃xff09;&＃xff0c;台湾的ITRI&＃xff0c;中华人民共和国所支持。

作者包括国民交通大学电子学院&＃xff0c;Hisn Chu&＃xff0c;台湾&＃xff0c;中华人民共和国

国际电子与电气工程师协会历史记录数字 8608010.

    In Fig.1, the six-terminal structure [4] is also incorporated and this has two merits. First, the voltage V can also be sensed between pad pairs 2-4, 2-5, and 2-6. The obtained values can be averaged with that of the pad pair 2-3 to reduce the error introduced by the misalignment between the isolation junction window and the contact window. Second, six-terminal measurement can be performed and the result can be compared with that obtained from the vertical structure.

在图1中&＃xff0c;6-终端结构&＃xff08;图标4&＃xff09;也被合并进来&＃xff0c;这样做有两个优点。第一&＃xff0c;电压V也能在衬垫对2-4、2-5、2-6之间产生。所获得的值能够被平均防止因为孤立界面和接触界面之间的未对准引入的错误。第二&＃xff0c;6-终端测量能够被执行&＃xff0c;结果能够与来自于垂直结构的值做比较。

III.  EXPERIMENTS AND RESULTS

Test structures of Fig.1 have been fabricated to compare the results obtained by the vertical structure measrement and the six-terminal measurement, respectively. Different contact systems of either A1(1-percent Si)/Si or A1(1-

((0741-3106/86/0400-0259$01.00  © 1986 IEEE

 percent Si)/TiSi2/Si were formed on n-type substrates of the resistivity of 3~5 Ω·cm. Contact regions of various dimensions (5X5 µm, 5X10 µm, 10X10 µm, 10X15 µm, 10X20 µm, 15X15 µm, 15X20 µm, 20X20 µm) were used. The alignment tolerance of the contact was 10 µm. The n&＃43; implanted bars and the p isolation junctions were formed by ion implantation and the junctions were kept as shallow as possible. The depths of n&＃43; junctions were 0.33-0.5 µm for 2 x 10(15)/cm2 to 6 x10(15)/cm2 ion doses of As&＃43; and the junction depths of p isolations were 0.7 pm for 2 x 10(13)/cm(2) ion dose of BF2&＃43;. The backsides of wafer were n&＃43;  diffused and A1 metallized. For each wafer, at least 200 test patterns were measured.

3. 实验和结果

图1的测试结构已经被制造来比较通过垂直结构测量和6-终端测量的结果。分别的&＃xff0c;不同的接触系统1%硅/硅或者1%硅/Ti硅/硅在电阻率为3~5欧姆/厘米的N型基片上加工成型。各种各样规模&＃xff08;5X5 纳米&＃xff0c;5X10 纳米&＃xff0c; 10X10纳米&＃xff0c; 10X15纳米&＃xff0c; 10X20 纳米&＃xff0c; 10X20 纳米&＃xff0c; 15X15 纳米&＃xff0c; 15X20 纳米&＃xff0c; 20X20 纳米&＃xff09;的接触面被使用。接触的定位公差是10纳米。n&＃43; 嵌入栅和p沟道通过离子嵌入和间隔被加工成型看起来是可能的和浅的。n&＃43; 沟道对于As&＃43;的2 x 10的15次方/平方厘米到6 x10的15次方/平方厘米的离子量的深度是0.33到0.5纳米&＃xff0c;对于BF2&＃43;的2 x 10的13次方/平方厘米离子量的p绝缘栅的深度为0.7纳米。晶片的背面是n&＃43; 扩散和A1金属化。对于每个晶片&＃xff0c;至少200个测试模式被执行。

For all the test structures fabricated, the I-V characteristics were linear at the current levels of -5 to 0.5 mA. Contact resistances, measured in the linear region, of various contact areas are shown in Figs.2. and 3 for A1(1-percent Si)/Si and A1(1-percent Si)/TiSi2/Si contact systems, respectively, where Rcv were measured from the vertical structure and Rcs were measured from the six-terminal method. In these two figures, for two sets of measured points, two straight lines obtained by the least-square fitting are drawn. In Fig.2, the slope of the Rcv straight line is -0.76. Forboth figures, the former are closer to the ideal value of -1. In these figures, it is also seen that Rcv values are always smaller than Rcs values and the differences become larger for larger contact areas. This is expected since the lateral current crowding and the sheet resistance effects (Rs &＃61; 22.67 Ω/cm and 46.5 Ω/cm for the case of Figs. 2 and 3, respectively) become serious in determining Rcs for the larger contact window if the alignment tolerance of the contacts is kept constant. This phenornenon had also been predicted by [6].

对所有测试结构制造&＃xff0c;在电流等级为-5到0.5mA时I-V特征是线性的。接触电阻&＃xff0c;在线性关系测量&＃xff0c;对于A1&＃xff08;1%硅&＃xff09;/硅和A1&＃xff08;1%硅&＃xff09;/Ti硅2/硅接触系统来说各种各样的接触面积被展示在图2和图3&＃xff0c;分别的&＃xff0c;Rcv被垂直系统测量&＃xff0c;Rcs被6-终端系统测量。在这两个图里面&＃xff0c;对于两个测试点的设置&＃xff0c;两条与最小平方一致的直线被画。在图2中&＃xff0c;Rcv直线的斜率是-0.76。对于这两个图来说&＃xff0c;过去比较接近理想值-1。在这些图里面&＃xff0c;看起来Rcv的值看起来一直比Rcs的值小&＃xff0c;对于比较大的接触面积差异变大。对于比较大的接触窗口如果接触的直线斜率保持不变时决定Rcs的值因为侧面电流聚集效应和电阻切面影响&＃xff08;分别得对于图2和图3的案例&＃xff0c;Rs &＃61; 22.67 Ω/cm and 46.5 Ω/cm&＃xff09;变的严重起来。这个结果已经在[6]节被预言。

    The apparent specific contact resistivities of contact systems were also measured with the vertical structure (

ρcv ) and the six-terminal method (ρcs ) for a constant contact area. Fig.4 is a plot of ρcv versus ρcs with the contact area of 20X20 µm. The various contact resistivities were obtained by implanting various doses of As&＃43; onto the contact window. It can be seen that ρcs /ρcv deviates more from 1 for smaller specific contact resistivities. Since it has been predicted that ρcs deviates more from the true value of ρc as ρc becomes smaller [7], this indicates that ρcv does give the closer value to the true specific contact resistivity.

In Fig.3, the Rcv line does not exhibit the ideal square law, i.e., the slope of -1. It is believed to be mainly caused by the nonuniformity of TiSi2 formation at the contact interface. (The microscopy observation on the contact areas after stripping the cntact metal confirmed this.) Besides the interface nonuni-formity, for this test structure, the errors mainly come from the misalignment of the first mask to form the isolation junction and the third mask to open the contact window, and the lateral diffusion of the isolation junction. These two effects will make the current flow not strictly vertical at the periphery of the contact region. However, the former error can be reduced by averaging the four Rcv values obtained by sensing on pad pairs of 2-3,2-4,2-5, and 2-6, respectively, and the latter error can be minimized by taking into account the lateral diffusion during the mask design. From the experimental results in Section III, the vertical test structure did give a more linear relation between contact resistance and contact area than the six-terminal test structure. Hence, it can be concluded that the vertical Kelvin test structure offers a potential method to measure the true contact resistance of contact sstems.

4. 讨论和结论

在图3里&＃xff0c;Rcv线没有展示出理想平方定律&＃xff0c;例如&＃xff0c;-1的斜率。这被确信是导致在接触窗口的TiSi2 结构不一致的主要原因。&＃xff08;在剥离金属接触确信这个后用显微镜在接触面观察&＃xff09;除了接触面不均匀&＃xff0c;对于这个测试结构来说&＃xff0c;错误主要来自于隔离交叉口的第一个膜和打开连接窗口的第三个膜未对准&＃xff0c;还有隔离交叉口的侧面扩散。这两个影响会使电流在接触面的外围非严格的垂直。之前的错误通过平均4个来自垫片对2-3,2-4,2-5,2-6的Rcv值能够被缩小&＃xff0c;最后的错误通过重视膜设计期间的侧面扩散问题。在第三部分的实验结果里&＃xff0c;垂直测试结构比6-终端测试系统给出了更多的接触电阻率和接触面积之间的线性关系。因此&＃xff0c;那是确定的&＃xff0c;垂直开尔文测试结构给出了一个潜在的方法来测量接触系统的真实的接触电阻率。

Fig. 2. The contact resistances of Al(1-percent Si)/Si are plotted versus the
contact area, where R,, were measured from the vertical structure and R,
were measured from the six-terminal structure. The straight lines are
obtained by the least-square-fitting method for two sets of measured values,
respectively.

图2&＃xff0c; A1(1%Si&＃xff09;/Si 的接触电阻率与接触面积的比&＃xff0c;Rcv 被垂直结构测量&＃xff0c;Rcs被6-终端结构测量。直线通过两个测量值的适当的最小平方方法获得。

Fig. 3. The contact resistances of Al(1-percent Si)/TiSiJSi are plotted
versus the contact area, where R,, were measured from the vertical
structure and R, were measured from the six-terminal structure. The
straight lines are obtained by the least-square-fitting method for two sets of
measured values. respectively.

图3&＃xff0c; A1(1%Si&＃xff09;/TiSi2/Si 的接触电阻率与接触面积的比&＃xff0c;Rcv 被垂直结构测量&＃xff0c;Rcs被6-终端结构测量。直线通过两个测量值的适当的最小平方方法获得。

Fig. 4. The apparent specific contact resistivities of contact systems measured from the vertical structure pe. are plotted with those measured from the six-terminal structure The various contact resistances were formed by implanting various doses of As &＃43; onto contact regions.

 图4. 来自垂直结构ρcv测量的接触系统的表面的特定的接触电阻率被绘制和来自6-终端的就诶哦股的测量结果。各种各样的接触电阻率被在接触区域注入不同剂量的As&＃43;构成。

REFERENCES

[1] H.H.Berger: "Models for contact to planer devices," Solid-State Electron., vol. 15,  pp 145-158, 1972.

[2] G. K. Reeves and H. B. Harrison, “Obtaining the specific contact
resistance from transmission line model measurements,” IEEE Electron
Device Lett., vol. EDL-3, p p ~1 11-113, 1982.
[3] S . S. Cohen, G. Gildenblat, M. Ghezzo. and D. M. Brown, “A1-0,9%
SiiSi ohmic contacts to shallow junction,” J. Electrochem. Soc., vol.
[4] S. J. Proctor, L. W . Linholm. and J. A. Mazer, *‘Direct measurements
of interfacial contact resistance. end contact resistance, and interfacial
contact layer uniformity,” IEEE Trans. Electron Devices, vol. ED-
[5] S . S . Cohen, “Contact resistance and methods for its determination,”
Thin Solid Films, vol. 104, pp. 361-379, 1983.
[6] M . Finetti, 4. Scorzoni, and G. Soncini, “Lateral crowding effects on
contact resistance measurement in four terminal resistor test patterns,”
IEEE Electron Device Lett., vol. EDL-5. pp. 524-526, 1984.
[7] W’. M. Loh, K. Saraswat, and R. W. Dutton, “.4nalysis and scaling of
Kelvin resistors for extraction of specific contact resistivity,” IEEE
Electron Device Lett., vol., EDL-6. pp. 105-108, 19P5.

对于不变的接触面积来说显然的接触系统的特定的接触电阻率也被垂直结构(ρcv )和6-终端方法(ρcs )测量。图4是一个接触面积为20X20的ρcv 和ρcs 的对比曲线图。不同的接触电阻率通过在接触窗口上植入不同量的As&＃43;来获得。对于特定的接触电阻率来说看起来ρcs /ρcv 偏离1更多。因为已经预计到ρcs与当ρc变的比较小[7]的时候ρc真实值偏离的更多。这指出了ρcv给出了与真实的特定电阻率比较接近的值。

IV. Discussions AND Conclusions