Stephen M. Gates

Silicon hydride etch products from the reaction of atomic hydrogen with Si(100)

The formation of silane, SiH4, on the Si(100) surface following atomic hydrogen chemisorption has been investigated using temperature programmed desorption (TPD) mass spectrometry and static secondary ion mass spectrometry (SSIMS). The yield of SiH3+ ions in SSIMS is correlated with trends observed in the SiH4 TPD yield, as the hydrogen surface coverage (θH) and the surface temperature during hydrogen exposure are varied. A mixture of silicon hydrides is formed on the Si(100)-(2×1) surface by adsorption of H atoms, including SiH3(a) which yields SiH4(g) during the temperature program. The peak temperature (Tp) for SiH4 in TPD occurs at 375 °C, which is 50 °C below the H2 β2 desorption peak. The maximum yield of SiH4 is observed at θH = 1.5 monolayers, with roughly 4% of all the surface hydrogen desorbing as SiH4, resulting in removal of about 1% of a monolayer of silicon atoms. The minimum hydrogen coverage needed for detectable SiH4 formation is 0.25–0.3 monolayers. The SiH4 TPD yield and the SiH3+ intensity in SSIMS are proportional to θH between θH = 0.25 and 0.5. As θH is increased beyond 0.5, the SiH4 TPD yield gradually saturates at the maximum value. Desorption of polysilicon hydrides, SixHy (x = 2,3,4) is also observed. These higher silicon hydride species desorb in TPD with Tp coincident with the β2 desorption peak for H2 at 425 °C. Molecular species Si2H6, Si3H8 and Si4H10 desorb. and mass spectrometry fragmentation patterns indicate that hydrogen deficient radical species such as Si2H2, Si2H, or Si2 are thermally desorbed. The silicon surface temperature during hydrogen adsorption dramatically affects the yields of all the silicon hydride products.

Adsorption kinetics of SiH4, Si2H6 and Si3H8 on the Si(111)-(7×7) surface

The rates and mechanisms of chemisorption on the Si(111)-(7×7) surface have been investigated under UHV conditions for SiH4, Si2H6, and Si3H8. Temperature programmed desorption of H2 has been used to quantitate adsorption of the silanes following calibrated exposures. A dramatic enhancement of the adsorption rate occurs with one SiSi bond in the molecule. A second SiSi bond has no effect on adsorption rate on the clean surface, but increases the reactivity with a hydrogen covered surface. The reactive sticking coefficient, SR, for silane is less than 0.001 near zero coverage on Si(111)-(7×7). SR (SiH4) is independent of temperature from 25 to 275°C. Reactive sticking coefficients of 0.47±0.1 for higher silanes Si2H6 and Si3H8 are observed at low coverage for 25°C surface temperature. The higher silanes adsorb on the clean surface through molecular precursor states, as evidenced by coverage independent sticking coefficients and by negative activation energies for adsorption. At low coverage, Si2D6 and Si2H6 exhibit the same sticking coefficient, so that no deuterium kinetic isotope effect is observed for the adsorption of disilane. This observation is consistent with SiSi bond breaking as the rate limiting step for conversion of the molecular precursor to chemisorbed species. The surface residence time of molecular SiH4 is inferred to be very short, relative to the residence times of molecular Si2H6 and Si3H8. SiH4 is relatively unreactive due to its short residence time and because adsorption requires SiH bond scission.

Kinetics of surface reactions in very low‐pressure chemical vapor deposition of Si from SiH4

A steady‐state kinetic model for the chemical vapor deposition (CVD) growth of Si films from SiH4 on Si(100) is presented. The only adsorbing species is SiH4 (absence of homogeneous SiH4 dissociation is presumed). Model predictions of surface hydrogen coverage and Si film growth rate as a function of growth temperature ( T ) are compared with literature values for these quantities. The rate of each reaction step is calculated at selected T. Adsorption of SiH4 and decomposition of SiH3 control the growth rate in the high T limit. In the low T limit, SiH4 adsorption is slowest but is not a simple rate determining step. The SiH4 adsorption rate is controlled by the rate of H2 desorption from two surface SiH species, producing dangling bonds.

Decomposition mechanisms of SiHx species on Si(100)‐(2×1) for x=2, 3, and 4

Silane adsorption at a surface temperature of 150 K and the surface decomposition of SiH3 and SiH2 have been investigated on the Si(100)‐(2×1) surface using static secondary ion mass spectrometry (SSIMS) and temperature programmed desorption (TPD). Silane dissociatively chemisorbs at 150 K to form SiH3 and H. At saturation, the combined coverage of these two is approximately 0.4 groups/1st layer Si atom (0.2 SiH4 adsorbed/1st layer Si atom). Using SiH4, the surface coverage of SiH3 species is varied, and the coverage‐dependant kinetics of SiH3 decomposition are examined using temperature programmed SSIMS. Changes in SiH4 exposure and source of SiH3 (di‐ vs monosilane) cause changes in surface SiH3 stability. The stability changes are interpreted as due to blocking of empty sites (dangling bonds, db) required for SiH3 decomposition to SiH2 and H. It is shown here that the decomposition temperature of SiH3 can vary from 200 to 600 K, depending on the dangling bond coverage (θdb). Subsequently, evidence for ...

Progress in the development and understanding of advanced low k and ultralow k dielectrics for very large-scale integrated interconnects—State of the art

The improved performance of the semiconductor microprocessors was achieved for several decades by continuous scaling of the device dimensions while using the same materials for all device generations. At the 0.25 μm technology node, the interconnect of the integrated circuit (IC) became the bottleneck to the improvement of IC performance. One solution was introduction of new materials to reduce the interconnect resistance-capacitance. After the replacement of Al with Cu in 1997, the inter- and intralevel dielectric insulator of the interconnect (ILD), SiO2, was replaced about 7 years later with the low dielectric constant (low-k) SiCOH at the 90 nm node. The subsequent scaling of the devices required the development of ultralow-k porous pSiCOH to maintain the capacitance of the interconnect as low as possible. The composition and porosity of pSiCOH dielectrics affected, among others, the resistance of the dielectrics to damage during integration processing and reduced their mechanical strength, thereby af...

ATOMIC H ABSTRACTION OF SURFACE H ON SI : AN ELEY-RIDEAL MECHANISM ?

The abstraction kinetics for atomic hydrogen (Hat) removal of chemisorbed D and atomic deuterium (Dat) removal of chemisorbed H are studied on single crystal Si surfaces. The surface H and D coverages are measured in real time by mass analyzing the recoiled H+ and D+ ion signals. On both Si(100) and Si(111) surfaces, the abstraction reactions are efficient, and have very low activation energies ≂0.5–1 kcal/mol. For abstraction from surfaces containing only monohydride species, the abstraction reaction probability is ≂0.36 times the adsorption rate of Hat or Dat. For the same Hat and Dat exposures, the reaction rates for Hat removal of adsorbed D and Dat removal of adsorbed H are nearly identical. All observations are consistent with a generalized Eley–Rideal abstraction mechanism, and a two‐dimensional quantum‐mechanical model is used to calculate reaction probabilities for these reactions. According to the model, the activation energies are due to enhanced abstraction rates from excited vibrational states of the adsorbed Si–H or Si–D bond. With SiH2 and SiH3 species present on the surface, the removal rate of H using Dat is decelerated, suggesting that the higher hydrides have a lower cross section for abstraction.

Decomposition of silane on Si(111)‐(7×7) and Si(100)‐(2×1) surfaces below 500 °C

Using static secondary ion mass spectrometry (SIMS) to observe the silicon hydride species formed by silane adsorption on atomically clean single crystal silicon surfaces, two distinct adsorption mechanisms are identified. Dissociation to SiH3 plus H occurs on the Si(100)‐(2×1) surface, which contains pairs of dangling bonds located on Si dimers (with Si–Si distance ≊2.4 A). In contrast, SiH2 formation in the adsorption step is indicated on the Si(111)‐(7×7) surface, where adjacent dangling bonds are separated by more than 7 A. Lower limits on the silane reactive sticking coefficient (SR) are evaluated using hydrogen coverage (ΘH) measurements after calibrated SiH4 exposures, and this limit is ≊10−5 for 25 °C gas and 100–500 °C surface temperatures. Within experimental error, SR is the same for both mechanisms on the two clean surfaces (ΘH near zero). Dependence of SR on ΘH is reported at 400 °C for both surfaces, and differences appear as ΘH exceeds 0.1 H/Si. Silane adsorption is weakly activated on Si(1...

Hydrogen coverage during Si growth from SiH4 and Si2H6

Time‐of‐flight direct recoiling (DR) measurements of surface hydrogen coverage (θH) are made in situ during chemical beam epitaxy growth of Si from Si2H6 on Si(100) as a function of temperature and disilane flux. Temperatures (T) of 300–900 °C and fluxes from 1015 to 1017 molecules cm−2 s−1 are used. Limited data for SiH4 are also presented. Predictions of θH from a steady state kinetic model are compared with the measurements, enabling the reactive sticking probability (S) of Si2H6 to be estimated at T≳500 °C.

Effect of plasma interactions with low- κ films as a function of porosity, plasma chemistry, and temperature

Integration of new low-κ interlayer dielectrics (ILD) with current damascene schemes is a continuing issue in the microelectronics industry. During integration of the ILD, processing steps such as plasma etching, resist strip, and chemical-mechanical planarization are known to chemically alter a layer of the dielectric. Here, porous organosilicate glass (OSG) ILD films, which—according to the 2004 edition of the International Technology Roadmap for Semiconductors—are projected for use in the 65 and 45 nm nodes, are investigated. spectroscopic ellipsometry, x-ray photoelectron spectroscopy, and Fourier transform infrared spectroscopy are used to characterize the modified layer of the ILD after exposure to O2 or H2 resist strip plasmas. The effects of the two types of plasma etch chemistries on the formation of the modified layer were found to differ significantly. These effects include both the degree of modification (i.e., chemical composition) and depth of the modified layer. A key difference between the...

Preparation and structure of porous dielectrics by plasma enhanced chemical vapor deposition

The preparation of ultralow dielectric constant porous silicon, carbon, oxygen, hydrogen alloy dielectrics, called “pSiCOH,” using a production 200mm plasma enhanced chemical vapor deposition tool and a thermal treatment is reported here. The effect of deposition temperature on the pSiCOH film is examined using Fourier transform infrared (FTIR) spectroscopy, dielectric constant (k), and film shrinkage measurements. For all deposition temperatures, carbon in the final porous film is shown to be predominantly Si–CH3 species, and lower k is shown to correlate with increased concentration of Si–CH3. NMR and FTIR spectroscopies clearly detect the loss of a removable, unstable, hydrocarbon (CHx) phase during the thermal treatment. Also detected are increased cross-linking of the Si–O skeleton, and concentration changes for three distinct structures of carbon. In the as deposited films, deposition temperature also affects the hydrocarbon (CHx) content and the presence of CO and CC functional groups.