Resist Film Thickness Research Articles

Near-field optical photolithography for high-aspect-ratio patterning using bilayer resist

We attempted to achieve a sufficiently large high-aspect-ratio patterning in near-field optical photolithography employing a bilayer resist process. In this experiment, we formed a resist pattern with a linewidth of 130nm and a depth of 550nm. This suggests that a practical method of near-field optical photolithography should become available. We also studied the dependency of an optical pattern on resist film thickness by the three-dimentional finite-difference time-domain method. The calculation results suggest that the resist film thickness affects the distribution of light in the resist.

Prediction of resolution using wet-develop type single layer and dry-development process for EUV lithography

This paper predicts the resolution of EUV lithography by using wet and dry developable resist process. We demonstrated that there is a possibility of the wet-develop type single layer resist whose type is positive-tone or negative-tone CA resist. The resist-film thickness is 200–400 nm. An exposure light (13.5 nm) reached at the bottom of resist film and the resists were reacted sufficiently by chemically amplification. We also showed that the EUV lithography process (13.5 nm, NA = 0.1) with the dry development prossess has the capability to resolve 50 nm L/S pattern. We can confirm that the space pattern of 50 nm can be resolved by dry-development.

Use of SU-8 photoresist for very high aspect ratio x-ray lithography

This paper describes the process of deep x-ray lithography using epoxy negative photoresist SU-8. Coating, softbake, exposure, post exposure bake, and development of the resist is characterised. Influence of x-ray source spectrum on the lithographic image contrast is calculated and optimal x-ray mask layers compositions for the spectrum in use are proposed. Method for resist film thickness control during casting step is reported. Temperature limit for the post exposure bake was found to ensure safe post bake to obtain maximum resist sensitivity. Optimised development and rinsing process is presented. Resist structures with aspect ratio as high as 100:1 (height:width) are demonstrated.

Control of photoresist properties: a Kalman filter based approach

The photolithography step in the semiconductor manufacturing process becomes increasingly critical as linewidths decrease for the next generation of integrated electronics. It therefore becomes necessary to reduce variations in photoresist parameters such as resist film thickness and photoactive compound concentration during manufacture. In this paper, we present a simple feedback scheme for accurately regulating these parameters. Our approach involves obtaining a simple, static process model the coefficients of which are recursively adjusted based on previous wafer measurements. We use this adaptive model to determine appropriate input setpoints for the next wafer. The effectiveness of this scheme in reducing process drift is exhibited by experimental data. Our results supports the widespread contention that modern feedback control offers the promise of improving the semiconductor manufacturing processes, often with relative ease and minimal capital cost.

Resist Performance in 5 nm Soft X-Ray Projection Lithography

Imaging experiments in 5 nm soft X-ray projection lithography (SXPL) were performed using 32:1 reduction Schwarzschild optics with NiCr/C multilayer, which was illuminated with synchrotron radiation (SR) from the SORTEC ring. Sensitivities of 0.7 µ m-thick polymethylmethacrylate (PMMA), ZEP and AZ-PN100 are 285, 34, and 15 mJ/cm2, respectively. Resist contrasts (γ-values) of 0.7 µ m-thick PMMA, ZEP and AZ-PN100 are 2.5, 1.6 and 3.1, respectively. A 0.15 µ m line-and-space pattern can be replicated by 5 nm exposure. Resist performance in 5 nm SXPL was investigated with a special emphasis on the effect of resist film thickness. Resist contrasts of PMMA did not change markedly with increasing resist thickness. By 5 nm exposure, a 0.3 µ m line-and-space pattern was clearly delineated in 0.9 µ m-thick PMMA. This result confirms that a single resist scheme is applicable in 5 nm SXPL.

The effects of gas phase convection on mass transfer in spin coating

The thickness uniformity of photoresist films deposited by spin coating critically influences the resolution of photolithography. This thickness uniformity depends on uniform evaporation from the film during drying. Simple scaling arguments demonstrate that, if the mass transfer coefficient at the surface of the wafer does not vary with radial position, then the dry coated resist film thickness will be independent of radial position. A model is presented for the compressible, laminar, steady-state, axisymmetric air flow in a spin coating apparatus for 6-in.-diam wafers. Flow fields computed by a finite-element–Newton method are used to evaluate the radial profile of the mass transfer coefficient at the surface of the rotating wafer, and to calculate the trajectories of particles that are generated as photoresist is flung from the edge of the spinning wafer. At a spin speed of 2000 revolutions/min and exhaust flow rate of 100 ℓ/min through the coater, the calculations predict that the mass transfer coefficient should be independent of radius. Comparison with film contours measured from experiments at these conditions indicates radial nonuniformities in the film thickness and suggests the importance of hydrodynamic instabilities in the gas on the uniformity of the coating.

Photobleachable negative resist for KrF excimer laser lithography.

Novel sensitizers, pyridine N-oxides, were investigated in negative deep-UV resists. Pyridine N-oxide, three isomers of methylpyridine N-oxides and aliphatic N-oxide compound were tested for their bleaching characteristics and resist characteristics in resist films, which contained these compounds as sensitizers and polyvinylphenol as a matrix polymer. 2-methyl isomer was the most sensitive in these compounds, while aliphatic N-oxide compound had very low sensitivity. Relationships between sensitizer concentration and bleaching characteristic were measured and results agreed with the Dill's model. Photoreaction of pyridine N-oxide was investigated in resist film. Pyridine was found and molecular weight of polyvinylphenol was not changed after UV irradiation. 0.4μm line and space patterns with 1.0μm resist film thickness were resolved, using a KrF excimer stepper(NA=0.42).

Novel Photoimaging Schemes at Polymer Surfaces: Their Role in High Resolution Lithography

ABSTRACTPolymer surfaces play essential roles in many technical arenas, but their importance in imaging science and technology has only recently been established. As optical lithography for integrated circuit applications pushes below 0.5μm feature sizes, ever shorter wavelengths and lenses with larger numerical aperatures are required. These narrow the depth-of-focus to such a degree that it eventually becomes less than the substrate topography, resist film thickness and wafer nonuniformity contributions to focus placement. Thick single-layer, surface-conforming resists will not be able to record the aerial image accurately.Surface imaging of planarized single-layer or bilayer organic films offers a means for minimizing the depth-of-focus constraints. This paper outlines this general concept which includes radiation-induced chemical changes in the surface and near-surface regions, amplification of these events by gas-solid reactions with inorganic and organometallic agents and plasma development using an oxygen plasma. Two approaches to surface imaging are discussed. The first employs Plasmask® resist, a very absorbing diazonaphthoquinone-functionalized novolac resin which is exposed and functionalized in the topmost several hundred nm of film. According to Coopmans, Roland and coworkers [14] and Pierrat, et al., [22] two tones are possible depending upon the processing. The present study reports results obtained for both processing modes using 248.4 nm lithography.The second approach involves imaging at the surface and utilizes one of the oldest polymer photoreactions, photo-oxidation, in the imaging step. Hydrophobic aromatic polymers are first irradiated in air to give hydrophilic groups. These sorb water selectively on the hydrophilic areas. The water is reacted in a separate step with an inorganic or organometallic compound such as TiC14 to give a metal oxide film (TiO2) on the exposed areas. Development with an oxygen plasma gives negative tone patterns because TiO2 reduces the etching rate by a factor of ∼ 500 in the exposed regions. Imaging is dependent upon the polymer structure, among other things, and is optimized at shorter wavelengths likely to be used in future exposure systems. From the present results we feel that surface imaging resists may realize 0.25 μm resolution at 193 nm and that <0.10 μm resolution may be achieved using x-ray radiation.

The effect of ion implantation on polymer mask resistance to ion beam etching

The effect of Ar + ion implantation with energy of 150 keV (doses 10 114−10 16 ions cm −2) into negative EN-14 electron resists on their resistance to subsequent ion-beam etching by argon ions has been investigated. Studies of structural reconstructions in resist films using the method of infrared spectroscopy show a significant change in the character of the infrared spectra after ion-beam implantation due to an increase in the atomic binding energy in the modified resist. This leads to a great reduction of the etch rate of modified resist layer (a factor of three reduction at doses of 10 15 ions cm −2). A decrease in resist film thickness is also observed. The simultaneous effect of these two factors is that the resulting resistance of modified resist masks to ion-beam etching does not increase. However, application of resist masks with a reduced etch rate modified by ion implantation permits one to decrease the lateral shrinkage of the etched structures and to increase the precision of the ion-beam etch process.

A new contrast enhancement technique for electron-beam lithography

A new electron‐beam process technique is proposed to improve resolution and aspect ratio in negative resist patterns. With this technique a modified surface layer, which acts as a dissolution barrier to the developer solution, is introduced on top of the negative resist film. The relationship between development rate and time is varied, and therefore, it becomes a nonlinear relationship. By exploiting this nonlinear development characteristic, resist film thickness loss during development in the electron exposed area can be suppressed significantly. This concept of contrast enhancement is confirmed using an electron‐beam resist profile simulator prior to experiment. Since negative resist microresist for shorter wavelengths (MRS) show strong absorption at deep UV, deep‐UV flood exposure on MRS provides ideally modified surface layers. The new process, combined deep‐UV and electron‐beam exposure, realizes 0.3–0.4‐μm line and space patterns delineated with very little film thickness loss. This technique is also an efficient way of enhancing the resist sensitivity.

Tapered Resist Wall Profiles by Flood Exposure

A resist wall profile tapering procedure is proposed for positive resist systems that are normally designed to produce near vertical wall profiles. A flood exposure step has been added, but no change in the softbake, develop, or postbake procedure is required. At the same contact hole dimension, different ratios of flood to pattern exposure dose produce different slopes for the resist profiles. The unexposed resist film thickness loss can be used to monitor the resist tapering results. SAMPLE resist exposure and development simulations are done to analyze the effects of substrate reflectivity, thickness variations, and postexposure bakes on the resist wall profile at various flood exposure dose fractions.

Transient behavior of spin-coated resist film thickness based on invariance of viscous fluid under similarity transformation

The spin-coating method that has been used as a standard procedure in the semiconductor industry is investigated theoretically. A phenomenological model is derived to describe transient behaviors of the resist film thickness on a rotating circular wafer. The model is constructed by leaving invariant the Navier-Stokes and continuity equations that describe the viscous resist fluid under the similarity transformation of variables. The film thickness H as a function of solvent concentration c for transient time t is generally expressed as H(t, r) = K(\Omega t, v(c)^{- 1/2} \Omega^{1/2}r) v(c)^{1/2}\Omega^{-1/2}, K being a dimensionless function. Here r; v; \Ω and R are the two-dimensional space vector in the wafer, viscosity, rotational speed, and wafer radius, respectively. This relationship is used to yield the transient behaviors of the film thickness that is described by a simple partial differential equation similar to a Fokker-Planck equation by assuming azimuthal symmetry in the wafer. By applying the singular perturbation technique, we obtain an analytic approximate solution that exhibits various important features such as a time development of the film thickness. The calculated film thickness is qualitatively consistent with experiments.

A mathematical model for spin coating of polymer resists

The success of lithographic processes in microelectronics fabrication depends on the reproducible generation of desired polymer resist film thickness and profile uniformity. Numerous process variables affect the outcome of spin coating of resists on wafers. A thorough understanding of the intricate interdependence of process parameters is essential to guide future process design and improvement. A mathematical model is derived to elucidate the dominant mechanisms governing film formation. The non-Newtonian character of the resist solution is taken into account, as well as the changes in resist viscosity and solvent diffusivity with changing polymer concentration. Results obtained from this model show that polymer film thickness is controlled by convective radial flow of the resist solution and solvent evaporation. The former process governs film thickness during the early stages of the process, while the latter becomes significant in later stages. The model accurately describes the experimentally observed dependence of film thickness on the variables affecting the spin-coating process.

An investigation of spin coating of electron resists

AbstractThis paper reports the results of spin coating studies of an electron resist using several different solvents. Two problems which were frequently observed during the spin coating studies were the appearance of orange peel coatings and the formation of cloudy films. The appearance of orange peel is attributed to the rapid drying of low‐boiling solvents. The formation of a cloudy film is believed to be caused by the hygroscopic nature of the solvents used. The resist film thickness can be approximated by the relationship l = KC2.1/ω0.5, where l is film thickness, C is volume fraction of polymer resist, K is a constant and ω is the rotation speed. The results are compared with the theoretical modelings of Washo (1) and Meyerhofer (2).