CN101192009B - Method for establishing OPC model - Google Patents

Abstract

A method of building optical proximity correction model comprises the following steps of collecting optical proximity correction data to a database from processes, determining an initial value X0, an end value Xn and a fixed step size h, calculating initial value slope coefficient according to the initial value X0, obtaining an initial optimal step size Y0 according to the initial value slope coefficient, obtaining a first additive value X1 according to the initial value X0 and the initial optimal step size Y0, repeating the steps, working out a second additive valueX2, a third additive value X3,..., till the end value Xn, fitting the initial value X0, the first additive value X1, the second additive value X2, the third additive value X3,..., till an n-1 additive value and the end value Xn. Through the above steps, as step size values can be regulated according to the slope coefficient worked out, regulation step number can be reduced and simulation time is shortened.

Description

Set up the method for OPC model

Technical field

The present invention relates to semiconductor fabrication process, particularly the photomask data processing technique is especially set up the optical proximity correction model.

Background technology

The current realization of photomask data deviation is usually based on some models, and described model is calibrated in concrete photoetching process, is commonly referred to as OPC model.For example, the correction of optical proximity effect (OPE) often need be attempted " calibration " photoetching process so that compensation OPE.Prior art comprises so-called calibration parameter is associated with OPC model that this need carry out one group of detailed optical parameter measurement at different feature locations.It is many more that measurement data is collected, and the accuracy of calibration parameter is just good more.Yet, for reliable OPC model parametric calibration different under environment, optical parameter measurement more than the common needs in different critical characteristic positions are hundreds of, this is labour intensive and time consuming work, what is worse owing to the reason of experience level, take what kind of optical parameter measurement often to become and depend on the operator, thus this obvious negative effect parametric calibration limited the overall validity of optical nearing correction technique.

U.S. Patent number is 6,563,566 disclose and carry out illumination by a series of calculating of attempting linearization mask transmission optimization and optimize and photomask optimization, wherein disclose maximization minimum NILS (normalized image log slope) and selection and will be used in various conditions in this calculating.Because rely on the symmetry of photomask, this calculating is restricted, yet the linearization of the photomask transmission of use need be used several approximate in calculating, rather than actual imaging equation itself, this use photomask formation institute important plan as the time can produce error.The linearization of photomask transmission also needs to use a lot of variablees, and this needs a large amount of computing times and carries out this calculating.Therefore along with the reducing of logical implication size, be necessary to provide the photomask implementation of using the minimum of computation time accurately to form desired image.

The method for making of existing OPC model, for example this optical parametric of logarithm value aperture is optimized, and as shown in Figure 1, S101 at first gathers the numerical aperture data to database from technology; S102 determines initial value X ₀, end value X _nWith fixed step size h; S103 is with initial value X ₀Carry out addition with fixed step size h, draw the first additive value X ₁S104 is with the first additive value X ₂Carry out addition with fixed step size h, draw the second additive value X ₂S105 continues above-mentioned steps, and cycle calculations goes out the value added X of third phase ₃, the 4th additive value X ₄To end value X _nS106 is with initial value X ₀, the first additive value X ₁, the second additive value X ₂, the value added X of third phase ₃, the 4th additive value X ₄End value X _nCarry out match.

OPC model is a function, has characterized the difference of the optical field distribution that the optical field distribution that calculates by optical theory and actual amount record.This function is relevant with some optical parametrics, such as numerical aperture.When that was some occurrences when numerical aperture, function had minimum value, and it is best just to illustrate that the light field of model description and actual light field are coincide.

In general our function F (X of investigating ₀) all be the multivariate function usually, its independent variable parameter has a plurality of, and what Fig. 2 represented is when independent variable X is two (two-dimensional function), F (X ₀) possible distribution shape.

As shown in Figure 3, suppose the function F (X of investigation ₀) be two-dimensional function, that at first determines the initial value X of two parameters before beginning modeling with the existing method of setting up OPC model ₀, end value X _NAnd fixed step size value h, the combination of the value of used parameter just can be regarded as on the lattice point that is distributed in a two-dimensional mesh (among the figure in the ellipse) in optimizing process so, and the parameter value combination outside the lattice point is not to be used in calculating.

The method for making of existing OPC model because step-length is a fixed value, if intermediate data is a lot of between initial value and the end value, can cause the adjusting step number too many, and simulated time is elongated.

Summary of the invention

The problem that the present invention solves provides a kind of method of setting up OPC model, prevents because step-length is a fixed value, if intermediate data is a lot of between initial value and the end value, can cause the adjusting step number too many, and simulated time is elongated.

For addressing the above problem, the invention provides a kind of method of setting up OPC model, comprise the following steps: from technology, to gather optical nearing correction data to database; Determine initial value X ₀, end value X _NWith fixed step size h; According to initial value X ₀Calculate the initial value slope

According to the initial value slope

Draw initial optimization step-length Y with fixed step size h ₀According to initial value X ₀With initial optimization step-length Y ₀Draw the first additive value X ₁Repeat above-mentioned steps, calculate the second additive value X ₂, the value added X of third phase ₃The n-1 additive value is to end value X _nWith initial value X ₀, the first additive value X ₁, the second additive value X ₂, the value added X of third phase ₃... n-1 additive value and end value X _nCarry out match.The formula that calculates additive value is: X _n=X _N-1+ Y _N-1, Y wherein _N-1For optimizing step-length, n is a natural number.

Described calculation optimization step-length Y _N-1Formula be: $Y_{n-1}=\frac{\∂F}{\∂X}{|}_{X=X_{n-1}}\×h,$ Wherein

Be slope.

The formula of described slope calculations is: $\frac{\∂F}{\∂X}{|}_{X=X_{n-1}}=\frac{F(X_{n-1}+\mathrm{\ΔX})-F\left(X_{n-1}\right)}{\mathrm{\ΔX}},$ Wherein Δ X is the increment of independent variable X, F (X ₀) equal X for optical nearing correction data ₀The time, the degree of agreement when OPC model and optical nearing correction actual correction.

Compared with prior art, the present invention has the following advantages: because step-length can be regulated according to the slope that calculates, can realize regulating step number like this and reduce, simulated time shortens.

Description of drawings

Fig. 1 is the process flow diagram that prior art is set up the OPC model method;

Fig. 2 is prior art F (X when working as independent variable X and being two ₀) possible distribution shape;

Fig. 3 is the grid chart that prior art is set up OPC model;

Fig. 4 is the process flow diagram that the present invention sets up the OPC model method;

Fig. 5 is F (X of the present invention ₀) curve map of function;

Fig. 6 is the effect comparison diagram of the OPC model method of prior art and the present invention's foundation.

Embodiment

The linearization of photomask transmission need be used a lot of variablees, this needs a large amount of computing times carries out this calculating, and prior art is because step-length is a fixed value, if intermediate data is a lot of between initial value and the end value, can cause the adjusting step number too many, simulated time is elongated.So reducing along with the logical implication size, be necessary to provide the photomask implementation of using the minimum of computation time accurately to form desired image, the present invention can realize regulating step number like this and reduce because step-length can be regulated according to the slope that calculates, and simulated time shortens.For above-mentioned purpose of the present invention, feature and advantage can be become apparent more, the specific embodiment of the present invention is described in detail below in conjunction with accompanying drawing.

Fig. 4 is the process flow diagram that the present invention sets up the OPC model method.As shown in Figure 4, execution in step S201 gathers optical nearing correction data to database from technology; S202 determines initial value X ₀, end value X _nWith fixed step size h; S203 is according to initial value X ₀Calculate the initial value slope

S204 is according to the initial value slope Draw initial optimization step-length Y with fixed step size h ₀S205 is according to initial value X ₀With initial optimization step-length Y ₀Draw the first additive value X ₁S206 repeats above-mentioned steps, calculates the second additive value X ₂, the value added X of third phase ₃The n-1 additive value is to end value X _nS207 is with initial value X ₀, the first additive value X ₁, the second additive value X ₂, the value added X of third phase ₃... n-1 additive value and end value X _nCarry out match.In the present embodiment, the formula that calculates additive value is: X _n=X _N-1+ Y _N-1, Y wherein _N-1For optimizing step-length, n is a natural number.The concrete first additive value X ₁=X ₀+ Y ₀, the second additive value X ₂=X ₁+ Y ₁, the value added X of third phase ₃=X ₂+ Y ₂N-1 additive value X _N-1=X _N-2+ Y _N-2

Described optimization step-length Y _N-1Computing formula be: $Y_{n-1}=\frac{\∂F}{\∂X}{|}_{X=X_{n-1}}\×h,$ Wherein

Be slope.In the present embodiment, concrete initial optimization step-length $Y_0=\frac{\∂F}{\∂X}{|}_{X=X_0}\×h,$ Second optimizes step-length $Y_1=\frac{\∂F}{\∂X}{|}_{X=X_1}\×h,$ The 3rd optimizes step-length $Y_2=\frac{\∂F}{\∂X}{|}_{X=X_2}\×h......$ N-1 optimizes step-length

$Y_{n-2}=\frac{\∂F}{\∂X}{|}_{X=X_{n-2}}\×h.$

In the present embodiment, the formula of described slope calculations is: $\frac{\∂F}{\∂X}{|}_{X=X_{n-1}}=\frac{F(X_{n-1}+\mathrm{\ΔX})-F\left(X_{n-1}\right)}{\mathrm{\ΔX}},$ Wherein Δ X is the increment of independent variable X, F (X ₀) equal X for optical nearing correction data ₀The time, the degree of agreement when OPC model and optical nearing correction actual correction.Concrete initial value slope is $\frac{\∂F}{\∂X}{|}_{X=X_0}=\frac{F(X_0+\mathrm{\ΔX})-F\left(X_0\right)}{\mathrm{\ΔX}},$ The first additive value slope is $\frac{\∂F}{\∂X}{|}_{X=X_1}=\frac{F(X_1+\mathrm{\ΔX})-F\left(X_1\right)}{\mathrm{\ΔX}},$ The second additive value slope is $\frac{\∂F}{\∂X}{|}_{X=X_2}=\frac{F(X_2+\mathrm{\ΔX})-F\left(X_2\right)}{\mathrm{\ΔX}},$ The value added slope of third phase is $\frac{\∂F}{\∂X}{|}_{X=X_3}=\frac{F(X_3+\mathrm{\ΔX})-F\left(X_3\right)}{\mathrm{\ΔX}},$ N-1 additive value slope is

$\frac{\∂F}{\∂X}{|}_{X=X_{n-1}}=\frac{F(X_{n-1}+\mathrm{\ΔX})-F\left(X_{n-1}\right)}{\mathrm{\ΔX}}.$

Fig. 5 is F (X of the present invention ₀) curve map of function.Suppose that we optimize F (X ₀) time only consider a variable parameter, function is with respect to the funtcional relationship of variable element such as the curve representation among Fig. 5 so, OPC model is from initial value X ₀Set out, find optimized parameter value (the curve the lowest point part among the figure), after the optimum value that finds all parameters, these the most optimized parameter values are carried out match, finish optimization OPC model.

Fig. 6 is the effect comparison diagram of the OPC model method of prior art and the present invention's foundation.As shown in Figure 6, when parameters optimization is 4,6 or 8 respectively, set up the method for OPC model and method that the present invention sets up OPC model is finished the needed time of last optimization with existing.When parameters optimization was 4, existing to set up the required time of OPC model be 12468s, and the present invention to set up the required time of OPC model be 4513s; When parameters optimization was 6, existing to set up the required time of OPC model be 30465s, and the present invention to set up the required time of OPC model be 10145s; When parameters optimization is 4, having now and setting up the required time of OPC model is 48627s, and the present invention sets up the required time of OPC model is 16031s, the present invention is owing to carry out slope calculating with each value, and then slope be multiply by fixed step size as optimizing step-length, make and regulate the step number minimizing, simulated time shortens.

Though the present invention with preferred embodiment openly as above; but it is not to be used for limiting the present invention; any those skilled in the art without departing from the spirit and scope of the present invention; can make possible change and modification, so protection scope of the present invention should be as the criterion with the scope that claim of the present invention was defined.

Claims (1)

1. a method of setting up OPC model is characterized in that, comprises the following steps:

From technology, gather optical nearing correction data to database;

Determine initial value X ₀, end value X _nWith fixed step size h;

According to initial value X ₀Calculate the initial value slope

According to the initial value slope

Draw initial optimization step-length Y with fixed step size h ₀

According to initial value X ₀With initial optimization step-length Y ₀Draw the first additive value X ₁

Repeat above-mentioned steps, calculate the second additive value X ₂, the value added X of third phase ₃... n-1 additive value X _N-1To end value X _n, n is a natural number;

The described first additive value X ₁, the second additive value X ₂, the value added X of third phase ₃... n-1 additive value X _N-1To end value X _n, obtain in the following way: X _n=X _N-1+ Y _N-1, Y wherein _N-1For optimizing step-length;

Described optimization step-length Y _N-1Obtain in the following way:

Wherein

Be slope;

With initial value X ₀, the first additive value X ₁, the second additive value X ₂, the value added X of third phase ₃... n-1 additive value X _N-1With end value X _nCarry out match, set up OPC model photoetching process is calibrated.

Images