Atomic Layer Deposition of TiO2 on Surface Modified Nanoporous

Sep 4, 2013 - This can be explained by the faster deposition of TiO2 in the hydrophilic top part of the film than in the hydrophobic bulk which leaves...
1 downloads 23 Views 942KB Size
Download PDF
Article pubs.acs.org/Langmuir

Atomic Layer Deposition of TiO2 on Surface Modified Nanoporous Low‑k Films Elisabeth Levrau,*,† Kilian Devloo-Casier,† Jolien Dendooven,† Karl F. Ludwig,‡ Patrick Verdonck,§ Johan Meersschaut,§ Mikhail. R. Baklanov,§ and Christophe Detavernier† †

Department of Solid State Sciences, COCOON, Ghent University, Krijgslaan 281/S1, B-9000 Ghent, Belgium Physics Department, Boston University, 590 Commonwealth Avenue, Boston, Massachusetts 02215, United States § IMEC, Kapeldreef 75, B-3001 Leuven, Belgium ‡

ABSTRACT: This paper explores the effects of different plasma treatments on low dielectric constant (low-k) materials and the consequences for the growth behavior of atomic layer deposition (ALD) on these modified substrates. An O2 and a He/H2 plasma treatment were performed on SiCOH low-k films to modify their chemical surface groups. Transmission FTIR and water contact angle (WCA) analysis showed that the O2 plasma changed the hydrophobic surface completely into a hydrophilic surface, while the He/H2 plasma changed it only partially. In a next step, in situ X-ray fluorescence (XRF), ellipsometric porosimetry (EP), and Rutherford backscattering spectroscopy (RBS) were used to characterize ALD growth of TiO2 on these substrates. The initial growth of TiO2 was found to be inhibited in the original low-k film containing only Si-CH3 surface groups, while immediate growth was observed in the hydrophilic O2 plasma treated film. The latter film was uniformly filled with TiO2 after 8 ALD cycles, while pore filling was delayed to 17 ALD cycles in the hydrophobic film. For the He/H2 plasma treated film, containing both Si-OH and Si-CH3 groups, the in situ XRF data showed that TiO2 could no longer be deposited in the He/H2 plasma treated film after 8 ALD cycles, while EP measurements revealed a remaining porosity. This can be explained by the faster deposition of TiO2 in the hydrophilic top part of the film than in the hydrophobic bulk which leaves the bulk porous, as confirmed by RBS depth profiling. The outcome of this research is not only of interest for the development of advanced interconnects in ULSI technology, but also demonstrates that ALD combined with RBS analysis is a handy approach to analyze the modifications induced by a plasma treatment on a nanoporous thin film.

■

INTRODUCTION With the downscaling of microelectronic devices, integration of low dielectric constant (low-k) materials into IC (Integrated Circuit) interconnects has become necessary. One important class of low-k dielectrics is the porous organosilicate glasses (SiCOH) containing an inorganic silica-like matrix and organic hydrophobic surface groups. Their hydrophobicity and high porosity reduce the k value down to k = 2.0, thus making them very promising for IC applications.1,2 Plasma technology is frequently used in semiconductor processing for etching, deposition, cleaning, and surface treatments of materials. Plasma species such as radicals, but also vacuum ultraviolet light (VUV), can damage the low-k films by irreversible and nonuniform chemical modifications of the surface and densification.3−5 For example, oxygen atoms can attack and remove the organic surface groups; hence, the low-k material turns more hydrophilic leading to a significant increase in dielectric constant and leakage current.6 One way to deal with this problem is to use less-damaging plasma chemistry such as H2-based instead of O2-based plasmas.7 Nonetheless, this softer H2-based plasma chemistry also induces damage to the material.8 Although extensively studied in the literature, the © 2013 American Chemical Society

extent of plasma damage is not yet completely understood. The radicals in a plasma limit the conformality of the treatment due to radical recombination. Consequently, the modifications induced by a plasma treatment can be limited to a certain depth and/or have a gradient in a porous thin film.9 Still, the modifications induced by these plasma treatments need to be characterized. The nonconformality of the plasma treatment can be used for local modifications of the porous material. Lowk materials have to be sealed from the environment to prevent deterioration of the film properties by moisture uptake and metal diffusion. This protection layer needs to be limited to the top few nanometers of the porous film; hence, plasma treatments can be used to locally activate the surface groups and a sealing layer can be deposited on this modified region.10 Atomic layer deposition (ALD) is a thin film deposition technique based on sequential, self-limiting surface reactions where the precursors adsorb on all active sites available in the growth chamber and hence also within pores of a nanoporous Received: July 20, 2013 Revised: September 3, 2013 Published: September 4, 2013 12284

dx.doi.org/10.1021/la4027738 | Langmuir 2013, 29, 12284−12289

Langmuir

Article

index of the porous low-k films could be extracted. The porosity can then be calculated via the Lorentz−Lorenz equation.18 Rutherford backscattering spectrometry (RBS) was used to get a depth profile of the deposited TiO2 in the low-k films. A He+ ion beam of 1.523 MeV was used to perform these measurements. All samples were measured with a glancing exit angle to improve depth resolution. To extract depth information the analysis was performed using a 4-layer model: one layer on the top and three buried layers representing various diffusion depths from the Ti into the low-k material.

thin film.11 By limiting the presence of active sites to the top surface of a porous low-k film, ALD might seal the film without coating the interior pore surface. In order to successfully integrate these materials it is thus useful to study the plasma modifications of O2 and H2 based plasmas and the ALD growth behavior into/onto these different plasma treated porous low-k films. In this work, an O2 and a He/H2 plasma treatment have been used to modify the surface of nanoporous SiCOH low-k films. TiO2 ALD growth into/onto these substrates is used to study the extent of plasma modifications in these low-k substrates. Using synchrotron based in situ XRF measurements the amount of deposited material is monitored during the ALD process and ex situ EP measurements are used to determine the remaining porosity that is accessible by the probe molecule toluene. Additionally, RBS measurements are performed to study the depth profile of the ALD deposited material. The unique combination of ALD with these techniques allows us to not only study the plasma modification profile in low-k materials, but also establish a model for the growth behavior of TiO2 deposited by ALD in surface modified porous materials.

■

■

RESULTS AND DISCUSSION The different types of low-k films were characterized prior to the ALD of TiO2. An overview of the substrate properties is presented in Table 1. According to EP measurements all the Table 1. Overview of Material Properties of the Different Substrates pristine SiCOH low-k

type of low-k film Thickness (nm) Porosity (%) Surface area (cm2/cm2 substrate) Water Contact Angle (WCA °) Chemical surface groups present (FTIR)

EXPERIMENTAL SECTION

In this study standard silicon substrates with nanoporous SiCOH lowk films were used as starting samples. The synthesis of these films has been described before by Urbanowicz et al.12 On these pristine films, two different plasma treatments were performed. One pristine substrate was treated with an inductively coupled plasma (ICP) of O2 gas during 200 s at 60 °C. During this treatment only top power was used as described by Kunnen et al.13 Another pristine substrate was exposed to a 5 s capacitively coupled plasma (CCP) of He/H2 gas at 370 °C. In both cases the wafers were located on top of the grounded electrode. The three types of substrates were first characterized with transmission FTIR and water contact angle (WCA) measurements. The FTIR measurements have been done on a VERTEX 70 V FTIR spectrometer with KBr beamsplitter and RT-DLaTGS detector (midinfrared). Deposition of TiO2 was done in a low-vacuum home-built ALD reactor using tetrakis(dimethylamino)titanium (TDMAT, Ti(NMe2)4, 99%, Strem Chemicals) as the titanium precursor and H2O vapor as the oxygen source. The TDMAT precursor was heated in a stainless steel container at 50 °C and its vapor entered the chamber with argon as carrier gas via tubes heated to 55 °C. Gas flows were controlled by computer controlled pneumatic valves and needle valves and kept at a reactor pressure of 0.3 Pa for both precursors. A typical ALD cycle consisted of 30 s of exposure to the TDMAT precursor vapor (and argon), followed by 60 s of pumping, 30 s of exposure to the H2O vapor, and again 60 s of pumping. The exposure times were chosen after performing saturation tests on the samples to ensure saturation of the growth rate. The substrate temperature was fixed at ca. 120 °C, below the decomposition temperature of TDMAT, for all the depositions while the chamber wall was heated to 90 °C. Before ALD treatment the low-k films were evacuated in vacuum at the deposition temperature for 1 h. In situ XRF measurements were carried out in the thin film growth facility installed at Beamline X21 of the National Synchrotron Light Source (NSLS) at Brookhaven National Laboratory (BNL).14−16 A Vortex Silicon Drift detector was used to measure the fluorescent radiation from the sample. The porosity of the porous low-k films was determined using a home-built ellipsometric porosimetry (EP) setup consisting of a vacuum chamber in which a sorbent can be systematically introduced until it reaches its vapor pressure. Here, toluene was used as an adsorptive because it is known to be suitable for RT porosimetry.17 On the measurement chamber a Woollam M-2000U ellipsometer (245−1000 nm) was mounted to monitor the polarization angles psi and delta of the sample during the adsorption−desorption process. Using a Cauchy model for the fitting of these data, accurate values for the thickness and refractive

O2 plasma treated

He/H2 plasma treated

95 45 39.5

72 35 19.5

91 46.5 36.5

91.4

19.4

23.8

Si-CH3

Si-OH

Si-CH3/ SiOH

films have a disordered structure of interconnected pores with pore diameters of approximately 3 nm. The pristine and He/H2 plasma treated films are very similar in thickness and porosity, while the O2 plasma treatment resulted in a reduction of the film thickness and porosity. Compared to the pristine film, the plasma treated films have a much lower water contact angle, indicating that both plasma treatments made the surface more hydrophilic. Ex situ transmission FTIR spectroscopy was used to identify the chemical surface groups present in each film. The results are shown in Figure 1 and summarized in Table 1. The vibration band at 1000−1200 cm−1, present in all the substrate spectra, can be assigned to the Si−O bonds that make up the skeleton of the nanoporous low-k films. The slight peak shift and the decrease of the right shoulder indicate that the Si−O network is changing into a more SiO2-like material due to the plasma

Figure 1. Normalized transmission FTIR spectra of the three surface modified low-k films prior to ALD deposition. Inset: zoom of the SiCH3 stretch vibration. 12285

dx.doi.org/10.1021/la4027738 | Langmuir 2013, 29, 12284−12289

Langmuir

Article

treatments. Around 3400 cm−1 a vibration band is noticeable in the spectrum of the O2 plasma treated film, which is also somewhat present in the spectrum of the He/H2 plasma treated film and can be assigned to H-bonded Si-OH surface groups, siloxane bridges, and H2O that remained in the pores due to a higher hydrophilic surface. This hydrophilicity of the two films is revealed by WCA measurements. The sharp peak at 1274 cm−1 that is clearly visible in the spectra of the pristine and He/ H2 plasma treated films represents hydrophobic Si-CH3 surface groups (see Figure 1 inset).7 The absence of this sharp peak in the spectrum of the O2 plasma treated film proves that the O2 plasma treatment replaced all the Si-CH3 surface groups by Si− OH groups. Thus, these data show that the pristine film contains only Si-CH3 surface groups, the O2 plasma converted all Si-CH3 surface groups to Si-OH groups and the He/H2 plasma only partially converted the Si-CH3 groups to Si-OH groups. After normalization of the FTIR spectra, the peak intensities of the Si-CH3 groups of the pristine and the He/H2 plasma treated films can be quantitatively compared. Since the Si-CH3 peak intensity is halved after He/H2 plasma treatment we can derive that approximately 50% of the Si-CH3 groups were converted to Si-OH surface groups by the He/H2 plasma treatment. In our previous work,14,15 we demonstrated that in situ XRF can be used to probe the conformality of the TiO2 ALD process in a nanoporous hydrophilic thin film. Therefore, the three substrates were coated with TiO2 in a UHV chamber installed at the National Synchrotron Light Source at Brookhaven National Laboratory, and the amount of deposited Ti atoms was monitored by XRF (Figure 2). The XRF intensity curves

This can be explained by the active sites present on the internal surface of the films. The importance of Si-OH surface groups for deposition of oxide films by ALD has been reported in the literature.11,14,19−21 Si-CH3 surface groups are much less reactive toward ALD precursor molecules than Si-OH surface groups. Since the pristine and He/H2 plasma treated films contain Si-CH3 groups, the TDMAT precursor will not immediately react with them. Thus, these Si-CH3 surface groups cause an inhibited ALD growth. The growth inhibition in the He/H2 plasma treated film is less than in the pristine film, most likely due to a smaller fraction of Si-CH3 groups present. Moreover, despite the inhibited initial growth, the pores in the He/H2 plasma treated film become inaccessible for TDMAT as fast as in the O2 plasma treated film, i.e., after 8 ALD cycles, while for the pristine film 17 cycles are needed to shrink the pore size below the kinetic diameter of TDMAT. This result suggests that the accessibility of the pores in the He/H2 plasma treated film is determined by TiO2 growth on the Si-OH groups rather than on the Si-CH3 groups. Assuming complete pore filling of the pores, the amount of TiO2 deposited inside the films should be proportional to the pore volume. The total pore volume of the film can be calculated by multiplying the thickness of the film with the open porosity. Therefore the total pore volume of the O2 plasma treated film is a factor (35 × 72)/(45 × 95) = 0.59 times smaller than the pore volume of the pristine film. The XRF signal of the O2 plasma treated film after pore filling (8 cycles) is approximately 1.5/2.5 = 0.6 times smaller than the corresponding signal of the pristine film (17 cycles). Hence, this is in excellent agreement, meaning that there is no significant difference in the relative amount of TiO2 deposited in the O2 plasma treated and the pristine sample. In contrast, the XRF data show that there is only half as much TiO2 deposited inside the He/H2 plasma treated film after pore filling than in the pristine film with an equal total pore volume. Using spectroscopic ellipsometry, the refractive indices of the porous low-k films are measured. They are shown in Figure 3 as

Figure 2. In situ XRF measurements during ALD of TiO2 on a planar SiO2 reference substrate and on the three surface modified nanoporous low-k films. The Ti Kα (4.5 keV) peak area (plotted on the y-axis) is proportional to the number of Ti atoms deposited.

measured on the three different porous films show a similar evolution. During the first ALD cycles, the intensity increases much faster in the porous films than on a planar, nonporous Si reference substrate, proving that TiO2 is deposited on the internal surface of the porous thin films.14 With progressing growth, the growth rate transitions to a lower value and becomes more similar to the growth rate on the planar reference sample. This suggests that the pores are no longer accessible for the TDMAT precursor and ALD continues on top of the coated porous films. The first part of the XRF intensity curves reveals that there is an inhibited initial growth in the pristine and He/H2 plasma treated films, whereas unhindered growth is observed in the O2 plasma treated film.

Figure 3. Refractive index change of the three surface modified low-k films with increasing number of ALD cycles, as measured with spectroscopic ellipsometry.

a function of the number of cycles. In general, the refractive indices of the films increase with the number of ALD cycles because the air in a porous film is gradually being replaced by a higher refractive index material (TiO2). Ellipsometric porosimetry allows us to monitor the change of refractive index during adsorption and desorption of toluene inside the pores of a low-k film. From this data, the accessible porosity of the porous low-k film is derived.22 As represented in Figure 4, the 12286

dx.doi.org/10.1021/la4027738 | Langmuir 2013, 29, 12284−12289

Langmuir

Article

Figure 4. Porosity change with increasing number of ALD cycles in the three surface modified nanoporous low-k films, as obtained from ellipsometric porosimetry measurements.

Figure 5. RBS measurements after 60 cycles of TiO2 deposition in the three surface modified nanoporous low-k films.

O2 plasma treated film becomes inaccessible to the probe molecule toluene after 8 cycles of TiO2 deposition. This is in agreement with the XRF data showing that TDMAT is no longer deposited inside the O2 plasma treated film after 8 cycles. On the other hand, for the He/H2 plasma treated film, EP indicates a reduction of the porosity to half of its original value after 8 cycles of TiO2 deposition. With the XRF results also taken into account, it is concluded that only half of the original pore volume is filled with TiO2, leaving a remaining porosity that is accessible by toluene but not by TDMAT. In order to get more insight into the depth profile of the TiO2 deposited in the porous low-k films, RBS measurements were performed under a glancing exit angle on all three substrates. A numerical representation of the fitted data is given in Table 2, while the RBS spectra are shown in Figure 5. The

the RBS results show that the bottom part of the film is hardly modified.6,23 This indicates that the H atoms themselves do not modify the low-k material, which is in agreement with literature.12 Consequently, the modification induced by the He/H2 plasma treatment comes from the VUV light of the He plasma. Since the depth of plasma modification is limited by the penetration depth of the photon fluence, our experimental conditions suggest that the depth of modification by VUV photons was about 40−50 nm, half of the film thickness.24 Since the initial ALD growth rate is higher on the Si-OH surface groups than on the Si-CH3 groups, the pores in the top part of the film get filled faster with TiO2 than those in the bottom part until the TDMAT can no longer pass through the narrowed pores in the top layer, thus leaving the bottom layer porous. This also explains the EP result indicating a remaining porosity equal to half of its original value after closure of the pores to TDMAT (8 cycles) and the XRF result indicating that only half as much TiO2 was deposited in the He/H2 plasma treated film as in the pristine film. Quantitatively the 50% value is in excellent agreement with the relative amount of Si-OH surface groups in the He/H2 plasma treated film as calculated from the FTIR data. In general, this result illustrates that ALD in combination with RBS depth profiling is a handy approach to analyze the plasma modification profile in porous thin films. A schematic summary of the proposed behavior of TiO2 ALD growth in the three films is depicted in Figure 6. ALD growth in the O2 plasma treated film proceeds readily from the start and TiO2 is deposited uniformly throughout the entire thickness of the porous layer. In the pristine film the deposited

Table 2. RBS Results after 60 Cycles of TiO2 Deposition in the Three Surface Modified Nanoporous Low-k Filmsa Type of low-k film Pristine SiCOH low-k He/H2 plasma treated O2 plasma treated

Areal density of Ti atoms (1015 at/ cm2)

Fraction of surface Ti (%)

Fraction of Ti in layer n-1 (%)

Fraction of Ti in layer n-2 (%)

Fraction of Ti in layer n-3 (%)

56.37

33

27

23

17

34.88

50

27

17

7

44.43

36

26

20

18

a

Measurements were performed under a glancing exit angle to improve depth resolution. A 4-layer model is used to extract depth information. Layer n-1 is closest to the surface, while n-3 is closest to the substrate.

results show that for the pristine and O2 plasma treated films TiO2 is uniformly deposited throughout the entire film, while for the He/H2 plasma treated film a gradient is observed with a high TiO2 concentration on the top surface and a fast decrease in concentration when going deeper inside the film. This indicates that the O2 plasma treatment resulted in uniformly distributed Si-OH surface groups throughout the entire film, while the He/H2 plasma treatment was more confined to the top part of the porous film resulting in a hydrophilic (Si-OH rich) top layer and a hydrophobic (Si-CH3 rich) bottom layer in the low-k film. Although the lifetime of H atoms is more than sufficient to penetrate into the film from the top to the bottom,

Figure 6. Proposed behavior of TiO2 ALD growth (red) in the three types of surface modified nanoporous low-k films having hydrophilic (green) and/or hydrophobic (blue) internal surfaces. The representation of porosity is not on scale with the film thickness. 12287

dx.doi.org/10.1021/la4027738 | Langmuir 2013, 29, 12284−12289

Langmuir

■

layer is also spread uniformly throughout the internal surface, but the initial growth is inhibited by the less reactive Si-CH3 surface groups. This means that the pore filling mechanism is delayed. In the He/H2 plasma treated film, TiO2 deposition takes place mainly on the hydrophilic top part and only a small amount is deposited in the hydrophobic bottom part. Consequently, the pores in this top part are narrowed fast and no more deposition inside the porous film is possible. This research demonstrates that in situ XRF in combination with EP can be used as an analytical method to investigate the pore narrowing and sealing properties of ALD treated nanoporous films and that ALD combined with RBS analysis can provide in-depth information on the modifications induced by a plasma treatment on a nanoporous thin film. Although the data show that TDMAT is not the most selective probe molecule for these surface groups, the principle of this method is clearly demonstrated. Because plasma technology is frequently used in the low-k industry, the presented approach can contribute to a better understanding of the modifications and the plasma damage profile in a low-k film.

REFERENCES

(1) Maex, K.; Baklanov, M. R.; Shamiryan, D.; Iacopi, F.; Brongersma, S. H.; Yanovitskaya, Z. S. Low dielectric constant materials for microelectronics. J. Appl. Phys. 2003, 11, 8793−8841. (2) Hatton, B. D.; Landskron, K.; Hunks, W. J.; Bennett, M. R.; Shukaris, D.; Perovic, D. D.; Ozin, G. A. Materials Chemistry for low-k Materials. Mater. Today 2006, 9, 22−31. (3) Shearn, M.; Sun, X.; Henry, M. D.; Yariv, A.; Scherer, A. Semiconductor Technologies, Advanced plasma processing: etching, deposition, and wafer bonding techniques for semiconductor applications, Grym, J. , Ed.; InTech, 2010; ISBN: 978−953−307−080−3; DOI: 10.5772/8564. (4) Baklanov, M. R.; Maex, K. Porous low dielectric constant materials for microelectronics. Phil. Trans. R. Soc. A 2006, 364, 201− 215. (5) Lee, J.; Graves, D. B. Roles of plasma-generated vacuumultraviolet photons and oxygen radicals in damaging nanoporous low-k films. J. Vac. Sci. Technol., A 2013, 31, 041302−1−11. (6) Braginsky, O. V.; Kovalev, A. S.; Lopaev, D. V.; Malykhin, E. M.; Mankelevich, Y. A.; Rakhimova, T. V.; Rakhimov, A. T.; Vasilieva, A. N.; Zyryanov, S. M.; Baklanov, M. R. The mechanism of low-k SiOCH film modification by oxygen atoms. J. Appl. Phys. 2010, 108, 073303− 1−10. (7) Baklanov, M. R.; De Marneffe, J.-F.; Shamiryan, D.; Urbanowicz, A. M.; Shi, H.; Rakhimova, T. V.; Huang, H.; Ho, P. S. Plasma processing of low-k dielectrics. J. Appl. Phys. 2013, 113, 041101. (8) Worsley, M. A.; Bent, S. F.; Gates, S. M.; Fuller, N. C. M.; Volksen, W.; Steen, M.; Dalton, T. Effect of plasma interactions with low-k films as a function of porosity, plasma chemistry, and temperature. J. Vac. Sci. Technol., B 2005, 23, 395−405. (9) Iacopi, F.; Choi, J. H.; Terashima, K.; Rice, P. M.; Dubois, G. Cryogenic plasmas for controlled processing of nanoporous materials. Phys. Chem. Chem. Phys. 2011, 9, 3634−3637. (10) Chang, C. C.; Pan, F.-M.; Chen, C.-W. Pore sealing of mesoporous silica low-k dielectrics by oxygen and argon plasma treatments. Microelectron. Eng. 2009, 86, 2241−2246. (11) George, S. M. Atomic Layer Deposition: An Overview. Chem. Rev. 2010, 110, 111−131. (12) Urbanowicz, A. M.; Vanstreels, K.; Verdonck, P.; Shamiryan, D.; De Gendt, S.; Baklanov, M. R. Improving mechanical robustness of ultralow-k SiOCH plasma enhanced chemical vapor deposition glasses by controlled porogen decomposition prior to UV-hardening. J. Appl. Phys. 2010, 107, 104122. (13) Kunnen, E.; Baklanov, M. R.; Franquet, A.; Shamiryan, D.; Rakhimova, T. V. Effect of energetic ions on plasma damage of porous SiCOH low-k materials. J. Vac. Sci. Technol., B 2010, 28, 450−459. (14) Dendooven, J.; Sree, S. P.; De Keyser, K.; Deduytsche, D.; Martens, J. A.; Ludwig, K. F.; Detavernier, C. In Situ X-ray Fluorescence Measurements During Atomic Layer Deposition: Nucleation and Growth of TiO2 on Planar Substrates and in Nanoporous Films. J. Phys. Chem. 2011, 115, 6605−6610. (15) Detavernier, C.; Dendooven, J.; Sree, S. P.; Ludwig, K. F.; Martens, J. A. Tailoring nanoporous materials by atomic layer deposition. Chem. Soc. Rev. 2011, 40 (11), 5242−5253. (16) Devloo-Casier, K.; Dendooven, J.; Ludwig, K. F.; Lekens, G.; D’Haen, J.; Detavernier, C. In situ Synchrotron based X-ray Fluorescence and Scattering Measurements during ALD. Appl. Phys. Lett. 2011, 98, 231905. (17) Dultsev, F. N.; Baklanov, M. R. Nondestructive Determination of Pore Size Distribution in Thin Films Deposited on Solid Substrates. Electrochem. Solid-State Lett. 1999, 2, 192. (18) Baklanov, M. R.; Mogilnikov, K. P.; Polovinkin, V. G.; Dultsev, F. N. Determination of pore size distribution in thin films by ellipsometric Porosimetry. J. Vac. Sci. Technol., B 2000, 18, 1385. (19) Lakomaa, E.-L.; Root, A.; Suntola, T. Surface reactions in Al2O3 growth from trimethylaluminium and water by atomic layer epitaxy. Appl. Surf. Sci. 1996, 107, 107.

■

CONCLUSIONS The effect of a saturating O2 and an unsaturating He/H2 plasma treatment on porous SiCOH low-k films has been investigated. It was seen that both plasmas induced hydrophilic groups on the surface of the originally hydrophobic porous film. Next, the behavior of TiO2 ALD growth in the surface modified films was investigated, leading to a model for the pore filling mechanism. In situ synchrotron XRF measurements showed that the initial growth of TiO2 was inhibited in a film with a hydrophobic internal surface, while the growth was unhindered in a film with a hydrophilic internal surface. In the He/H2 plasma treated porous film that had a hydrophilic top layer and a hydrophobic bulk, TiO2 grew faster in the hydrophilic part, leading to faster reduction of the pore size in the top and remaining porosity in the bulk. These results can be of interest to the low-k community that is looking for sealing the bulk dielectric from its environment without losing porosity. Alternatively, the sensitivity of the initial stages of TiO2 ALD growth to the underlying surface groups also enables the use of ALD as a probe tool to study the extent of damage induced by the plasma treatments. RBS depth profiling of TiO2 in a surface modified low-k film provides insights into the plasma modification profile in the nanoporous film, as demonstrated here for the He/H2 plasma treated film.

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS C.D. is grateful for financial support to the European Research Council through an ERC Starting Grant (Grant No. 239865), the Flemish FWO and the Special Research Fund BOF of Ghent University (GOA project). K.F.L. acknowledges the DOE for funding (Grant No. DE-FG02-03ER46037). Use of the National Synchrotron Light Source, Brookhaven National Laboratory, was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC02-98CH10886. 12288

dx.doi.org/10.1021/la4027738 | Langmuir 2013, 29, 12284−12289

Langmuir

Article

(20) Elam, J. W.; Libera, J. A.; Huynh, T. H.; Feng, H.; Pellin, M. J. Atomic Layer Deposition of Aluminum Oxide in Mesoporous Silica Gel. J. Phys. Chem. C 2010, 114, 17286−17292. (21) Kobayashi, N. P.; Donley, C. L.; Wang, S.-Y.; Williams, R. S. Atomic Layer Deposition of Aluminum Oxide on Hydrophobic and Hydrophilic Surfaces. J. Cryst. Growth 2007, 299, 218−222. (22) Dendooven, J.; Devloo-Casier, K.; Van Hove, R.; Levrau, E.; Detavernier, C. Ellipsometric porosimetry for the characterization of coatings grown by atomic layer deposition in nanoporous thin films. Langmuir 2012, 28, 3852−3859. (23) Rakhimova, T. V.; Braginsky, O. V.; Kovalev, A. S.; Lopaev, D. V.; Mankelevich, Y. A.; Malykhin, E. M.; Rakhimov, A. T.; Vasilieva, A. N.; Zyryanov, S. M.; Baklanov, M. R. Recombination of O and H atoms on the surface of nanoporous dielectrics. IEEE Trans. Plasma Sci. 2009, 37, 1697−1704. (24) akhimova, T. V.; Rakhimov, A. T.; Mankelevich, Yu. A.; Lopaev, D. V.; Kovalev, A. S.; Vasil’eva, A. N.; Proshina, O. V.; Braginsky, O. V.; Zyryanov, S. M.; Kurchikov, K.; Novikova, N. N.; Baklanov, M. R. Modification of organosilicate glasses low-k films under extreme and vacuum ultraviolet radiation. Appl. Phys. Lett. 2013, 102, 111902.

12289

dx.doi.org/10.1021/la4027738 | Langmuir 2013, 29, 12284−12289