The following discusses use of the new technique and Method of Direct Coulomb Explosion in Laser Ablation of Semiconductor Structures for laser ablation of silicon semiconductor structures. It is understood that the Method of Direct Coulomb Explosion in Laser Ablation of Semiconductor Structures of the present description may be used to perform material removal on any semiconductor structure, the discussion of silicon semiconductor structures considered to be exemplary only and in no way limiting in scope.
As discussed in the background, the activation of the Coulomb explosion mechanism in laser ablation is predicated on the efficient and rapid evacuation of electrons. In the prior art activation of Coulomb explosion in typical dielectrics, a seed carrier density is first created though multi-photon absorption in an intense laser field, followed by the onset of avalanche photoionization. The avalanche photoionization term is due to free carrier absorption and is therefore also operative in metals, although no seed stage is necessary. In the disclosure contained herein, the wavelength of the laser is selected to maximize the linear optical absorption such that the avalanche process becomes secondary or unnecessary. In order to understand this, consider the exemplary case of an approximately 400 nm wavelength femtosecond laser applied to the ablation of silicon. At this wavelength, well above the band edge of silicon, the linear absorption, coefficient is substantial and any multi-photon absorption may be neglected. Neglecting the carrier relaxation term, the equation which describes carrier production in a laser field becomes:
dN
e
/dt=[α(F)+αD(Ne)]×I,
where α(F) is the linear optical absorption coefficient, which depends on the electric field F, and αD(Ne) is the Drude carrier absorption, which depends on the carrier density. Note from this expression the formal connection between the Drude absorption and the avalanche absorption coefficient may be identified. The semiconductor optical absorption will dominate the carrier generation process provided a(F)≧αD(Ne). At 400 nm wavelength, the Drude absorption is well approximated by αD(Ne)≈1/ncτ×Nee2/εom*ω2, where n is the index of refraction at the laser wavelength and τ is the plasma relaxation time. The condition α(F)≧αD(Ne) implies Ne≦ncτm*ω2εoα(F)/e2. Taking n≈5.6, τ≈200 fs, m*≈0.3 me, λ≈400 nm, and α(F)≈1×106/cm (constant), the condition α(F)≧αD(Ne) is satisfied for carrier densities Ne≦7×1025/cm3. This carrier density exceeds, by several orders of magnitude, the threshold density at which material is expected to undergo ablation due to Coulomb explosion. Hence, femtosecond Coulomb explosion of Si with 400 nm wavelength does not invoke or require the avalanche photoionization mechanism. Equivalently stated, the short absorption depth of Si at 400 nm wavelength would require a plasma density ˜1025/cm3 in order to attain comparable plasma skin depths. In practice, this never occurs.
Without the avalanche photoionization term, the solution to the differential equation at the end of the laser pulse becomes: Ne≈α(F)×I×τp, where τp is the laser pulse width. Assuming values of α(F)≈1×106/cm, τp≈150 fs, and Nth≈1×1023/cm3, where Nth is the approximate Coulomb explosion threshold, a threshold intensity of Ith≈3.3×1011 W/cm2 may be estimated. This is well below the ˜1013-1014 W/cm2 thresholds commonly seen for avalanche photoionization. For pulse lengths up to the plasma relaxation time τ, a longer laser pulse requires proportionally less intensity, since it is the laser fluence which governs the final carrier density. The presence of a relaxation term will suppress the induced carrier density by a factor of ˜exp{−τp/τ}. For Si and the carrier densities under discussion herein, relaxation times are commonly estimated to be in the range of ˜20-200 fs, corresponding to Si mobilities μ≈102-103 cm2/V·s.
An instructive counterpoint to this example is obtained when the activation of Coulomb explosion in silicon through the conventional avalanche photoionization is considered. In an exemplary case, consider an approximately 800 nm wavelength laser applied to the conventional avalanche photoionization activation of Coulomb explosion in silicon. This wavelength is above the band edge of silicon, so multi-photon absorptions may again be neglected. The linear absorption must develop a seed plasma density of Ne≈1021/cm3 within approximately the first half of the pulse. Using Ne≈α×I×τp÷2, with α≈1×103/cm and τp≈200 fs, the estimated intensity to seed the avalanche I≈2.5×1012 W/cm2, or greater. An analogous estimate of the intensity required to seed the avalanche, for the case of 400 nm radiation of silicon, results in I≈5×109 W/cm2. However, as discussed, this estimate is purely hypothetical since the linear optical absorption will continue to dominate the avalanche photoionization even well beyond the Coulomb explosion threshold density.
Another key advantage inherent to the Method of Direct Coulomb Explosion in Laser Ablation of Semiconductor Structures is the ability to use an externally applied electric field to enhance the linear optical absorption. Nearby to strong features in the optical absorption spectra of a semiconductor material, the field dependence of the optical absorption coefficient will be large. In particular, the optical absorption of semiconductors and dielectrics undergoes significant redshifting in strong electric fields. This effect is known as the Franz-Keldysh effect [Keldysh, 1958], and may be explained as follows. Since the strong electric field tilts the semiconductor bandstructure spatially, electrons may tunnel from the valence band to the conduction band. Thus, at photon energies just below strong absorption features, the redshift of the absorption is due to photon-stimulated tunneling [Yu & Cardona, 2001]. The dependence of the absorption coefficient on electric field F is given by:
α(F)=KF2 exp{−C/F},
where K and C are constants which depend on the “zero field” absorption, the effective electronic mass, and the difference in laser wavelength from the strong absorption feature. The “zero field” absorption is the optical absorption of the material wherein no electric field is applied. The expression above describes the redshifting of an optical absorption edge in the presence of a strong electric field [Keldysh, 1958]. The 400 nanometer wavelength laser is nearby to such an absorption edge in silicon, which occurs at a wavelength of approximately 375 nanometers. As an externally applied field approaches 107 V/cm, the silicon optical absorption coefficient increases about an order of magnitude from 105/cm to 106/cm. Thus, an externally applied electric field may be used to enhance and control the optical absorption coefficient.
It is therefore instructive to understand the ablation threshold for Coulomb explosion in the case of electric fields F ˜106-107 V/m. For the laser pulse lengths under consideration herein, the electrons will thermally equilibrate but will not have time to transfer their energy to the lattice via collisions or thermal diffusion. The kinetic energy the electron attains from the field is given by ΔK=m*vS2/2, where vS is the saturation velocity (ΔK is on the order ˜10 meV, so it does not add substantial energy to the electron) [Yu & Cardona, 2001,]. The equation for the average change in electron temperature Te (or energy) due to laser absorption is given by:
c
e(Te)Ne×∂Te/∂t=−∂S/∂x; S=(1−R)Io×exp{−2x/δ},
where ce is the specific heat of conduction electrons, and S is the absorbed energy flux as a function of depth (R is the reflectivity and Io is the incident intensity) [Chichkov, 1996]. Near the laser ablation threshold Te≈EF, and ce≈3/2. The average change in electron energy then may be expressed: Te≈4/3×(1−R)Iτp/δNat×exp{−2x/δ}. The threshold fluence for ablation is conventionally determined by the requirement the electron energy Te reach, within a surface layer x<<δ, a value equal to the sum of the atomic binding energy and the electron ionization energy, i.e. Te≧Eb+Ip. This is the threshold condition for energetic electrons to escape the solid and produce a strong charge separation field which then results in the Coulomb explosion of the surface layers [Gamaly, 2002]. Solving the relation Te≧Eb+Ip for the fluence Ithτp, yields the threshold dependence on the static field:
(1−R)Ithτp≈3/8×Nat/α(F)×(Eb+Ip).
For Si at 400 nm, the reflectivity is roughly 50%. The Si binding energy is Eb≈4.62 eV/atom, and the ionization potential is Ip≈1.12 eV. The predicted threshold fluence is then Ithτp≈20 mJ/cm2. For a pulse length of approximately 200 fs, this corresponds to a pulse intensity of approximately 1011 W/cm2. This estimate agrees with the earlier estimate from use of the linear semiconductor absorption to generate threshold carrier densities. It is instructive to compare this to the expression for threshold derived for avalanche activated ablation. For a plasma skin layer, we have 1−R≈2 ωδ/c, which is justified at critical density. The threshold fluence then becomes Ithτp≈3/8×λNat/2π×(Eb+Ip). From this, it is seen the ablation threshold in the Method of Direct Coulomb Explosion in Laser Ablation of Semiconductor Structures is reduced by a factor≈2π/λ·α(F) relative to the conventionally used avalanche ablation mechanism. For the exemplary case of ablation of silicon using 400 nm radiation herein, this results in a more than six (6) times reduction in the threshold intensity. Using the relation α=4πk/λ, where k is the imaginary part of the complex index of refraction, known as the extinction coefficient, it may be seen the ablation process will proceed through semiconductor linear optical absorption, and not the avalanche (Drude) absorption, provided k≧½. Also note in the Method of Direct Coulomb Explosion in Laser Ablation of Semiconductor Structures, k depends on the applied electric field and therefore field induced shifts of the optical absorption may be used to effect reductions in ablation threshold.
In order to estimate the ablation rates attained in the Method of Direct Coulomb Explosion in Laser Ablation of Semiconductor Structures, we note that due to the exponential decrease of the electron energy into the surface, the ablation depth is of the order of the absorption depth and exhibits a logarithmic dependence on fluence. The ablation depth may therefore be written Lab˜δ/2×ln{I/Ith}. The number of particles ablated per unit area per unit time can be estimated as j˜LabNat/τp˜δ/2×Nat/τp×ln{I/Ith}. For 400 nm irradiation of silicon using a femtosecond laser pulse, the ablation rate is estimated ˜1.67×1029/cm2·s×ln{I/Ith}. For strong fields ˜107 V/m, we obtained an estimate of Ith˜5×1011 W/cm2. Assuming an operational intensity such that I/Ith≈2, we find j˜1029/cm2·s. The number of ablated ions per laser pulse is Lab×Nat×S, where S is the surface area available for ablation. In the exemplary planar configuration, S is the focal area≈π×10−8 cm2 (S≈π
Therefore, as disclosed herein, the Method of Direct Coulomb Explosion in Laser Ablation of Semiconductor Structures provides a new and remarkable capability to effect and control laser ablation, and in so doing, substantially departs from the conventional concepts and designs of the prior art.
As to a further discussion of the manner of usage and operation of the present disclosure, the same should be apparent from the above description. Accordingly, no further discussion relating to the manner of usage and operation will be provided.
With respect to the above description then, it is to be realized that the optimum dimensional relationships for the parts of the disclosure, to include variations in size, materials, shape, form, function and manner of operation, assembly and use, are deemed readily apparent and obvious to one skilled in the art, and all equivalent relationships to those described in the specification are intended to be encompassed by the present disclosure.
Therefore, the foregoing is considered as illustrative only of the principles of the disclosure. Further, since numerous modifications and changes will readily occur to those skilled in the art, it is not desired to limit the disclosure to the exact construction and operation described, and accordingly, all suitable modifications and equivalents may be resorted to, falling within the scope of the disclosure.
The present invention claims benefit of U.S. Provisional Patent Application Ser. No. 60/815,109, filed on Jun. 20, 2006, which is incorporated herein in its entirety by reference.
Number | Date | Country | |
---|---|---|---|
60815109 | Jun 2006 | US |