Role of the Reducing Agent in the Electroless ... - ACS Publications

Apr 25, 2017 - Role of the Reducing Agent in the Electroless Deposition of Copper on Functionalized SAMs. Ashley A. Ellsworth. † and Amy V. Walker*,...
0 downloads 0 Views 4MB Size
Download PDF
Article pubs.acs.org/Langmuir

Role of the Reducing Agent in the Electroless Deposition of Copper on Functionalized SAMs Ashley A. Ellsworth† and Amy V. Walker*,†,‡ †

Department of Chemistry and Biochemistry and ‡Department of Materials Science and Engineering, University of Texas at Dallas, Richardson, Texas 75080, United States S Supporting Information *

ABSTRACT: Metallized organic layer constructs have a wide range of technological applications. Electroless deposition is an attractive technique by which to deposit metal overlayers because it is inexpensive and can be performed at low temperatures, compatible with organic materials. Amine borane reducing agents are versatile and are capable of depositing metals, semiconductors, and even insulators. We have investigated the role of amine borane reducing agents in the electroless deposition of copper on −CH3-, −OH-, and −COOH-terminated SAMs adsorbed on gold using time-offlight secondary ion mass spectrometry, optical microscopy, and complementary MP2 calculations. Three reducing agents were studied: amine borane, dimethylamine borane, and trimethylamine borane. At pH >9, −COOH-terminated SAMs form coppercarboxylate complexes, which serve as nucleation sites for subsequent copper deposition. The rate of copper deposition is dependent on the strength of the B−N bond of the amine borane reducing agent. Similarly, if the terminal group is nonpolar such as a −CH3 functionality, then the rate of copper deposition is dependent on the amine borane B−N bond strength. However, in contrast to −COOH-terminated SAMs, copper deposition does not begin immediately. If the terminal group contains polar bonds, such as the C−OH bond of −OH-terminated SAMs, deposition is dominated by the interaction of the reducing agent with the terminal group rather than the relative bond strengths of the amine borane reducing agents.

1. INTRODUCTION Metallized organic layers, such as polymers and self-assembled monolayers (SAMs), have many important technological applications including energy harvesting,1,2 sensing,3 and electronics.4−7 Electroless deposition (ELD) on organic surfaces is attractive for the deposition of metals, semiconductors, and even insulators. In ELD, the substrate is simply immersed in a deposition bath that contains REDOX agents and additives, which help control the deposition process and influence the properties of the resulting deposit. ELD processes have many advantages. First, they are solution-phase processes and so do not require vacuum equipment.8,9 Second, they can be performed at low substrate temperatures, which are compatible with organic surfaces.10−15 Third, unlike electroplating, ELD involves REDOX reactions and so does not require the application of an external potential, making it compatible with insulating materials such as polymers.8,9 Finally, ELD is a soft deposition technique and can prevent metal penetration through organic layers for low- to mediumreactivity metals.11,12 Although electroless deposition is widely used in industrial processes for the metallizaton of organic substrates,8,9 there have been few studies of the reaction pathways of ELD on organic substrates. Early studies demonstrated that stable, adherent metallic overlayers are produced if stable metal− terminal group complexes are formed,10−14 so the chemical © XXXX American Chemical Society

identity and position of the terminal group are important. For example, Zangmeister and van Zee12 observed, using the reduction of Cu2+ ions by formaldehyde, that copper layers were formed on 4-mercaptobenzoic acid self-assembled monolayers (SAMs) but not on 3-mercaptobenzoic acid and octadecanethiolate SAMs. Later experiments using the same ELD process showed that small amounts of copper could be deposited on −CH3-terminated alkanethiolate SAMs adsorbed on Au.10,13 However, these studies also showed that copper pentrated through −CH3- and −COOH-terminated SAMs to the Au/S interface.10 To prevent copper penetration through the SAM and improve the deposition selectivity, Lu and Walker11 observed that by using a deposition temperature of 45 °C and a bath additive, adenine, strongly adherent Cu films could be selectively deposited on the −COOH-terminated areas of a patterned −COOH/−CH3-terminated SAM sample without Cu penetration to the Au/S interface. Furthermore, the use of adenine improved the morphology of the deposited copper layer. There have been even fewer studies of the effect of different reducing agents on ELD processes. It is well known that on Special Issue: Surfaces and Interfaces for Molecular Monitoring Received: January 11, 2017 Revised: April 14, 2017 Published: April 25, 2017 A

DOI: 10.1021/acs.langmuir.7b00103 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Hill, MA). Copper(II) sulfate pentahydrate (CuSO4·5H2O, 98+%), ethylenediaminetetraacetic acid (EDTA, 98%), dimethylamine borane complex (97%), ammonia borane complex (97%), trimethylamine borane complex (97%), hexadecanethiol (HDT, 99+%), and mercaptohexadecanoic acid (MHA, 99+%) were acquired from Sigma-Aldrich, Inc. (St. Louis, MO). 16-Hydroxy-1-hexadecanethiol (MHL) (99+%) was acquired from Frontier Scientific (Logan, UT). Anhydrous ethanol (A.C.S. grade) was purchased from Aaper Alcohol (Shelbyville, KY). Concentrated sulfuric acid (95%) was purchased from BDH Aristar, Inc. (Chester, PA). All reactants were used without further purification. Silicon wafers (⟨111⟩ orientation) were acquired from Addison Engineering Inc. (San Jose, CA) and cleaned using an RCA SC-1 etch (H2O/NH4OH/H2O2 = 5:1:1) for 20 min prior to use. 2.2. Preparation of Self-Assembled Monolayers. The preparation of self-assembled monolayers (SAMs) is well known.27,28 In brief, chromium (∼50 Å) and then gold (∼1000 Å) were sequentially thermally deposited onto clean Si wafers. Well-ordered SAMs were then formed by immersing the gold substrate into 1 mM ethanolic solutions of each alkanethiol for 24 h at ambient temperature (22 ± 1 °C). The samples were then rinsed with ethanol and dried using nitrogen gas. To ensure that the SAMs were free of significant chemical contamination, for each batch one sample was taken and characterized using single-wavelength ellipsometry (Gaertner Scientific Corp., Skokie, IL) and TOF SIMS. 2.3. Copper Electroless Deposition. The copper electroless deposition solution was composed of 0.032 M copper(II) sulfate pentahydrate, 0.24 M triethanolamine (TEA), 0.037 M EDTA, and 0.067 M amine borane reducing agent (amine borane, dimethylamine borane, or trimethylamine borane). The reported bath pH is the measured pH after it was adjusted using dilute sulfuric acid before the addition of the reducing agent. Normally, deposition was carried out at pH 9. To investigate the role of bath pH, depositions were also carried out at pH 6 and 12. All experiments were carried out at room temperature (22 ± 1 °C). Depositions were performed for periods of time ranging from 2 min to 15 h. After deposition, each substrate was rinsed with deionized water and ethanol and dried with N2 gas. The resulting films were immediately studied using time-of-flight secondary ion mass spectrometry (TOF SIMS) and optical microscopy. To check that the deposited film was metallic copper, X-ray photoelectron spectra were also obtained using each reaction condition (data not shown). 2.4. Time-of-Flight Secondary Ion Mass Spectrometry. TOF SIMS measurements were performed using an ION TOF IV spectrometer (ION TOF Inc., Chestnut Hill, NY) with a Bi liquid metal ion gun primary ion source. The instrument has three chambers for sample introduction, preparation, and analysis. The pressure in the analysis and preparation chambers is typically less than 5 × 10−9 mbar. The Bi+ primary ions had a kinetic energy of 25 keV. The probe beam diameter was ∼100 nm and rastered over a (500 × 500) μm2 area during data acquisition. The spectra were acquired using an ion dose of less than 1010 ions·cm−2, which is within the static regime.29 For each experiment, a minimum of three samples were made in three separate deposition baths (on separate days), and three separate areas on each sample were analyzed. The reported data therefore represents an average of at least nine measurements, and the error bars, the corresponding estimated standard deviation. 2.5. Optical Microscopy. Optical microscopy was performed using a Keyence VHX-2000 digital microscope. Dark-field images were obtained from representative samples with 200× magnification. The images shown are representative of the data obtained. 2.6. MP2 Calculations. For each reducing agent, MP2 calculations were used to determine the optimized geometries, the bond energies, the dipole moment, and the electronic charge on B and N. The MP2 approach was chosen because it has been shown that MP2 gives the best agreement in predicting B−N bond dissociation energies and the structures of amine boranes.30 The dipole moments and the electronic charges were calculated using natural bond order (NBO) population analyses. All structures reported in this article are minima; frequency calculations have been made to confirm this in every case. Calculations

inorganic substrates reducing agents change the bath stability as well as the deposited film properties.8,9,16,17 Recently, it has also been demonstrated that the interaction of the reducing agent with organic substrates is an important experimental parameter. Shi and Walker15 showed that the selective deposition of nickel via reduction of Ni2+ by dimethylamine borane could be achieved in two different ways. First, on −COOH/−CH3- and −COOH/−OH-patterned SAMs, deposition occurred on the −COOH-terminated areas as a result of the formation of Ni− carboxylate complexes, which acted as the nucleation sites for metal overlayer formation. However, surprisingly, on −OH/− CH3-patterned SAMs, deposition was observed on the −CH3terminated SAM areas. In this case, there is no Ni−surface complex formation but rather the deposition selectivity is due to the interaction of the reducing agent, dimethylamine borane, with the methyl and hydroxyl terminal groups. In later experiments, Walker and co-workers11,18 exploited this effect to deposit metallic nanostructures, including nanowires, nanochannels, and nanopores, on patterned SAM surfaces using a new technique, electroless nanowire deposition on micropatterned substrates (ENDOM). In this article, we systematically investigate the effect of amine borane reducing agents on Cu electroless deposition on −CH3-, −COOH-, and −OH-terminated SAMs. In ELD, amine borane reducing agents are very versatile8,9,16,19 and can be used to deposit metals including gold,9,20 nickel,9,15,16,19 silver,20 cobalt,9,20 palladium,16,21 iron,21 copper,9 alloys,9,22,23 semiconductors,24,25 and insulators.26 Other advantages of amine borane reducing agents include their ability to produce relatively pure metallic films,16 their wide range of operating conditions,17 and their reduced environmental impact.17 To investigate the effect of the amine borane structure and the deposition process, three reducing agents were employed: ammonia borane (AB), dimethylamine borane (DMAB), and trimethylamine borane (TMAB). We note that methylamine borane is also of interest but is not commercially available and so was not used in this study. Previous studies of these reducing agents on glass substrates indicate that the critical factor in the ELD process is the N−B bond strength of the amine borane.19 Our data show that the interaction of the amine borane reducing agent with the surface and the formation of surface complexes as well as the N−B bond strength of the amine borane affect the ELD process. On −CH3- and −COOHterminated SAMs, at pH 9 the major factor in the Cu ELD process is the strength of the N−B bond, which indicates that the first step in the reaction is the adsorption and decomposition of the amine borane. At pH 9, deposition starts immediately on −COOH-terminated SAMs because of the formation of Cu(II)−carboxylate complexes that act as the nuclei for copper overlayer formation. On −CH3-terminated SAMs, no Cu(II)−surface complexes form, so there is an induction time for copper deposition to begin. At pH 9 for −OH-terminated SAMs and at a lower deposition bath pH (pH 6) on −COOH-terminated SAMs, the deposition process is dominated by the interaction of the amine borane reducing agents with the −OH and −COOH terminal groups. In these cases, the ELD process is significantly slowed because the reducing agent is unable to adsorb on the SAM surface as a result of electrostatic repulsion.

2. EXPERIMENTAL SECTION 2.1. Materials. Gold (99.995%), chromium (99.995%), and triethanolamine (98+%) were obtained from Alfa Aesar, Inc. (Ward B

DOI: 10.1021/acs.langmuir.7b00103 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir were carried out using Gaussian 09.31 The basis set employed was the 6-311+G(2d,p) basis set, which incorporates polarization and diffuse functions.32−35

3. RESULTS AND DISCUSSION 3.1. Bare SAMs. The positive and negative secondary ion mass spectra of bare −CH3-, −OH-, and −COOH-terminated SAMs have been discussed in detail previously.28,36−38 Briefly, for all monolayers in the positive ion mass spectra, prominent features include (CH2)x(CH)y+, [AuS(CH2)x]H+, and S(CH2)x+. In addition, for −OH-terminated SAMs, ions of the form [(CH2)xCO]+ are observed, whereas in the −COOHterminated SAMs, [(CH2)xCO]+ and [(CH2)xCO2H]H+ ions are detected. The negative ion mass spectra contain a number of high-mass cluster ions including AuxSy−, Au2M−, and AuM2−, where M is the intact SAM adsorbate species. Fragment ions are also observed that are indicative of the SAM terminal group, such as (CH2)xCO2− in the −COOHterminated SAM spectra, as well as the backbone of the molecule, for example (CH2)xSy−. 3.2. Copper Deposition. For all reducing agents studied, copper was observed to deposit on −CH3-, −OH-, and −COOH-terminated SAMs. Table 1 summarizes the fragment

Figure 1. High-mass negative ion resolution mass spectra centered at m/z 163 after Cu electroless deposition on a −COOH-terminated SAM at pH 12, 9, and 6. Deposition at pH 6, 9, and 12 was carried out at 15 h, 25 min, and 75 min, respectively. The mass spectra are normalized to the intensity of [C10H11O2]− to make clear the changes in the mass spectra upon copper deposition.

complex formation are observed (Figure 1). Dark-field optical images also show differences in the morphology of the deposited film after Cu ELD at pH 6 and 12 and are consistent with the SIMS data (Supporting Information Figure S3). At pH 6, the deposited copper layer is nonuniform and composed of large islands. In contrast, at pH 12, the deposited copper layer is more uniform, which suggests that the formed copper− carboxylate complexes serve as nuclei for copper deposition. In contrast to −CH3- and −COOH-terminated SAMs, the mass spectra indicate that for −OH-terminated SAMs copper deposits atop the −OH-terminated SAM and also penetrates through the SAM to the Au/S interface. In the SIMS spectra, ions of the form CuxOy± and CuO(CH2)x+ are observed, indicating that copper has deposited at the SAM/vacuum interface, whereas ions of the form AuxCuySz− are also observed, indicating that copper has also penetrated to the Au/S interface. Further information about the deposited Cu can be obtained by an examination of the intensities of the molecular cluster ions, such as Au2M−, which involve the adsorbed intact SAM molecule (Figure 3). For −COOH-terminated SAMs, upon Cu ELD using DMAB and TMAB the attenuation of the Au2M− ion intensity is much faster than for −OH- and −CH3terminated SAMs, suggesting that a copper overlayer is forming, which blocks ion ejection from the substrate (Figure 2a). Using amine borane (AB), the intensity of the Au2M− ion increases and then decreases as the deposition continues, suggesting that the deposited layer is nonuniform. The increase in molecular ion intensity is likely due to electron transfer from the more electropositive copper to the electronegative Au and molecular clusters leaving the surface.38 Upon completion of a copper monolayer, the molecular ion intensity decreases and then increases again as another monolayer is deposited. For both −OH- (Figure 2b) and −CH3-terminated (Figure 2c) SAMs, initially the Au2M− ion intensity remains approximately constant or even increases, indicating that some of the SAM molecules are left chemically intact after Cu deposition and copper deposits as islands on the SAM surface. At longer deposition times, for −OH-terminated SAMs the intensity of Au2M− is approximately constant and ∼50% of its initial intensity (Figure 2b), indicating that some SAM molecules are never covered by deposited copper and suggesting that copper

Table 1. Fragment and Cluster Ions Observed in the TOF SIMS Spectra of −CH3-, −OH-, and −COOH-Terminated SAMs after Cu ELD SAM terminal group −CH3 −OH −COOH

fragment ions observed Cux+ (x = 1−3), Cu(CH2)x(CH)y±, Cu(CH2)x(CH3)y+ Cux+ (x = 1−3), AuxCuySz±, CuO(CH2)x+, Cux(OH)y±, CuxOy+ Cux+ (x = 1−3), CuCO(CH2)x+, Cux(OH)y±, Cu2COO(CH2)x(CH)y±

and cluster ions observed in the TOF SIMS spectra upon Cu ELD. The data indicate that for −CH3- and −COOHterminated SAMs that copper is deposited atop the SAM. We observed methyl-terminated ions of the forms Cu(CH2)x(CHy)± and Cu(CH2)x(CH3)y+, which increase with deposition time, but no AuxCuySz± ions, which are indicative of metal atom penetration through the SAM (Supporting Information Figure S1). We note that this is in contrast to previous studies using physical vapor deposition39 and electroless deposition using formaldehyde as a reducing agent,10 which showed that deposited copper penetrated through −CH3-terminated SAMs to the Au/S interface. In agreement with previous experiments, −COOH-terminated SAMs upon Cu deposition in the mass spectra ions of CuxCOO(CH2)x(CH)y± and CuCO(CH2)x+ are found, indicating that the copper has interacted with the −COOH terminal group. Because the pKa of mercaptohexadecanoic acid is 8 to 9,40 under our reaction conditions it is likely that the −COOH terminal group is deprotonated and these ions are indicative of copper−carboxylate complex formation, which serve as the nucleation sites for copper deposition. To test this hypothesis, copper ELD was carried out at pH 6 and 12. At pH 6, the −COOH terminal group is almost completely protonated, and in agreement with our hypothesis, we observe lower ion intensities of CuxCOO(CH2)x(CH)y± and CuCO(CH2)x+ (Figure 1 and Supporting Information Figure S2). In contrast and also in agreement with our hypothesis, at pH 12, higher intensities of ions indicative of copper−carboxylate C

DOI: 10.1021/acs.langmuir.7b00103 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure S5). For copper deposition using ammonia borane (AB), on −COOH- and −CH3-terminated SAMs the films appear to be nonuniform even at the earliest deposition times (Figure 3, left column, and Supporting Information Figures S4 and S5). Thus, on −CH3- and −COOH-terminated SAMs the molecular ion does not decrease as quickly using AB as a reducing agent (Figure 2a,c). On −OH-terminated SAMs, copper deposition appears to be nonuniform; large copper islands are observed using all reducing agents and at all deposition times studied (Figure 3 and Supporting Information Figure S6). 3.3. Variation of Copper Deposition with Reducing Agent Identity. The data clearly indicate that copper ELD on functionalized SAMs is dependent on both the SAM terminal group and the chemical identity of the reducing agent. To try to quantify the amount of copper deposited using AB, DMAB, and TMAB, we examined the intensities of the Cu+ signals with deposition time (i.e., increasing Cu deposition). Figure 4 displays the ratio of the 63Cu+ ion intensity (peak area) to C4H9+ (m/z 57), a characteristic ion of the alkanethiolate SAM backbone, with deposition time at 22 °C and pH 9. We note that secondary ion intensities are nonlinear and dependent on the matrix surrounding the structure of interest, so quantitation is relatively difficult.29,39 We therefore employed the ion intensity of C4H9+, which is a characteristic ion of the alkanethiolate SAM backbone, as an internal standard to perform a relative quantitation of the amount of Cu deposited.15 On −COOH-terminated SAMs, copper deposition begins immediately upon immersion of the sample into the deposition bath: the relative ion intensity of 63Cu+ immediately starts to rise, reaching a plateau after ∼10 min of deposition time (Figure 4a). At deposition times longer than 5 min, the deposition rate appears to slow down. It is also interesting to note that the amount of copper deposited using AB and DMAB reducing agents as judged by the 63Cu+ to C4H9+ ratio is approximately 3 times larger than for TMAB. In contrast, on −CH3- and −OH-terminated SAMs, there is a time delay before the ratio of 63Cu+ to C4H9+, and copper is observed to deposit. For −CH3-terminated SAMs, some copper is observed to deposit in the first 5 min of deposition (Figure 4c) whereas on −OH-terminated SAMs copper deposition appears to start slightly later (5−10 min for AB and DMAB and 10−15 min for TMAB) (Figure 4b). After ∼20 min and ∼15 min on methyland hydroxyl-terminated SAMs, respectively, the deposition rate appears to slow toward a steady state. Similar to −COOHterminated SAMs, on −CH3-terminated SAMs the amount of copper is approximately 3 times larger than for TMAB. On −OH-terminated SAMs, a similar amount of copper is deposited using AB, TMAB, and DMAB as indicated by the ratio of 63Cu+ to C4H9+. The initial fast increase in the amount of copper deposited on the functionalized SAMs followed by a slowing of the deposition rate indicates that Cu electroless deposition using amine borane reducing agents is an autocatalytic reaction, in agreement with previous studies on inorganic substrates.9,16,19 3.4. Reaction Pathways in the Electroless Deposition of Copper on Functionalized SAMs. The data show that copper electroless deposition using amine borane reducing agents is similar on −COOH- and −CH3-terminated SAMs; for these SAMs, a uniform overlayer appears to form at long deposition times using DMAB and TMAB (Figure 4). Furthermore, the amount of copper deposited, as evaluated

Figure 2. Integrated ion intensities (peak areas) of Au2M− with deposition time for (a) −COOH- (M = −S(CH2)15COOH m/z 681), (b) −OH- (M = −S(CH2)15CH2OH, m/z 667), and (c) −CH3terminated (M = −S(CH2)15CH3 m/z 651) SAMs. The reducing agents employed were ammonia borane (AB), dimethylamine borane (DMAB), and trimethlamine borane (TMAB).

deposits in islands. For methyl-terminated SAMs, at longer deposition times the intensity of the Au2M− ion significantly decreases, indicating that a copper overlayer forms (Figure 2c). However, unlike −COOH-terminated SAMs the Au2M− ion intensity does not decrease to within the noise level of the mass spectrum, suggesting that some −CH3-terminated SAM molecules remain intact. Dark-field optical images are consistent with the intensities of the molecular cluster ions (Figure 3 and Supporting Information Figures S4−S6). At short deposition times on −CH3-terminated SAMs, islands are formed using DMAB and TMAB (Supporting Information Figure S4). In contrast, at longer times, the film appears to be quite uniform (Figure 3, top row: middle and right columns). Using DMAB and TMAB, the deposited film appears to be more uniform on −COOHterminated SAMs for all deposition times studied than on −CH3-terminated SAMs (Figure 3 and Supporting Information D

DOI: 10.1021/acs.langmuir.7b00103 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 3. Dark-field optical images of the deposited copper layer after 25 min of Cu ELD using ammonia borane (AB), dimethylamine borane (DMAB), and trimethlamine borane (TMAB) on −CH3-, −OH-, and −COOH-terminated alkanethiolate SAMs. Deposition conditions: pH 9 and 22 °C.

by the 63Cu+ to C4H9+ ratio, using AB and DMAB is ∼3 times greater than by employing TMAB as the reducing agent. However, copper deposition starts immediately on −COOHterminated SAMs, but there is an induction time for Cu deposition to begin on −CH3-terminated SAMs (Figure 4). In contrast, on −OH-terminated SAMs copper deposition is nonuniform (Figure 3), some deposited copper penetrates through the SAM to the Au/S interface, and there is little difference in the amount of copper deposited using the reducing agents studied (Figure 4b). Furthermore, it takes slightly longer for Cu deposition to begin on −OH-terminated SAMs than on −CH3-terminated SAMs. At first glance, these observations are surprising for two reasons. First, one would expect copper would be faster on the hydrophilic −OHterminated SAM than on the hydrophobic −CH3-terminated SAM. Second, in contrast to previous studies of physical vapor deposition and Cu ELD using formaldehyde, copper does not penetrate through the methyl-terminated SAM to the Au/S interface.10,39 Thus, any proposed reaction mechanism for copper electroless deposition using amine borane reducing agents needs to take into account both the chemistry of the surface and the identity of the reducing agent. Although the exact reaction mechanism for copper electroless deposition using amine borane reducing agents is not known, it can be described by the following chemical reaction:9

(CH3)3 − x NHxBH3(aq) → (CH3)3 − x NHxBH3(ads) + H+ → (CH3)x NHx +(aq) + BH3(ads)

(2)

From the above equations, it can be seen that the Cu deposition rate will be dependent on the strength of the N−B bond because it determines the extent of amine borane dissociation on the surface (eq 2). The strength of the B−N bond is dependent in part by the degree of the lone electron pair of the N atom shared with the B atom (i.e., the polarity of the amine borane). As the number of methyl groups in the amine borane increases, the electron density on the N atom facilitates electron donation to the borane and the strengthening of the B−N bond. Thus, using TMAB, the electroless deposition rate is predicted to be the lowest. The relative charge on the borane (BH3) fragment, which is a Lewis acid, also affects its interaction with and adsorption on functionalized SAMs. It is expected that the ELD reaction is slower if there is repulsion between the slightly negatively charged BH3 fragment and the surface. For example, if the terminal bond possesses polar bonds, such as the C−OH bond of hydroxyl-terminated SAMs, then the O atom is slightly negatively charged (δ−) and so the electroless deposition reaction is slowed.15 To investigate further the effect of the amine borane reducing agent on copper electroless deposition, MP2 calculations were performed (Table 2). In agreement with previous MO calculations,19 the polarity and B−N bond strength increase with the number of methyl substituents in the amine borane. Furthermore, the B−N bond strengths and dipole moments are in good agreement with reported experimental data.30,41−43 Using NBO, the charge on the boron atom also increases with the number of methyl groups of the amine borane. In agreement with our experimental

3Cu 2 +(aq) + (CH3)3 − x NHxBH3(aq) + 3H 2O(l) → 3Cu(s) + (CH3)x NHx +(aq) + H3BO3(aq) + 5H+(aq) (1)

It is thought that the initial reaction is the adsorption of the reducing agent to the surface followed by cleavage of the N−B bond:9,16,19 E

DOI: 10.1021/acs.langmuir.7b00103 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

induction time for copper deposition appears to be slightly longer on −CH3- and −OH-terminated SAMs than when using DMAB, which in turn is longer than for AB (Figure 4). This can be attributed to the relative strengths of the B−N bonds; for TMAB, larger deposition nuclei (catalytic centers) are required to catalyze the dissociation of TMAB, so a longer induction time is required for the deposition of these nuclei. Hydroxyl-terminated SAMs have the least amount of copper deposited as evaluated by the 63Cu+ to C4H9+ ratio (Figure 4). The difference between deposition on the −OH-terminated SAMs and the −CH3-terminated SAMs may be explained by the adsorption of the BH3 fragment on these surfaces. (We note that deposition on −COOH-terminated SAMs at pH 9 is dominated by the formation of copper−carboxylate complexes that act as the deposition nuclei, so the deposition mechanism is initially different.) The MP2 calculations show that the BH3 fragment has a slight negative charge except for TMAB, which has almost no charge (Table 2). The C−OH bond of the −OH-terminated SAM is covalent with the electrons displaced toward the hydroxyl group, and it has a partial negative charge (δ−). In contrast, for −CH3-terminated SAMs the C−H bond is not polar and has no partial charge. Consequently, BH3 is repelled by the −OH-terminated SAM surface, and the BH3 fragment is able to easily adsorb on the −CH3-terminated SAM surface. Thus, the calculations suggest that less copper is deposited on −OH-terminated SAMs than on −CH3terminated SAMs, in agreement with the experimental data. We further investigated the role of surface charge on Cu ELD using −COOH- and −CH3-terminated SAMs. Electroless deposition was performed at pH 6 and 12 on −CH3- and −COOH-terminated SAMs using dimethylamine borane. Figure 5 shows the amount of copper deposited, as measured by the ratio of 63Cu+ to C4H9+ ion intensities, with deposition time. We note that at pH 6, copper deposition is much slower than at pH 9 or 12. At high pH, by Le Chatelier’s principle the Cu ELD equilibrium is moved to the product side (eq 1), so more copper is deposited. Although a shift in equilibrium does not necessarily imply an increase in reaction rates, it is not surprising that the deposition process appears to be faster at a higher pH. At pH 6, deposition begins later on −COOHterminated SAMs than on −CH3-terminated SAMs (Figure 5a). Because the pKa of the −COOH-terminated SAM is 8 to 9,40 at pH 6 the carboxylic acid group is almost fully protonated (∼99.9% protonated), so the surface is polar with a small negative charge (δ−) on the O atoms. Thus, at pH 6, the Cu ELD process is dominated by the repulsion of the BH3 fragment by the polar −COOH-terminated SAM surface; copper deposition occurs earlier on −CH3-terminated SAMs because there is a higher concentration of BH3 at the SAM− solution interface. In contrast, at pH 12 more copper is deposited on −COOH-terminated SAMs at a given deposition time (Figure 5b). At pH 12, the carboxylic acid is almost fully

Figure 4. Variation of the 63Cu+ to C4H9+ ratio with deposition time on (a) −COOH-, (b) −OH-, and (c) −CH3-terminated SAMs at pH 9 and 22 °C. The reducing agents used are ammonia borane (AB), dimethylamine borane (DMAB), and trimethylamine borane (TMAB). The dashed lines are shown as guides to the eye.

observations (Figure 4), the calculations predict that the lowest amounts of copper are deposited using TMAB. This is because the B−N bond strength is the largest for TMAB and the first step in the reaction is the breaking of the B−N bond (eq 2). Furthermore, we also observe that using AB copper ELD is fastest (Figure 4). This is because the B−N bond is the weakest (Table 2), so the adsorption and decomposition of AB are relatively facile (eq 2). We also note that by using TMAB the

Table 2. MP2 Calculations of the Charge on N and B, the Dipole Moment, and the B−N Bond Energy of the Amine Boranes Used in This Study

a

reducing agent

charge on N

charge on B

dipole moment (D) (N → B)

B−N bond energy (kJ mol−1)

H3NBH3 CH3H2NBH3a (CH3)2HNBH3 (CH3)3NBH3

−0.857 −0.721 −0.612 −0.546

−0.011 −0.014 −0.012 0.001

5.50 5.46 5.34 5.15

122.9 152.3 168.2 173.5

Methylamine borane was included for completeness. F

DOI: 10.1021/acs.langmuir.7b00103 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

dissociation of TMAB, so a longer induction time is required for the deposition of these nuclei. If the terminal group contains polar bonds, such as the C−OH bond of −OHterminated SAMs, then the Cu ELD process is governed by the interaction of the reducing agent with the terminal group rather than the relative bond strengths of the amine borane reducing agents.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.7b00103. Variation of ion intensities of AuCuS2− and Cu(CH2)12(CH3)2+ with deposition time on −CH3terminated SAMs; high-mass negative ion resolution mass spectra centered at m/z 164 and dark-field optical images after Cu ELD on −COOH-terminated SAMs performed at pH 6 and 12; dark-field optical images of the deposited copper layer on −CH3-, −COOH-, and −OH-terminated SAMs using amine borane (AB), dimethylamine borane (DMAB), and trimethylamine borane (TMAB). (PDF)

■

Figure 5. Variation of the 63Cu+ to C4H9+ ratio with deposition time using DMAB on −COOH- and −CH3-terminated SAMs at (a) pH 6 and (b) pH 12. The dotted lines are guides for the eye.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Ph: +1 972 883 5780. Fax: +1 972 883 5725. ORCID

deprotonated (∼99.9% carboxylate), so copper−carboxylate complexes can form. Because these complexes serve as nucleation sites for copper deposition, more copper is deposited on the −COOH-terminated SAM than on the −CH3-terminated SAM. Taken together, the data indicate that the Cu ELD reaction is dominated by the adsorption and dissociation of the amine borane reducing agent if the metal cation does not form complexes with the surface. However, if metal−surface terminal group complexes can form, then these complexes serve as the nucleation sites for deposition, and deposition begins immediately.

Ashley A. Ellsworth: 0000-0001-8912-3645 Amy V. Walker: 0000-0003-2114-3644 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors gratefully acknowledge support from the National Science Foundation (CHE1213546). Nicole A. Ward, an undergraduate researcher, helped measure the amount of copper deposited using DMAB and AB on −COOHterminated SAMs, which is shown in Figure 4a.

4. CONCLUSIONS The data clearly show that the B−N bond strength of amine borane reducing agents and the chemistry of the organic surface strongly affect copper electroless deposition on functionalized SAMs. At high deposition bath pH, the Cu ELD process is determined by the relative strengths of the B−N bonds on carboxylic acid-terminated SAMs. At pH >9, −COOHterminated SAMs form copper−carboxylate complexes, which serve as nucleation sites for subsequent copper deposition. The rate of copper deposition is dependent on the strength of the B−N bond of the amine borane reducing agent with the least amount of copper deposited using TMAB, which has the strongest, largest B−N bond energy. Similarly, if the terminal group is nonpolar such as for −CH3-terminated SAMs then the rate of copper deposition is dependent on the amine borane B−N bond strength for all deposition bath pH values studied. However, in contrast to −COOH-terminated SAMs, copper deposition does not begin immediately with the longest induction time observed for TMAB. This can be attributed to the relative strengths of the B−N bonds; for TMAB, larger deposition nuclei (catalytic centers) are required to catalyze the

■

REFERENCES

(1) Gledhill, S. E.; Scott, B.; Gregg, B. A. Organic and NanoStructured Composite Photovoltaics: An Overview. J. Mater. Res. 2005, 20, 3167−3179. (2) Hsu, J. W. P.; Lloyd, M. T. Organic/Inorganic Hybrids for Solar Energy Generation. MRS Bull. 2010, 35, 422−428. (3) Favier, F.; Walter, E. C.; Zach, M. P.; Benter, T.; Penner, R. M. Hydrogen Sensors and Switches from Electrodeposited Palladium Mesowire Arrays. Science 2001, 293, 2227−2231. (4) Reed, M. A. Molecular-scale electronics. Proc. IEEE 1999, 87, 652−658. (5) Angelopoulos, M. Conducting polymers in microelectronics. IBM J. Res. Dev. 2001, 45, 57−75. (6) Kugler, T.; Logdlund, M.; Salaneck, W. R. Polymer surfaces and interfaces in light-emitting devices. IEEE J. Sel. Top. Quantum Electron. 1998, 4, 14−23. (7) Bharathan, J. M.; Yang, Y. Polymer/metal interfaces and the performance of polymer light-emitting diodes. J. Appl. Phys. 1998, 84, 3207−3211. (8) Schlesinger, M.; Paunovic, M. Modern Electroplating, 4th ed.; John Wiley & Sons: New York, 2000.

G

DOI: 10.1021/acs.langmuir.7b00103 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir (9) Mallory, G. O.; Hajdu, J. B. Electroless Plating: Fundamentals and Applications; Noyes Publications/William Andrew Publishing: Norwich, NY, 1990. (10) Lu, P.; Walker, A. V. Investigation of the Mechanism of Electroless Deposition of Copper on Functionalized Alkanethiolate Self-Assembled Monolayers Adsorbed on Gold. Langmuir 2007, 23, 12577−12582. (11) Lu, P.; Shi, Z.; Walker, A. V. Selective Electroless Deposition of Copper on Organic Thin Films with Improved Morphology. Langmuir 2011, 27, 13022−13028. (12) Zangmeister, C. D.; van Zee, R. D. Electroless Deposition of Copper onto 4-Mercaptobenzoic Acid Self-Assembled on Gold. Langmuir 2003, 19, 8065−8068. (13) Garno, J. C.; Zangmeister, C. D.; Batteas, J. D. Directed Electroless Growth of Metal Nanostructures on Patterned SelfAssembled Monolayers. Langmuir 2007, 23, 7874−7879. (14) Dressick, W. J.; Dulcey, C. S.; Georger, J. H., Jr.; Calvert, J. M. Photopatterning and Selective Electroless Metallization of SurfaceAttached Ligands. Chem. Mater. 1993, 5, 148−150. (15) Shi, Z.; Walker, A. V. Investigation of the Mechanism of Nickel Electroless Deposition on Functionalized Self-Assembled Monolayers. Langmuir 2011, 27, 6932−6939. (16) Lelental, M. Dimethylamine Borane as the Reducing Agent in Electroless Plating Systems. J. Electrochem. Soc. 1973, 120, 1650−1654. (17) Jagannathan, R.; Krishnan, M. Electroless plating of copper at a low pH level. IBM J. Res. Dev. 1993, 37, 117−123. (18) Shi, Z.; Walker, A. V. Synthesis of Nickel Nanowires via Electroless Deposition on Micropatterned Substrates. Langmuir 2011, 27, 11292−11295. (19) Lelental, M. Effect of amine borane structure on activity in electroless plating. J. Catal. 1974, 32, 429−433. (20) Watanabe, T. Film Formation Mechanism in Electrodeposition. Nano Plating: Microstructure Formation Theory of Plated Films and a Database of Plated Films; Elsevier: Oxford, 2004; pp 95−139. (21) Schmeckenbecher, A. F.; Lindholm, J. A. Potentiometric determination of dimethylamine borane and hypophosphite. Anal. Chem. 1967, 39, 1014−1016. (22) Yokoshima, T.; Nakamura, S.; Kaneko, D.; Osaka, T.; Takefusa, S.; Tanaka, A. Micropattern Formation for Magnetic Recording Head Using Electroless CoFeB Deposition. J. Electrochem. Soc. 2002, 149, C375−C382. (23) Yokoshima, T.; Kaneko, D.; Akahori, M.; Nam, H.-S.; Osaka, T. Electroless CoNiFeB soft magnetic thin films with high corrosion resistance. J. Electroanal. Chem. 2000, 491, 197−202. (24) Izaki, M.; Omi, T. Transparent Zinc Oxide Films Chemically Prepared from Aqueous Solution. J. Electrochem. Soc. 1997, 144, L3− L5. (25) Murase, K.; Tada, H.; Shinagawa, T.; Izaki, M.; Awakura, Y. QCM Studies of Chemical Solution Deposition of ZnO in Aqueous Media Containing Zinc Nitrate and Dimethylamineborane. J. Electrochem. Soc. 2006, 153, C735−C740. (26) Chigane, M.; Izaki, M.; Hatanaka, Y.; Shinagawa, T.; Ishikawa, M. Effect of dimethylamine borane concentration on antireflection properties of silica thin films via redox deposition. Thin Solid Films 2006, 515, 2513−2518. (27) Nuzzo, R. G.; Dubios, L. H.; Allara, D. L. Fundamental Studies of Microscopic Wetting on Organic Surfaces. 1. Formation and Structural Characterization of a Self-Consistent Series of Polyfunctional Organic Monolayers. J. Am. Chem. Soc. 1990, 112, 558−569. (28) Fisher, G. L.; Walker, A. V.; Hooper, A. E.; Tighe, T. B.; Bahnck, K. B.; Skriba, H. T.; Reinard, M. D.; Haynie, B. C.; Opila, R. L.; Winograd, N.; Allara, D. L. Bond Insertion, Complexation and Penetration Pathways of Vapor-Deposited Aluminum Atoms with HO- and CH3O- Terminated Organic Monolayers. J. Am. Chem. Soc. 2002, 124, 5528−5541. (29) Vickerman, J. C.; Briggs, D. TOF-SIMS: Materials Analysis by Mass Spectrometry, 2nd ed.; IM Publications LLP and SurfaceSpectra: Chichester, U.K., 2013.

(30) Gilbert, T. M. Tests of the MP2Model and Various DFT Models in Predicting the Structures and B−N Bond Dissociation Energies of Amine−Boranes (X3C)mH3‑mB−N(CH3)nH3‑n (X = H, F; m = 0−3; n = 0−3): Poor Performance of the B3LYP Approach for Dative B−N Bonds. J. Phys. Chem. A 2004, 108, 2550−2554. (31) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M. J.; Heyd, J.; Brothers, E. N.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A. P.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, N. J.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö .; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09; Gaussian, Inc.: Wallingford, CT, 2009. (32) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. Selfconsistent molecular orbital methods. XX. A basis set for correlated wave functions. J. Chem. Phys. 1980, 72, 650−654. (33) McLean, A. D.; Chandler, G. S. Contracted Gaussian basis sets for molecular calculations. I. Second row atoms, Z = 11−18. J. Chem. Phys. 1980, 72, 5639−5648. (34) Lee, C.; Yang, W.; Parr, R. G. Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785−789. (35) Becke, A. D. A new mixing of Hartree−Fock and local densityfunctional theories. J. Chem. Phys. 1993, 98, 1372−1377. (36) Fisher, G. L.; Hooper, A. E.; Opila, R. L.; Allara, D. L.; Winograd, N. The Interaction of Vapor-Deposited Al Atoms with CO2H Groups at the Surface of a Self-Assembled Alkanethiolate Monolayer on Gold. J. Phys. Chem. B 2000, 104, 3267−3273. (37) Fisher, G. L.; Hooper, A. E.; Opila, R. L.; Jung, D. R.; Allara, D. L.; Winograd, N. The Interaction Between Vapor-Deposited Al Atoms and Methylester-Terminated Self-Assembled Monolayers Studied by ToF-SIMS, XPS and IRS. J. Electron Spectrosc. Relat. Phenom. 1998, 98−99, 139−148. (38) Hooper, A.; Fisher, G. L.; Konstadinidis, K.; Jung, D.; Nguyen, H.; Opila, R.; Collins, R. W.; Winograd, N.; Allara, D. L. Chemical Effects of Methyl and Methyl Ester Groups on the Nucleation and Growth of Vapor-Deposited Aluminum Films. J. Am. Chem. Soc. 1999, 121, 8052−8064. (39) Nagy, G.; Walker, A. V. Dynamics of the Interaction of VaporDeposited Copper with Alkanethiolate Monolayers: Bond Insertion, Complexation, and Penetration Pathways. J. Phys. Chem. B 2006, 110, 12543−12554. (40) Kanicky, J. R.; Shah, D. O. Effect of Degree, Type, and Position of Unsaturation on the pKa of Long-Chain Fatty Acids. J. Colloid Interface Sci. 2002, 256, 201−207. (41) Haaland, A. Covalent versus Dative Bonds to Main Group Metals, a Useful Distinction. Angew. Chem., Int. Ed. Engl. 1989, 28, 992−1007. (42) Weaver, J. R.; Parry, R. W. Dipole Moment Studies. III. The Dipole Moments of the Methylamine Boranes. Inorg. Chem. 1966, 5, 713−718. (43) Durig, J. R.; Li, Y. S.; Odom, J. D. Microwave spectrum, structure and dipole moment of trimethylamine-borane. J. Mol. Struct. 1973, 16, 443−450.

H

DOI: 10.1021/acs.langmuir.7b00103 Langmuir XXXX, XXX, XXX−XXX