Pattern collapse

The collapse of patterns is a phenomenon that occurs during the rinse step of the development process of resists. As developer and rinse deionized water are removed from the developed wafers, surface tension forces pull closely spaced adjacent lines together (see Fig. 11.44). The outside features are more susceptible to collapse than those on the inside. Factors that influence pattern collapse include the aspect ratio (height/width), pitch, and mechanical properties of the resist. 

The origin of this phenomenon can he traced to the drying step of the liquid development process. During the development step, after the resist-patterned wafer has been contacted with the developer solution for a given length of time and subsequently rinsed with deionized water, the level of the rinse liquid at some point attains a condition similar to that shown in Fig. 11.45, where the space between adjacent resist lines is partially filled with fluid. The fluid meniscus exhibits a curvamre due to the differences in pressure across the fluid interface that result from surface tension in the confined space between the resist lines. Tanaka et al. developed a cantilever beam mechanical model for describing pattern collapse. The Laplace equation relates the pressure differential across the meniscus 

Graffenberg, S. Patel, G.K. Rich, H.B. Cao, and P.F. Nealey, Pattern collapse in high aspect ratio DUV and 193 nm resists, Proc. SPIE 3999, 313 321 (2000). 

232t ]y[ Morgami, and N. Atoda, Mechanism of resist pattern collapse during development 

The pressure differential across the meniscus exerts capillary forces that act perpendicularly and inward from the resist sidewall. Thus, the capillary force on the resist line is given by  

---

Note that the quantity (1- Q D)) can be thought of as a disorder parameter distinguishing the different regimes of behavior. Numerical measurements of this disorder parameter around the transition point at. which the zigzag pattern collapses spontaneously (i.e. around etc 3.92) suggest a critical-like behavior [kaneko89a] ... 

Here cto is the isotropic shift (1/3 [cti, 0-21 -1- ct3,]) and tree is the projection of the chemical shift tensor along the spinning axis and defined analogously to Eq. (3). In general this latter term produces a powder pattern when summed over nuclei at all orientations. However, when cos 0 = l/J3, the powder pattern collapses and only the isotropic value remains, i.e., CTjj/ = 0-5,. If this magic angle is misset by e radians, then (9)... 

The exposure time in selective exposure increases as the pattern becomes thinner, causing pattern collapse as observed by SEM. The results indicate the size of the width of the pattern at which the pattern collapse occurs. 

Sensitivity Limiting Pattern Collapse Contact Angle... 

The Effect of Adsorbed Cationic Surfactant on the Pattern Collapse of Photoresist Lines in Photolithographic Processes... 

Abstract A crucial problem in the manufacturing of high aspect ratio structures in the microchip production is the collapse of photoresist patterns caused by imbalanced capillary forces. A new concept to reduce the pattern collapse bases on the reduction of the capillary forces by adsorption of a cationic surfactant. The application of a cationic surfactant rinse step in the photolithographic process leads to a reduction of the pattern collapse. Physicochemical investigations elucidate the mechanism of surfactant adsorption... 

Keywords Adsorption Capillary forces Cationic surfactant Hydrophobizing Pattern collapse... 

Fig. 1 Cross-section SEM images of free standing photoresist lines (left) and photoresist structures after pattern collapse (right).

To maintain the reduction of the feature size, it is of essential importance to prevent the pattern collapse. Several approaches have been proposed ... 

The highest efficiency to reduce pattern collapse was found for concentrated surfactant solutions. This may, however, have serious disadvantages. Watanabe et al. [15] noticed that high surfactant concentrations melt the photoresist and amplify the pattern collapse. In contrary, Miyahara et al. [16] found that some nonionic surfactants stabilize the photoresist surfaces and prevent melting. In addition, concentrated surfactant solutions tend to foam which also might cause structure defects. Zhang et al. [17] therefore proposed a less foaming diol-type surfactant for the reduction of pattern collapse. 

It can be summarized that the addition of surfactants to the rinse liquid in many cases leads to reduced pattern collapse. There is, however, no systematic study of the performance of surfactants neither of the surfactant concentration nor of the surfactant type or chain length. None of the authors investigated the adsorption of the surfactant at the interface photoresist - fluid in more detail. Therefore the effect of surfactants to reduce pattern collapse is not really understood by now. 

On one hand it is investigated how the addition of cationic surfactants affects the pattern collapse of 193 nm photoresist lines. On the other hand, the adsorption of the surfactant on model photoresist surfaces is explored by a variety of surface chemical methods. Of special interest is how the surfactant changes the surface properties of the photoresist as surface potential and wettability. For an optimum modelling of the properties of real photoresist structures, both unexposed photoresists and photoresists that have been UV exposed, baked and developed are studied. 

As a measure of pattern collapse, the maximum usable exposure latitude (MUEL) was defined ... 

Influence of the Cationic Surfactant Rinse on the Pattern Collapse... 

To investigate whether the rinse with cationic surfactants does affect the pattern collapse, the photolithographic process as described in Experimental was performed without surfactant rinse and with an additional rinse step with four different concentrations of the surfactant. The results of these experiments have been discussed in detail by Wun-nicke et al. [21], Here only a short summary is given. 

In Fig. 2, top-down SEM images of photoresist structures of variable line width are shown after cationic surfactant rinse (upper row) and after conventional rinse without surfactants (lower row). The improvement is evident While without surfactant rinse the lines start to collapse at a width between 65 and 70 nm they collapse after surfactant rinse below 60 nm. Defect density measurements, too, showed a significant reduction of defects due to pattern collapse by surfactant rinse [21],... 

Fig. 3 Maximum usable exposure latitude (MUEL) as a function of the concentration of the surfactant solution used in the surfactant rinse. The dashed line gives the concentration cefr assumed to have the highest efficiency to reduce pattern collapse...

Further investigations have been carried out to elucidate the influence of the puddle time of the surfactant rinse on the pattern collapse. It was varied between 1 and 900 s. It has been found that the performance of the surfactant rinse did not depend on the time [21], This fact is very important since in manufacturing processes short process times are essential. 

Dry photoresist layers. The capillary forces causing pattern collapse act on the vertical sidewall of the photoresist structures. These surfaces are formed during the development at the interface between soluble and insoluble photoresist. It is generally accepted that a photoresist with a deprotection level above 80% is soluble in the developer [3], Therefore it can be concluded that the structure sidewalls consist of a blend of more than 20% protected and less than 80% deprotected photoresist. Because of their small dimensions, the sidewalls are not accessible for most surface chemical methods. To allow an investigation, they have to be modelled by flat photoresist surfaces. To simulate the properties of the real sidewalls, these surfaces have to be processed, i.e. exposed, baked and developed. One goal of this study is to find out which of the flat... 

Of special importance for the capillary forces responsible for the pattern collapse is the wettability of the photore-... 

To understand the mechanisms that contribute to the reduction of the pattern collapse by surfactant rinse it was necessary to study the adsorption behaviour of the surfactant molecules on the photoresist surface directly in the surfactant solution. 

To estimate the influence of the surfactant adsorption on the capillary forces, the wetting tension yiv cos was calculated from the values given in Fig. 10a. The results drawn in Fig. 10b show for both measurement series a minimum of the capillary forces exactly at the concentration ceff. The capillary forces are reduced by about 20% compared to water. This confirms the hypothesis that the reduction of the pattern collapse is caused by a hydropho-bizing of photoresist processed with the threshold dose by cationic surfactant adsorption. Unfortunately the inverse ADS A method could not be applied at relative surfactant concentrations >0.2 since the bubbles became unstable due to the lower surface tension. Thus it cannot be estimated how the wetting tension evolves at higher concentrations. 

Nevertheless a good coincidence between the reduction of the capillary forces under laboratory conditions with the reduction of pattern collapse in the photolithographic process was found. This supports our hypothesis that cationic surfactants are able to reduce pattern collapse by hydrophobizing the photoresist surface. It shows also that the hydrophobizing must occur in a similar way under real conditions. Thus, flat photoresist layers processed at the threshold dose are appropriate model surfaces for the sidewalls of photoresist structures. 

Goal of the present study was to investigate how solutions of cationic surfactants affect the pattern collapse in the photolithographic sub-100 nm structuring. On one hand, the solutions were applied directly in the photolithographic process to investigate their ability to reduce the pattern collapse. On the other hand, the adsorption of the surfactant on flat model photoresist layers was studied using a variety of physicochemical characterization methods and its influence on the capillary forces was determined. 

The investigation of the pattern collapse in the photolithographic process revealed that a short rinse step with cationic surfactant solutions yields a maximum pattern collapse reduction at a surfactant concentration far below the cmc. The rinsing time had no influence on the pattern collapse. 

The minimum of the capillary forces in the model system correlates with the maximum of the pattern collapse reduction in the real photolithographic process. It is concluded that the sidewalls of the photoresist structures are hydrophobized in a similar way as the model surfaces by the cationic surfactant rinse even although the conditions in the process differ from those of the characterization experiments. The resulting reduction of the capillary forces during the drying of the structures is the essential factor in the reduction of pattern collapse. 

---