Superconformal filling composition and superconformally filling a recessed feature of an article
10889908 ยท 2021-01-12
Assignee
Inventors
Cpc classification
C25D5/605
CHEMISTRY; METALLURGY
C25D7/00
CHEMISTRY; METALLURGY
C25D5/18
CHEMISTRY; METALLURGY
C25D5/627
CHEMISTRY; METALLURGY
C25D5/10
CHEMISTRY; METALLURGY
C25D5/625
CHEMISTRY; METALLURGY
Y02E60/50
GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
C25D7/123
CHEMISTRY; METALLURGY
H01L21/2885
ELECTRICITY
H01L21/76879
ELECTRICITY
International classification
H01L21/768
ELECTRICITY
C25D5/10
CHEMISTRY; METALLURGY
C25D5/18
CHEMISTRY; METALLURGY
C25D7/00
CHEMISTRY; METALLURGY
Abstract
Superconformally filling a recessed feature includes: contacting the recessed feature with superconformal filling composition that includes: Au(SO.sub.3).sub.2.sup.3 anions; SO.sub.3.sup.2 anions; and Bi.sup.3+ cations; convectively transporting Au(SO.sub.3).sub.2.sup.3 and Bi.sup.3+ to the bottom member of the recessed feature; subjecting the recessed feature to an electrical current to superconformally deposit gold from the Au(SO.sub.3).sub.2.sup.3 on the bottom member relative to the sidewall and the field, the electrical current providing a cathodic voltage; and increasing the electrical current subjected to the field and the recessed feature to maintain the cathodic voltage between 0.85 V and 1.00 V relative to the SSE during superconformally depositing gold on the substrate to superconformally fill the recessed feature of the article with gold as a superconformal filling of gold, the superconformal filling being void-free and seam-free.
Claims
1. A process for superconformally filling a recessed feature of an article with gold, the process comprising: providing the article comprising: a substrate; a field disposed on the substrate; the recessed feature disposed on the substrate and surrounded by the field, the recessed feature comprising: a bottom member; a sidewall that separates the bottom member from the field, the recessed feature having an aspect ratio of a depth to a width from 0.5 to 100 before superconformally filling the recessed feature, the aspect ratio decreasing during superconformally filling the recessed feature; and an overlayer disposed on the article such that the field and the recessed feature are fully metallized for contact with a superconformal filling composition; contacting the field and the recessed feature with the superconformal filling composition in an absence of cyanide and lead, the superconformal filling composition having a near-neutral pH and comprising: a plurality of Au(SO.sub.3).sub.2.sup.3 anions as a source of gold for superconformally depositing gold in the recessed feature; a plurality of SO.sub.3.sup.2 anions; and a plurality of Bi.sup.3+ cations as a brightener and an accelerator for superconformally depositing gold in the recessed feature; convectively transporting the Au(SO.sub.3).sub.2.sup.3 anions and the Bi.sup.3+ cations to the bottom member by actively moving the substrate relative to the superconformal filling composition; subjecting the field and the recessed feature to an electrical current to superconformally deposit gold from the Au(SO.sub.3).sub.2.sup.3 anions on the bottom member relative to the sidewall and the field, the electrical current providing a cathodic voltage (V.sub.SSE) from 0.66 V to 1.00 V relative to a saturated sulfate electrode (SSE), and a first deposition ratio of a first deposition rate of gold on the bottom member relative to a second deposition rate of gold on the sidewall being from 1.5 to 10.sup.6; and increasing the electrical current subjected to the field and the recessed feature to maintain the V.sub.SSE from 0.66 V to 1.00 V relative to the SSE during superconformally depositing gold on the substrate to superconformally fill the recessed feature of the article with gold as a superconformal filling comprising gold, the superconformal filling being void-free and seam-free, such that in a presence of the superconformal filling composition: passivation of the field and the recessed feature occurs at the V.sub.SSE greater than 0.65 V relative to the SSE, sub-conformal deposition of gold occurs at the V.sub.SSE less than 1 V relative to the SSE, and superconformal deposition of gold occurs at the V.sub.SSE from 0.66 V to 1.00 V relative to the SSE.
2. The process of claim 1, further comprising changing the V.sub.SSE from a first potential to a more negative potential during superconformally depositing gold.
3. The process of claim 1, further comprising catalyzing superconformal deposition of gold with underpotential deposited Bi from the Bi.sup.3+ cations.
4. The process of claim 1, wherein actively moving the substrate relative to the superconformal filling composition comprising rotating the substrate at a rotation rate from 25 revolutions per minute (rpm) to 2000 rpm.
5. The process of claim 4, further comprising changing the rotation rate from a first rotation rate to a second rotation rate during superconformally depositing gold.
6. The process of claim 1, further comprising: maintaining the VSSE from 0.66 V to 1.00 V relative to the SSE until the recessed feature is partially filled with the aspect ratio being less than or equal to 0.5; and thereafter changing a deposition condition to fill the recessed feature sub-conformally, conformally, or a combination comprising at least one of the foregoing non-superconformal filling regimes.
7. The process of claim 1, wherein the V.sub.SSE is maintained from 0.66 V to 1.00 V relative to the SSE until the recessed feature is completely filled with the superconformal filling.
8. The process of claim 1, wherein the Au(SO.sub.3).sub.2.sup.3 anions are present in the superconformal filling composition in an amount from 50 millimolar (mM) to 350 mM.
9. The process of claim 1, wherein the SO.sub.3.sup.2 anions are present in the superconformal filling composition in an amount from 0.1 molar (M) to 1 M.
10. The process of claim 1, wherein Bi.sup.3+ cations are present in the superconformal filling composition in an amount from 1 micromolar (mmM) to 50 mmM.
11. The process of claim 1, wherein contacting the field and the recessed feature with the superconformal filling composition occurs in an absence of thallium.
12. The process of claim 1, wherein contacting the field and the recessed feature with the superconformal filling composition occurs in an absence of a suppressor.
13. The process of claim 1, wherein the superconformal filling composition consists essentially of the Au(SO.sub.3).sub.2.sup.3 anions, the SO.sub.3.sup.2 anions, the Bi.sup.3+ cations, and an additive in an aqueous liquid in an absence of a suppressor.
14. The process of claim 1, wherein the near-neutral pH of the superconformal filling composition is from 6.5 to 10.
15. The process of claim 1, wherein the superconformal filling is crystalline.
16. The process of claim 1, wherein the superconformal filling is nonporous.
17. The process of claim 1, wherein the superconformal filling further comprises cobalt, and the superconformal filling is a cobalt-gold alloy.
18. The process of claim 13, wherein the cobalt-gold alloy is a single phase.
19. The process of claim 1, wherein the superconformal filling is shiny and planar at an exposed surface of the superconformal filling on a submicron level, a brightness of the superconformal filling occurring in an absence of dendrites on the exposed surface.
20. The process of claim 1, wherein a second deposition ratio of the first deposition rate of gold on the bottom member relative to a third deposition rate of gold on the field is from 1.5 to 10.sup.6.
21. The process of claim 1, wherein superconformally filling the recessed feature is bottom-up filling.
22. The process of claim 17, wherein the field is passivated during bottom-up filling.
23. The process of claim 1, wherein the bottom-up filling is uniform over the entire bottom member.
24. The process of claim 1, wherein a deposition ratio of a thickness of gold deposited on the field to a thickness of gold deposited on the bottom member is from 1:20 to 1:50.
25. The process of claim 1, wherein depositing the gold on the field and the recessed feature automatically stops before completely filling the recessed feature with gold while the V.sub.SSE is from 0.66 V to 1.00 V relative to the SSE.
26. The process of claim 1, wherein the recessed feature comprises a trench, a via, a contact hole, or a combination comprising at least one of the foregoing recessed features.
27. The process of claim 1, wherein the overlayer comprises a gold seed crystal.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
(1) The following descriptions should not be considered limiting in any way. With reference to the accompanying drawings, like elements are numbered alike.
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
(21)
(22)
(23)
(24)
(25)
(26)
(27)
(28)
(29)
(30)
(31)
(32)
(33)
(34)
(35)
(36)
(37)
(38)
(39)
(40)
(41)
(42)
(43)
(44)
(45)
(46)
(47)
(48) 110
directions, the nanowires are inclined from the substrate surface by 54.7. Bottom left panel shows selected area electron diffraction pattern recorded from a single nanowire in the [1 1 0] zone-axis orientation. The bottom right panel shows a dark field image of an individual nanowire recorded with the 111-type reflection strongly excited by slight tilting off the [1 1 0] orientation;
(49)
(50)
(51)
(52)
(53)
(54)
(55)
(56)
(57)
DETAILED DESCRIPTION
(58) A detailed description of one or more embodiments is presented herein by way of exemplification and not limitation.
(59) It has been discovered that a process for superconformally filling a recessed feature of an article with gold unexpectedly fills trenches and vias with a gold superconformal filling from a superconformal filling composition. Beneficially, the superconformal filling can be an interconnect in microelectronics. In an aspect, the process fills the recessed feature from a bottom with the superconformal filling that is seam-free and void-free. Gold is deposited almost exclusively within the recessed feature such that little deposition occurs on the field relative to the bottom to minimize waste and reduce time spent on postdeposition processing. Surprisingly, the superconformal filling composition is an electrolyte that forms the superconformal filling in an absence of a lead additive or a polymer additive and is non-cyanide and nearly neutral. Superconformal deposition of gold can be performed at room temperature.
(60) In an embodiment, with reference to
(61) In an embodiment, with reference to
(62) The process also can include disposing overlayer 212 on article 200 such that field 204 and recessed feature 250 are fully metallized for contact with superconformal filling composition 214. Disposing overlayer 212 on article 200 can include evaporation, sputter deposition, chemical vapor deposition, or atomic layer deposition. In an embodiment, disposing overlayer 212 includes sputter deposition of a layer of titanium followed by a layer of gold.
(63) In the process, contacting field 204 and recessed feature 250 with superconformal filling composition 214 can include transferring a wafer patterned with recessed features into the superconformal filling composition.
(64) In the process, convectively transporting Au(SO.sub.3).sub.2.sup.3 anions and Bi.sup.3+ cations from superconformal filling composition 214 to bottom member 260 includes actively moving substrate 202 relative to superconformal filling composition 214. Actively moving substrate 202 relative to superconformal filling composition 214 can include displacing superconformal filling composition 214 across bottom member 206, exposed surface 216, sidewall 208, or field 204. Displacing can include rotating substrate 202, bubbling a gas (e.g., argon, nitrogen, carbon dioxide, and the like) through superconformal filling composition 214, stirring superconformal filling composition 214, heating superconformal filling composition 214 or substrate 202, recirculating superconformal filling composition 214, and the like. In an embodiment, actively moving includes rotating the patterned wafer using equipment for rotating disk electrodes to which the patterned wafer is attached and suspended within the superconformal filling composition. In an embodiment, actively moving substrate 202 relative to the superconformal filling composition includes rotating substrate 202 at a rotation rate from 0 revolutions per minute (rpm) to 2000 rpm, specifically at rotation rates from 100 rpm to 1600 rpm. The rotation rate can be variable or fixed.
(65) The process can include changing a rate of superconformal deposition of gold or changing from superconformally depositing gold to conformally or sub-conformally depositing gold. Here, it is contemplated that the process includes changing the rotation rate from a first rotation rate to a second rotation rate during superconformally depositing gold. The first rate can be, e.g., from 1200 rpm to 3000 rpm, specifically 1600 rpm, and the second rate can be from 400 rpm to 100 rpm, specifically 100 rpm.
(66) In the process, subjecting field 204 and recessed feature 250 to an electrical current can include attaching the specimen a Pt holder that is rotating with a contact to a potentiostat that applies current (or potential).
(67) In the process, increasing the electrical current subjected to field 204 and recessed feature 250 to maintain the V.sub.SSE between 0.85 V and 1.00 V relative to the SSE can include stepping or ramping the potential or current to maintain the potential in the range 0.85 V and 1.00 V relative to the SSE. As shown in
(68) According to an embodiment, the V.sub.SSE is maintained between 0.85 V and 1.00 V relative to the SSE until recessed feature 250 is completely filled with superconformal filling 210. In an embodiment, the VSSE is maintained between 0.85 V and 1.00 V relative to the SSE until recessed feature 250 is partially filled with the aspect ratio being less than or equal to 0.5; and thereafter the process includes changing a deposition condition to fill recessed feature 250 sub-conformally, conformally, or a combination of at least one of the foregoing non-superconformal filling regimes (i.e., sub-conformally filling or conformally filling).
(69) Substrate 202 can include a material such as silicon, silicon dioxide, germanium, or a compound semiconductor such as silicon nitride, gallium nitride. Such materials can be used for applications in electrodepositing gold on substrates for microelectronics or jewelry. In an embodiment, substrate 202 is a semiconductor, e.g., silicon. Substrate 202 can be multi-layered such a first layer is disposed on a second layer. The first layer can be, e.g., a semiconductor, and the second layer, e.g., can be a high-K dielectric such as a nitride of the material of the first layer, e.g., silicon nitride. It is contemplated that field 204 and recessed feature 250 including sidewall 208 and bottom member 206 are fully metallized to be electrically conductive for electrodeposition of gold thereon through superconformally depositing gold.
(70) Overlayer 212 provides full metallization of field 204 and recessed feature 250. Overlayer 212 can include a material such as gold, platinum, iridium, nickel, titanium, tantalum. Such materials can be used for adhesion to the substrate or wetting of the superconformal filling composition and superconformal filling. In an embodiment, overlayer 212 is a transition metal, e.g., Ti, Ta, or a combination thereof. A thickness of overlayer 212 can be from 1 nm to 1 m, specifically from 10 nm to 100 nm or specifically from 100 nm to 1 m. It is contemplated that in some embodiments overlayer 212 is an electrically conductive composite such as an electrically conductive polymer or an electrically conductive glass. Exemplary electrically conductive composites include indium tin oxide and the like.
(71) Substrate 202 has recessed feature 250 that includes bottom member 206 and sidewall 208. Recessed feature 250 can a trench, via, or another feature in which an electrical interconnect (here, superconformal filling 210) is formed. A shape of recessed feature as viewed from field 204 toward bottom member 206 can be via, trench, or a combination thereof. Before superconformally filling recessed feature 250, recessed feature 250 can have an aspect ratio of depth D to width W from 0.5 to 100, specifically from 1 to 20, wherein the aspect ratio increases during superconformally filling recessed feature 250. A size of width W can be from 10 nm to 50 m, specifically from 1 m to 10 m. A size of depth D can be from 100 nm to 50 m, specifically from 0.5 m to 5 m.
(72) Superconformal filling 210 is void-free and seam-free. It is contemplated that, in a presence of superconformal filling composition 250, passivation of field 204 and recessed feature 250 occurs at V.sub.SSE greater than 0.85 V relative to the SSE. Further in a presence of superconformal filling composition 250, sub-conformal deposition of gold occurs at the V.sub.SSE less than 1 V relative to the SSE, and superconformal deposition of gold occurs at the V.sub.SSE between 0.85 V and 1.00 V relative to the SSE. Accordingly, superconformal deposition of gold occurs when V.sub.SSE is maintained between 0.85 V and 1.00 V relative to the SSE.
(73) With reference to
(74) Superconformal filling 210 includes gold or an alloy of gold. Exemplary alloys include gold-cobalt, and gold-nickel. Elements in the alloy can be provided in superconformal filling composition 214. A purity of superconformal filling 210 can be from 98 at % Au to 100 at % Au, specifically based on the elements in the superconformal filling. An alloying element or contaminant can be present in super conformal filling 210 in an amount from 0 at % to 2 at %, based on the elements in the superconformal filling. Exemplary alloying elements are Co and Ni and exemplary contaminants include Na, Pb, Tl, Bi and the like.
(75) Advantageously, and unexpectedly, superconformal filling 210 is crystalline, dense, void-free, and seam-free of the macroscale, microscale, and nanoscale. In an embodiment, superconformal filling 210 is completely crystalline and is not amorphous. Crystalline domains of superconformal filling 210 include face centered cubic gold. Voids and seams include voids and seams along the centerline of the feature as well as pores within grains, and along grain boundaries, and the like, which are absent in superconformal filling 210 using an electron microscope at magnifications up to 100,000. As used here in, macroscale refers to dimensions that are of size 100 m to 1 mm. As used here in, microscale refers to 0.1 m to 100 m. As used here in, nanoscale refers to 1 nm to 0.1 m. In this regard, superconformal filling 210 is shiny and planar at exposed surface 216 on a submicron level with a brightness of superconformal filling 210 occurring in an absence of dendrites on exposed surface 216. It will be appreciated that electrochemical growth of surface deposits under certain conditions provides conformal fillings, sub-conformal fillings, or superconformal fillings (not superconformal filling 210) that have dendrites projecting away from the surface of the such filling. The dendrites are made of metal and decrease the brightness of such fillings such that such fillings are not as shiny superconformal filling 210 that lacks dendrites on exposed surface 216.
(76)
(77) Superconformal filling composition 214 forms superconformal filling 210 on bottom member 206 and has a near-neutral pH. As used herein, near-neutral pH refers to a pH that is from 5 to 10, with the lower value provided by instability of the superconformal filling composition. In an embodiment, the near-neutral pH of the superconformal filling composition is from 6.5 to 9. Moreover, superconformal filling composition 214 includes a plurality of Au(SO.sub.3).sub.2.sup.3 anions as a source of gold for superconformally depositing gold in recessed feature 250; a plurality of SO.sub.3.sup.2 anions; and a plurality of Bi.sup.3+ cations as a brightener and an accelerator for superconformally depositing gold in recessed feature 250. The Au(SO.sub.3).sub.2.sup.3 anions can be provided by a compound that includes the Au(SO.sub.3).sub.2.sup.3 anions such as Na.sub.3Au(SO.sub.3).sub.2. The Au(SO.sub.3).sub.2.sup.3 anions can be present in superconformal filling composition 214 in an amount from 5 millimolar (mM) to 350 mM, specifically from 40 mM to 80 mM. The SO.sub.3.sup.2 anions can be provided by a compound that includes the SO.sub.3.sup.2 anions such as K.sub.2SO.sub.3, Na.sub.2SO.sub.3, or a combination thereof. The SO.sub.3.sup.2 anions can be present in superconformal filling composition 214 in an amount from 0.1 molar (M) to 1 M, specifically from 0.6 M to 0.7 M. It is contemplated that a source of the Au(SO.sub.3).sub.2.sup.3 anions and the SO.sub.3.sup.2 anions are independent. Moreover, the concentration of SO.sub.3.sup.2 anions is independent from the concentration of Au(SO.sub.3).sub.2.sup.3 anions in that the SO.sub.3.sup.2 anions are not part of the Au(SO.sub.3).sub.2.sup.3 anions present in superconformal filling composition 214. The Bi.sup.3+ cations can be provided by a compound that includes the Bi.sup.3+ cations such as bismuth sulfate or bismuth oxide or may be added through electrochemical dissolution from a Bi metal source. The Bi.sup.3+ cations can be present in superconformal filling composition 214 in an amount from 1 micromolar (mmM or M) to 50 mmM, specifically from 2 M to 20 M. Without wishing to be bound by theory, it is believed that in the process for superconformally filling, the superconformal deposition of gold is catalyzed by Bi or its complexes adsorbed on the surface from the Bi.sup.3+ cations.
(78) Superconformal filling composition 214 can include an additive. Exemplary additives include hardeners, such as As and Sb, surfactants or deposition suppressing additives. In an embodiment, superconformal filling composition 214 includes Bi in the absence of additive.
(79) According to an embodiment, superconformal filling composition 214 consists essentially of Au(SO.sub.3).sub.2.sup.3 anions, SO.sub.3.sup.2 anions, Bi.sup.3+ cations in an aqueous liquid.
(80) In an embodiment, contacting field 204 and recessed feature 250 with superconformal filling composition 214 and obtaining bottom-up gold deposition occurs in an absence of thallium. In an embodiment, contacting field 204 and recessed feature 250 with superconformal filling composition 214 and obtaining bottom-up gold deposition occurs in an absence of lead. In an embodiment, contacting field 204 and recessed feature 250 with superconformal filling composition 214 and obtaining bottom-up gold deposition occurs in an absence of a suppressor.
(81) In the process, the electrical current provides a cathodic voltage (V.sub.SSE) between 0.85 V and 1.00 V relative to a saturated sulfate electrode (SSE), specifically from 0.85 V to 0.95 V. In an embodiment, the V.sub.SSE is 0.90 V. In an embodiment, the electrical current changes from 0.1 mA/cm.sup.2 to 10 mA/cm.sup.2, to maintain V.sub.SSE from 0.85 V to 1.00 V.
(82) With V.sub.SSE between 0.85 V and 1.00 V, a first deposition ratio of a first deposition rate of gold on bottom member 206 relative to a second deposition rate of gold on sidewall 208 can be from 2 to 1000 In an embodiment, a deposition ratio of a thickness of gold deposited on field 204 to a thickness of gold deposited on bottom member 206 is from 1:20 to 1:100.
(83) According to an embodiment, superconformally filling recessed feature 250 is bottom-up filling. In an embodiment, field 204 is passivated during bottom-up filling. In a particular embodiment, the bottom-up filling is uniform over the entirety of bottom member 206.
(84) Article 200 and the process for superconformally filling recessed feature 250 with gold has numerous beneficial uses, including defect-free Au metallizations in recessed geometries. In an embodiment, a process for making an electronic device with article 200 includes, includes: the superconformal deposition process described herein followed by chemical mechanical planarization to remove the electrically conductive overlayer leaving gold superconformal filling as electrically isolated wires and vias disposed in the substrate.
(85) Moreover, the process and superconformal filling 210 have numerous advantageous and beneficial properties. In an aspect, the process yields defect-free metal features for electrical conduction. The superconformal nature of the process yields reduced metal deposition on the field that decreases process cost, including reduced Au consumption in the superconformal filling composition as well as time and cost required to remove Au from the field.
(86) The process and superconformal filling 210 unexpectedly enables strictly bottom-up superconformal filling of recessed features as well as more general superconformal filling of recessed features, selectable using processing parameters including concentrations, convectively transporting, potential, or adjusting temperature.
(87) The articles and processes herein are illustrated further by the following Examples, which are non-limiting.
EXAMPLES
Example. Bottom-Up Filling of Damascene Trenches with Gold in a Sulfite Electrolyte
(88) In this Example, superconformally depositing Au is demonstrated in a Na.sub.3Au(SO.sub.3).sub.2Na.sub.2SO.sub.3 based superconformal filling composition electrolyte using underpotential deposited (upd) Bi to catalyze the reduction of Au(SO.sub.3).sub.2.sup.3. Micromolar additions of Bi.sup.3+ to the electrolyte give rise to acceleration of the Au deposition rate as evidenced by hysteretic voltammetry and rising chronoamperometric transients that convolve the diffusion, adsorption kinetics and coverage isotherm of the upd Bi catalyst. Void-free bottom-up filling of Damascene trenches using potential, as well as current, control is demonstrated for different Bi.sup.3+ concentrations. The bottom-up filling is accompanied under certain conditions by near-passivation of the field, resulting in deposition that resembles through-mask plating in the recessed features. The observed behavior is discussed in the context of previously described mechanisms for superconformal electrodeposition.
(89) Superconformal film growth through additive-based electrodeposition underlies successful implementation of Cu in state-of-the-art multilevel interconnect metallization for the silicon based semiconductor industry. Gold metallization for compound semiconductors and related optoelectronics enabling interconnect networks of arbitrary design and complex architecture involve Au Damascene processes and electroplating chemistries like those that provide void-free filling of recessed surface features for fabrication of Cu interconnects. Several Au-additive systems have been shown to provide such feature filling. Void-free superfill of Damascene features in accord with the Curvature Enhanced Accelerator Coverage (CEAC) mechanism has been achieved using under potential deposited (upd) Pb as a surfactant catalyst for Au deposition from KAu(CN).sub.2 electrolytes. However, both electrolyte toxicity and the aggressive nature of the alkaline electrolyte towards photoresist materials are serious shortcomings of the system. Micromolar concentrations of Pb added to Na.sub.3Au(SO.sub.3).sub.2+Na.sub.2SO.sub.3 electrolyte also yield Au superfill of Damascene features through the CEAC mechanism. However, Restrictions On Hazardous Substances (ROHS) precludes use of Pb at even these extremely low concentrations. The addition of mercaptopropane sulfonic acid (MPS) to a commercial Na.sub.3Au(SO.sub.3).sub.2+Na.sub.2SO.sub.3 with a pH buffer and Tl.sup.+ as a grain refiner enables superconformal deposition consistent with leveling through an incorporation-derived gradient of suppressor concentration, indicated by electroanalytical measurements to be the MPS. Additive incorporation in the Au deposit explicit in the mechanism is a concern with this approach, as is the use of Tl, again due to ROHS. A different suppressor, polyethyleneimine, exhibiting sharply defined suppression breakdown in a sulfite electrolyte manifest as S-shaped negative differential resistance (S-NDR) during voltammetry, yields superconformal Au filling that is entirely localized within recessed through silicon vias (TSV). However, the deposits contain nanoscale porosity due to additive incorporation that is part of the filling mechanism.
(90) The coinage metals Au, Ag and Cu can be deposited superconformally using electrolyte-additive systems containing accelerators that enable the CEAC mechanism of superfill. In these systems, decrease of surface area during growth on concave surfaces increases the coverage of adsorbed accelerator. The CEAC mechanism thus yields accelerated filling of recessed (i.e., concave) features through positive feedback. Examples include the alkaline cyanide and near-neutral sulfite Au electrolytes noted previously, Cu sulfate electrolytes and Ag cyanide electrolytes. Feature filling models based on the CEAC mechanism accurately predict the observed superconformal feature filling. CEAC-based models also predict observed surface smoothing and stabilization of surface planarity derived through the same mechanism.
(91) Near-neutral sulfite electrolytes are environmentally and practically superior to cyanide. They do suffer from disproportionation, SO.sub.2 formation, or dithionnite-based decomposition in acidic media. However, their instability can be mediated using additives such as ethylenediammine or electrolyte modification, i.e., thiosulfate, enabling operation to pH values as low as 4.0. Sulfite plating baths, like cyanide, can exhibit constrained kinetics that are at least partly associated with the formation of inhibiting species on the surface. For superconformal deposition through the CEAC mechanism, this native inhibition can be lifted by the addition of Pb, Tl, Bi, or Sb, yielding depolarization in voltammetry so that these additives are classified as accelerators of the Au deposition. When optimized, their addition permits fabrication of bright Au films.
(92) The utility of upd Bi as a surfactant catalyst for superconformal Au deposition in sub-micrometer trenches is explored in the Na.sub.3Au(SO.sub.3).sub.2+Na.sub.2SO.sub.3 electrolyte. Voltammetry and chronoamperometry on a rotating disk electrode (RDE) provide data that assist in understanding the observed feature filling behavior.
Experimental Details
(93) The Au deposition used an industrial sodium gold sulfite source (Na.sub.3Au(SO.sub.3).sub.2) containing 2 troy ounces of Au per liter, equivalent to 0.32 mol/L, that was diluted to one-fourth its original concentration using 18 M.Math.cm water. Deposition studies were conducted in a cell containing 40 mL of the 0.080 mol/L Na.sub.3Au(SO.sub.3).sub.2 with and without 0.64 mol/L of Na.sub.2SO.sub.3 supporting electrolyte. Bismuth was introduced by anodic dissolution of 99.999% pure Bi, the stated concentrations assuming 100% efficiency and a Bi.sup.3+ state. The electrolyte pH was 9.5, as determined using a pH sensor calibrated to buffer solutions of pH 7.0 and pH 10.0. Cyclic voltammometry and chronoamperometry were performed on Au coated rotating disk electrode (RDE) of 1.0 cm diameter Ag, polished to 1200 grit SiC paper, and current densities were obtained from the measured current using its nominal area. The electrolyte was sparged with argon before the cyclic voltammetry to reduce the amount of dissolved oxygen. Feature filling was studied using 3 mm11 mm fragments of patterned wafers having a 0.2 m thick Au seed in the field and a lesser amount on the side walls and bottoms of the patterned trenches and vias. The patterned substrates were rotated about one end from a Pt spindle during deposition, like a helicopter blade (the patterned surface facing upwards) to give definition to the metal ion and additive transport. Based on the 1 cm distance between the rotational axis and the imaged features, a 200 rad/min (100 rpm) rotation rate corresponds to an estimated 10 cm/s flow rate over the surface. The electrolyte was at room temperature (23 C.) during deposition. A saturated Hg/Hg.sub.2SO.sub.4/saturated K.sub.2SO.sub.4 reference electrode (SSE), connected to the working electrode compartment via a fritted bridge filled with saturated solution of potassium sulfate, was used for all experiments. The platinum counter electrode was held in a frit-separated cell immersed within the main cell.
(94) Electrochemical Measurements on Planar Substrates
(95) Cyclic voltammetry in electrolyte containing different Bi.sup.3+ concentrations is shown in
(96)
(97) Current density approaches an asymptotic value more rapidly with higher Bi.sup.3+ concentration and at more negative potential. The dependence on concentration is not surprising as it is consistent with either transport or interface kinetics limited adsorption of the dilute Bi.sup.3+ from the electrolyte. The dependence on potential suggests a change in the kinetics of adsorption or rearrangement of the surface. The current density of the plateau increases with potential and with Bi.sup.3+ concentration. Analogous behavior was noted in the voltammetry and suggests an increase of the saturation adsorbate coverage with increase of the Bi.sup.3+ concentration; higher concentration would otherwise shorten the time to saturation without affecting the saturation current density. Without wishing to be bound by theory, it is believed that consumption of the adsorbed Bi occurs during Au deposition and concentration-dependent equilibrium coverage.
(98) Feature Filling
(99) Trench filling is shown in
(100) The impact of substrate rotation rate suggested by the results in
(101) Maps that summarize the nature of feature filling systematically over a range of conditions including the set included in
(102) The filling behavior here with use of only an accelerating additive differs from previous observations of electrolytes that provide superconformal Au, Ag, and Cu filling of features using accelerators. While these other systems also exhibit an incubation period prior to bottom-up filling, it is accompanied by some amount of conformal growth on all surfaces. Furthermore, accelerated deposition in these other systems initiates at the bottom corners of the feature rather than the nominally uniform activation of the entire bottom surface captured in
(103) There are aspects where deposition is consistent with CEAC superfill. These include growth in wider features, where nucleation does occur in the lower corners. Furthermore, where significant deposition does occur on the field, it is very smooth (
(104) These results are consistent with the current transients in
(105) Moreover, current densities on the planar RDE are well below the gold ion transport limit at the 0.90 V and 0.95 V potentials yielding bottom-up deposition at the various Bi.sup.3+ concentration (
Comparative Example A. Superfilling Damascene Trenches with Gold in a Sulfite Electrolyte Including Lead
(106) Superconformal Au deposition is demonstrated in a Na.sub.3Au(SO.sub.3).sub.2Na.sub.2SO.sub.3 based electrolyte using underpotential deposited (upd) Pb to catalyze the reduction of Au(SO.sub.3).sub.2.sup.3. Micromolar additions of Pb.sup.2+ to the electrolyte give rise to acceleration of the Au deposition rate as evidenced by hysteretic voltammetry and rising chronoamperometric transients that convolve the diffusion, adsorption kinetics and coverage isotherm of the upd Pb catalyst. The catalytic activity on the Au deposition reaction may be quenched, or reset, by stepping the potential to more positive values to deactivate or strip the Pb catalyst. Void-free Damascene trench superfilling at various fixed potentials and Pb.sup.2+ concentrations is demonstrated and shown to be congruent with the curvature enhanced accelerator model (CEAC) of superconformal growth. This includes enhanced deposition due to catalyst enrichment on concave surface segments at the bottom corners of the trenches, followed by rapid bottom-up growth and, for certain conditions, bump formation above the filled trenches. It should be appreciated that Pb is toxic and not included in the superconformal filling composition described in the Detailed Description and the Example.
(107) Superconformal film growth by additive-based electrodeposition underlies the successful implementation of Cu in state-of-the-art multilevel interconnect metallization for the silicon based semiconductor industry. In contrast, electrodeposited Au remains the metallization of choice for interconnects and contacts in compound semiconductors and related optoelectronics applications as detailed in recent reviews. To date, the majority of these applications have involved through-mask-electrodeposition. However, increasing circuit densification demands a move to three-dimensional metallization to enable the fabrication of interconnect networks of arbitrary design and complex architectures. Implementation of the Damascene process calls for Au electroplating chemistry that can provide void-free filling of recessed surface features. Previously two different Au electrolytes have been examined and shown to be capable of void-free feature filling. The first derives from the original observation of McIntryre and Peck that underpotential deposited (upd) Pb catalyzes Au deposition from KAu(CN).sub.2-based electrolytes. When this acceleration is combined with mass conservation of the surfactant in the presence of electrode area change, superconformal Au void-free feature filling occurs in accord with the CEAC model. Practically speaking, two short-comings of the system are electrolyte toxicity and the aggressive nature of the alkaline media towards photoresist materials. In order to address these challenges a sulfite-based Au electrolyte was evaluated by Hu and Ritzdorf. The addition of mercaptopropane sulfonic acid (MPS) to a commercial electrolyte comprised of Na.sub.3Au(SO.sub.3).sub.2, Na.sub.2SO.sub.3, a pH buffer and Tl.sup.+ as a grain refiner enabled void-free superconformal deposition. Electroanalytical measurements indicate that MPS serves as an inhibitor that facilitates feature filling in line with traditional leveling concepts although no quantitative modeling was performed. In this report the utility of upd Pb as a surfactant catalyst for Au superfilling of sub-micrometer trenches is revisited using a slightly alkaline Au(SO.sub.3).sub.2.sup.3 electrolyte. Electroanalytical and feature filling experiments are shown to be in good agreement with the CEAC mechanism of superconformal deposition.
(108) The CEAC feature filling mechanism derives from the competition between accelerating and rate inhibiting adsorbates for available surface sites. Successful filling is ascribed to the increasing coverage of the more strongly bound accelerator surface phase (e.g., upd Pb in the present case) during area reduction that accompanies deposition on concave recessed surface segments. The positive feedback due to coupling of decreasing area and increasing accelerator adsorbate coverage at the bottoms of filling features leads to locally increased deposition rate and accelerated bottom-up superfilling. Several feature filling models based on the CEAC mechanism have been shown to accurately predict superfilling of patterned features. Examples can be found ranging from copper deposition in damascene features and microvias from acid sulfate-based electrolytes to the superfilling of trenches and vias with silver and gold deposited from KSeCN, SC(NH.sub.2).sub.2 or Pb catalyzed cyanide-based electrolytes. The surface smoothing capacity and stability associated with the CEAC mechanism has been evaluated as has the influence of surface diffusion 24 and surface derivatization.
(109) Research on sulfite chemistry for Au electrodeposition reportedly dates back to 1842. Compared to cyanide, sulfite electrolytes are environmental friendly but suffer from disproportionation, SO.sub.2 formation, and/or dithionnite-based decomposition in acidic media. The instability can be mediated through additive additions, e.g., ethylenediammine, and/or electrolyte modification, i.e., thiosulfate, to enable operation to be extended to pH values as low as 4. Cyanide and sulfite plating baths both exhibit constrained kinetics at least partly associated with the formation of inhibiting species on the surface. The native inhibition of these electrolytes may be lifted by the addition of Pb, Tl, or Sb resulting in depolarization, i.e., acceleration of the deposition rate, that when optimized yields bright Au films.
Experimental
(110) The depositions detailed in this work utilized an industrial sodium gold sulfite source (Na.sub.3Au(SO.sub.3).sub.2) stated to contain 2 troy ounces of gold per liter, equivalent to 0.32 mol/L (197 g/mol gold with 31.1 g=1 troy ounce). This solution was diluted to one-fourth its original concentration using 18 M.Math.cm water and Na.sub.2SO.sub.3 was added as a supporting electrolyte. Deposition studies were conducted in a cell containing 100 mL of the 0.080 mol/L Na.sub.3Au(SO.sub.3).sub.2-0.32 mol/L Na.sub.2SO.sub.3 electrolyte with sequential additions of aliquots drawn from a 1 mmol/L Pb(ClO.sub.4).sub.2 solution. The electrolyte pH was 9.5, as determined using a pH sensor calibrated to buffer solutions of pH 7.0 and pH 10.0. According to Pourbaix's atlas, Pb(ClO.sub.4).sub.2 (s) dissolution in pH 9.5 solution results in complex formation with two species present in the ratio [HPbO.sub.2.sup.]/[Pb.sup.2+]3.49 The electrolyte was sparged with nitrogen prior to use in order to remove dissolved oxygen. Cyclic voltammometry and chronoamperometry were performed using 100 nm thick planar Au films prepared by electron beam evaporation on Si wafers. The electroanalytical screening studies were performed under quiescent conditions although the vertical orientation of stationary substrates results in density-gradient-driven convection as the Au(SO.sub.3).sub.2.sup.3 is depleted in the electrode boundary layer by deposition. Feature filling experiments were conducted with the patterned specimens held in the same orientation but rotating at 160 rad/min (80 rpm) with the rotation axis oriented vertically within the plane of the specimen. This enhanced and fixed the Au(SO.sub.3).sub.2.sup.3 transport, necessary because CEAC-based superfilling fails when metal ion transport dominates interface kinetics. The electrolyte was at room temperature (23 C.) blanketed with nitrogen during deposition. A saturated Hg/Hg.sub.2SO.sub.4/saturated K.sub.2SO.sub.4 reference electrode (SSE) was used for all experiments. The reference electrode was connected to the working electrode compartment via a fritted salt bridge filled with saturated solution of potassium sulfate while a platinum counter electrode was held in a frit-separated cell immersed within the main cell.
(111) The planar substrates were pieces of a polished silicon wafer coated with 100 nm Au on a 10 nm Ti adhesion layer by electron beam evaporation. These specimens were masked with plater's tape to expose a circular region 0.54 cm.sup.2 in area; current densities are obtained from the measured current divided by this area. Studies of superfill in trenches were conducted on fragments of patterned wafers at fixed potential. The substrates with an array of trench widths were purchased from Sematech with a copper seed and Ta barrier already in place. Depositions were performed at fixed potential and each specimen included all the trench sizes studied.
(112) Electrochemical Measurements on Planar Substrates
(113) Cyclic voltammetry in
(114)
(115) Evolution of the boundary layer is convolved with the increasing metal deposition that accompanies increasing adsorbate coverage, particularly evident in the first several tens of seconds as the density gradient that drives the fluid flow first develops. The current density approaches an asymptotic value more rapidly for higher additive concentration, consistent with the dependence on additive concentration (albeit at the interface). Deposition exhibiting this behavior creates mirror-like deposits despite deposition rates approaching substantial fractions of the transport limit. This is in accord with modeling that quantifies the substantial surface stabilization that can arise from the CEAC mechanism (bottom-up filling is effectively a brightening process that flattens surfaces by removing micrometer to nanometer size scratches).
(116) Increase of the additive concentration in the electrolyte yields a larger steady-state current density, from which an increase of the maximum adsorbate coverage is inferred. This suggests some consumption of the adsorbate during Au deposition and/or a concentration-dependent equilibrium coverage. While the former was demonstrated and quantified for lead additive in cyanide-based superfilling electrolyte (preadsorbed lead gradually consumed during deposition in lead-free electrolyte), the latter evidently plays a substantial role here.
(117)
(118) The results in
(119) Feature Filling
(120) All feature filling was initiated with a 5 s pulse at 1.5 V, to obtain high nucleation density and protect the Cu seed from corrosion, followed immediately by deposition at the indicated potential for the indicated time. Without the initial step the Cu seed corroded through galvanic displacement.
(121)
(122) Predicted and previously observed manifestations of the CEAC mechanism in multiple superfilling systems are evident: an incubation period of conformal growth (appropriate for no initial adsorbate coverage), inception of superfill through enhanced deposition at the bottom corners of the trench (locations of highest curvature), the signature bottom-up filling that earned the process its moniker and the overfill bumps referred to as momentum plating. Deposition at 1.2 V also yields deposits consistent with CEAC smoothing.
(123) Faceted whiskers indicative of morphological instability are evident at the shortest deposition times for all four conditions (
(124) Superfilling is most evident for conditions where there is both substantial hysteresis in the cyclic voltammetry (
(125) Deposition times for feature filling were generally such that corresponding current densities on planar substrates were significantly below the gold ion transport limit (
(126) Surface diffusion of the CEAC-redistributed adsorbate, previously shown to explain the rounded contours of Au superfilling in cyanide electrolyte as opposed to the sharp corners observed in Cu superfill might also be surmised to explain the more limited rounding evident in
(127) Optimizing adsorbate coverage for superfill generally requires additive concentration scale inverse to feature size to ensure appropriate accelerator coverage is reached during the feature size-dependent fill time. In this system in particular, for the lower additive concentration required for larger features, CEAC-induced coverage increase might be countered by desorbtion where such enhanced coverage exceeds the correspondingly low equilibrium coverage, diminishing or eliminating CEAC-induced superfilling.
(128) Such concerns notwithstanding, the potential-dependent saturation coverage indicated by
(129) Bottom-up superfill of Au has been demonstrated in sub-micrometer trenches using Pb upd as an accelerator in a Na.sub.3Au(SO.sub.3).sub.2 electrolyte. The system exhibits rising chronoamperometric transients and hysteretic voltammetry that are ubiquitous characteristics of superfilling electrolytes. Feature filling is accompanied by surface evolution that reflects enrichment of the Pb upd accelerator on surface segments where area reduction occurs, including growth from the bottom trench corners, v-notch formation, upward accelerating flat bottom surface and bump formation. For application beyond the Damascene feature filling demonstrated herein pulse plating might be used to reset the accelerated gold deposition to specific accelerated, or unaccelerated, rates by stepping the potential to appropriate values. This has implications for processing across other materials systems.
Comparative Example B. Extreme Bottom-Up Filling of Through Silicon Vias and Damascene Trenches with Gold in a Sulfite Electrolyte
(130) Superconformal Au deposition is demonstrated in a Na.sub.3Au(SO.sub.3).sub.2 based electrolyte that exhibits critical behavior with addition of the deposition-rate suppressing additive polyethyleneimine (PEI). The critical behavior results from s-shaped negative differential resistance that is stabilized by the resistive electrolyte as evident in hysteretic cyclic-voltammetry. Specifically, the hysteretic potential range indicates conditions under which two states can exist: one in which adsorbed PEI suppresses gold deposition and the other in which ongoing gold deposition continues to disrupt and bury arriving suppressor. The specimen history determines the state observed. Extreme bottom-up filling of Through Silicon Vias (TSVs) and superconformal filling of submicrometer Damascene trenches are demonstrated and explained in this context.
(131) Superconformal film growth by additive-based electrodeposition underlies Cu processing in state-of-the-art multilevel interconnect metallization for the silicon based semiconductor industry. On the other hand, electrodeposited Au remains the metallization of choice for interconnects and contacts in compound semiconductors and related optoelectronics applications. While the majority of these applications presently involve through-mask-electrodeposition, increasing circuit densification requires a move to three-dimensional metallization and complex interconnect structures. The implementation of Damascene processing requires Au electroplating chemistries that can provide void-free filling of recessed surface features. To date, three different Au electrolytes have been shown to be capable of void-free feature filling. One electrolyte utilizes an underpotential deposited (upd) Pb process to accelerate Au deposition from KAu(CN).sub.2-based chemistry; a second chemistry uses a similar upd Pb process to accelerate Au deposition from a Na.sub.3Au(SO.sub.3).sub.2-based electrolyte. Both of these electrolytes provide superfill and void-free feature filling in accord with Curvature Enhanced Accelerator Coverage (CEAC) models that account for coupling of the rate-acceleration and mass conservation of the surfactant in the presence of electrode area change. The third electrolyte is a commercial sulfite-based Au electrolyte to which mercaptopropane sulfonic acid (MPS) has been added. Electroanalytical measurements indicate that the MPS serves as an inhibitor, with feature filling attributed to traditional leveling; i.e., additive incorporation in the deposit yields an MPS concentration gradient within the filling feature, the decreasing suppressor concentration inducing a smoothly varying, monotonic increase of metal deposition rate down the feature.
(132) This comparative example describes Au deposition from an electrolyte containing a deposition-rate suppressing additive, in this case polyethyleneimine (PEI). However, as will be detailed, the electrochemical signature and Au deposits, particularly within filling TSVs, are consistent with the recently detailed two-state extreme bottom-up filling of TSV rather than traditional leveling. The mechanism derives from s-shaped negative differential resistance (S-NDR) observed in cyclic voltammetry (CV) measurements of suppressor-containing electrolyte. In particular, over a limited range of deposition potential two states can exist: one in which adsorbed PEI suppresses gold deposition and the other in which ongoing gold deposition continues to disrupt and bury arriving suppressor. The spatial distribution of the bifurcation between the two states is coupled to the distributed electrolyte resistance within the deposition system. Temporal evolution of the gradient of suppressor within the feature dictates the geometrical distribution of the suppressed and active regions. This same mechanism was also suggested to explain superconformal filling of Damascene trenches by Cu as well as iron group alloys in resistive, additive-suppressed electrolytes and to underlie the symmetric butterfly pattern Cu deposits reported in through-holes.
(133) Sulfite electrolytes are environmentally friendlier than cyanide. While they suffer from disproportionation, SO.sub.2 formation, and/or dithionite-based decomposition in acidic media, the electrolyte instability can be mediated with additives, e.g., ethylenediamine, and/or electrolyte modification, i.e., thiosulfate, extending operation to pH values as low as 4.0. Sulfite plating baths exhibit constrained kinetics at least partly associated with the formation of inhibiting species on the surface. There has been substantial study of the lifting of such native inhibition because of the role that such depolarization plays in the formation of bright, smooth deposits. However, inhibition as studied herein is more relevant for filling of TSVs because it benefits from gradients of additives that are to be anticipated during filling of tall features.
Experimental
(134) The depositions detailed in this comparative example use an industrial aqueous gold sulfite source (Na.sub.3Au(SO.sub.3).sub.2 dissolved in H.sub.2O stated to contain 2 troy ounces of gold per liter, equivalent to 0.32 mol/L (197 g/mol gold and 31.1 g=1 troy ounce). This solution was used in its undiluted form as the working electrolyte. Supporting electrolyte was omitted because higher electrolyte resistivity favors extreme bottom-up feature filling and process control with the otherwise unstable processes. Deposition studies were conducted in a cell containing approximately 60 mL of the 0.32 mol/L Na.sub.3Au(SO.sub.3).sub.2 electrolyte with additions of aliquots drawn from a 10 mmol/L PEI made from 1800 MW PEI diluted in 18 M.Math.cm water. The electrolyte pH was 9.0, as determined using a pH electrode calibrated to buffer solutions of pH 7.0 and pH 10.0.
(135) Cyclic voltammetry and chronoamperometry were performed on 100 nm thick planar Au films prepared by electron beam evaporation on Si wafers. The electroanalytical screening studies were performed under quiescent conditions; the vertical orientation of stationary substrates results in density-gradient-driven convection as the Au(SO.sub.3).sub.2.sup.3 in the boundary layer is depleted by deposition. Feature filling experiments were conducted with the patterned specimens held in the same orientation but rotating at 160 rad/min (80 rpm) with the rotation axis oriented vertically within the plane of the specimen unless otherwise indicated. The electrolyte was at room temperature (23 C.) under a covering flow of nitrogen during deposition. A saturated Hg/Hg.sub.2SO.sub.4/saturated K.sub.2SO.sub.4 reference electrode (SSE) was used for all experiments. The reference electrode was connected to the working electrode compartment using a fritted salt bridge filled with saturated solution of potassium sulfate. A platinum counter electrode was held in a frit-separated cell immersed within the main cell. Nitrogen sparging was used to remove dissolved oxygen. Depositions were performed at controlled applied potential.
(136) The planar substrates were masked with plater's tape to expose a circular region 0.54 cm.sup.2 in area; current densities are obtained from the measured current divided by this area. Feature filling was studied on fragments of patterned wafers. A wafer with TSVs, already coated with 2 m thick Au seed in the field, and a lesser amount on the features side walls, thicker toward the top, was provided for this study by Hughes Research Laboratory. The 50 m deep TSVs have a square cross-section 30 m on a side.
(137) Electrochemical Measurements on Planar Substrates
(138)
(139) In contrast, the subsequent positive-going sweep varies smoothly with potential. Evidently, PEI adsorbed at the slower deposition rates of less negative potentials retards Au deposition as the potential moves toward the critical value (i.e., hindering access and/or reduction of the Au(SO.sub.3).sub.2.sup.3 at the interface); in contrast, the more rapid Au deposition at more negative potentials enables continued Au deposition (i.e., through burial of adsorbing suppressor) as the potential returns to values positive of the transition potential. The resulting hysteresis (e.g., from approximately 1.2 V to the transition at 1.28 V for 110 mol/L PEI) suggests the possibility of a two-state system in this potential range. While not addressing the stability of the states, or any dependence on potential, the 35 s the scan takes to traverse the 70 mV width of the hysteretic region suggests the ongoing deposition can retard buildup of the blocking PEI layer for at least this long. Gas bubble formation at the perimeter of the substrate increased with higher PEI concentration, yielding an associated increase of the current density at the most negative potentials (
(140)
(141)
(142) Such growth instabilities have been previously seen on planar deposits from the Cu electrolyte that yields extreme bottom-up fill of TSV and exhibits analogous hysteresis and critical behavior. A possible explanation for the transition from uniform deposition to unstable growth going up the planar substrates can be found in analysis of that Cu system; specifically, where two solutions to the equation for steady state adsorbate coverage do exist, changing the ratio of the suppressor and metal ion concentrations can change an unsuppressed surface to full coverage (and suppression), or vice-versa.
(143) Rotation of the substrates during deposition yielded optically smooth deposits even for applied potential near the critical value. The unusual rotation smears pattern development by disturbing the electrical field and concentration gradients preventing formation of regions of enhanced/decreased adsorbate coverage even as it enhances Au(SO.sub.3).sub.2.sup.3 and additive transport.
(144) Feature Filling
(145) Filling of the TSV was accomplished by pre-wetting with ethanol followed by immersion in the electrolyte at open circuit potential followed by prolonged, high speed rotation (at least 1 min at 400 revolutions per minute) to remove the alcohol from the features prior to application of the deposition potential. Without the alcohol predip air bubbles were frequently trapped in the TSVs.
(146)
(147)
(148) To avoid the transition to conformal filling, and thus reduce filling time,
(149) Superconformal filling of submicrometer Cu-seeded trenches is shown in
(150) The CEAC mechanism, while it has been observed and modeled at TSV length scales, is not at play here with a suppressing additive and enhanced deposition at locations independent of local curvature (e.g.,
(151) In contrast to the results for Cu filling of TSV, the Au deposition rate in the off state is still significant. Evident in the negative-going sweep of the CV measurements, this is also captured in the current density at the inflection points as summarized in
(152) The Au filling contrasts with the case of extreme bottom-up Cu filling of annular TSV in that the ratio of on and off deposition rates was orders of magnitude higher with the Cu deposition. The Cu filling also involved unsuppressed deposition current densities whose absolute values were nearly an order of magnitude higher than those here; the 1 mA currents associated with Au fill on the TSV die fragments yield a potential drop from the applied value of 10 mV across the 9 cell resistance. As a result of both effects, the iR potential drop in these filling experiments does not vary the overpotential available for deposition at the substrate as substantially during feature filling. It is thus likely for these reasons that the strict localization of deposition at the bottom of the feature and upward progress of filling observed at fixed applied potential with the Cu are not observed here. The images in
(153) One might confuse the Damascene filling images in
(154) Modeling quantifies observed effects in terms of metal ion and suppressor concentrations, applied potential and the electrical and transport characteristics of the cell in order to provide a-priori predictions of feature filling. Evaluation of kinetic factors for such modeling will need to remove ambiguity associated with non-planar deposition on planar substrates (
(155) Bottom-up superfill of Au has been demonstrated in this comparative example in 50 m deep re-entrant TSV and sub-micrometer trenches using PEI as a suppressor in a Na.sub.3Au(SO.sub.3).sub.2 electrolyte, which will be appreciated to be different than the superconformal filling composition described in the Detailed Description and Example. The electrolyte exhibits hysteretic S-NDR behavior in cyclic voltammetry. Deposition shifts from passive to active surfaces over only a few tens of mV on the negative going sweep, with deposition on the return sweep remaining active to potentials positive of the critical value. The S-NDR mechanism coupled with the gradient of suppressor concentration within filling features yields variants of extreme bottom-up filling of TSV and superconformal filling of submicrometer Damascene trenches. Potential drop within the electrolyte, for applied potential in the hysteretic range, permits controlled deposition. The results presented here involve a high PEI concentration; filling in additive-stabilized sulfite and sulfite-thiosulfite electrolyte at lower pH are natural extensions and might be anticipated to permit lower impurity content. The results extend the generality of the previously detailed S-NDR behavior, linking extreme bottom-up Cu fill and superconformal Ni fill of features that differ in size by nearly two orders of magnitude, in a single electrolyte.
Comparative Example C. Morphological Transitions During Au Electrodeposition: From Porous Films to Compact Films and Nanowires
(156) Gold electrodeposition was studied in a sulfite electrolyte to which micromolar concentrations of Tl.sub.2SO.sub.4 were added. Hysteresis and a regime of negative differential resistance (NDR) evident in electroanalytical measurements are correlated with deposit morphology and interpreted through measurements of thallium underpotential deposition (upd). Deposit morphologies range from specular surfaces to highly faceted dendrite-like grains of moderate aspect ratio and, for potentials within the NDR region, sub-50 nm diameter, high aspect ratio 110 oriented single crystal nanowires. The nanowires exhibit an epitaxial relationship to the substrate that permits one step fabrication of surfaces densely covered with high aspect ratio nanowires having controlled orientations. The NDR and nanowires are a consequence of the non-monotonic relationship between Tl coverage and growth velocity; at low coverage Tl accelerates Au deposition while at higher coverage it inhibits deposition. Immiscibility of the Tl and Au supports on-going surface segregation during area expansion that accompanies nanowire growth leading to greater dilution of the additive coverage and more rapid growth at the nanowire tips, while the sidewalls remain passivated by a saturated Tl coverage. It should be appreciated that Tl is absent in the superconformal filling composition described in the Detailed Description and Example in which the superconformal filling composition provides advantageous and unexpected benefits.
(157) Interest in nanostructures is driven by the unique properties associated with high surface area to volume materials that express quantum or mesoscale phenomena of importance to diverse subjects ranging from biology and medicine to optics and energy conversion. Of particular interest is the substantial literature on solution or vapor processing to yield a wide variety of nanostructures including high aspect ratio nanowires. Solution-based, chemical reduction processes are attractive because they are inexpensive and scalable. Subsequent processing involving particle packing and/or substrate interactions can lead to higher order organization and colloidal crystals. Alternatively, nanostructures can be grown on or in surfaces that have been shaped by lithography or other means to obtain control of orientation and/or placement. Electrochemical processing may involve through-mask plating in patterned templates or superconformal electrodeposition of metals and alloys including gold in high aspect ratio features. However, these processes require nanoscale patterned topography to create such structures and the associated patterning often comes with increased processing complexity.
(158) Controlled anisotropic growth and orientation, if not position, is possible through vapor-liquid-solid (VLS) growth processes. Fabrication of semiconductor and oxide nanowires typically uses dewetting of a low melting point metal catalyst where the resulting clusters set the feature size while epitaxy to the underlying substrate can control the growth orientation. More recently, an electrochemical electrolyte-liquid-solid analog has been demonstrated for growth of Ge nanowires. In another approach nanowire growth of metallic nanostructures capitalizes on the intrinsic or additive engineered anisotropic energetics and/or growth kinetics on different crystal facets. For example, adsorption of anions or dilute metal ion additives, via under potential deposition (upd), enables fabrication of structures such as Au nanospikes of moderate aspect ratio using a Pb additive and highly faceted, moderately elongated dendrite-like grains using Tl as an additive on unpatterned substrates. This work, part of a broader study of the impact of Pb.sup.2+ and Tl.sup.+ additives on the morphology of Au electrodeposits on planar and patterned substrates, demonstrates the use of dilute Tl.sup.+ for controlled growth of single crystal nanowires of much higher aspect ratio on unpatterned surfaces. The additive selection is based on its surfactant quality, derived from the absence of TIAu intermetallic phases and negligible mutual solubility (in contrast, there are a number of AuPb stoichiometric compounds).
(159) Deposition was conducted in a non-cyanide, alkaline Au(SO.sub.3).sub.2.sup.3 electrolyte at pH 9. At lower pH, sulfite-based electrolytes suffer from disproportionation, SO.sub.2 formation, and/or dithionnite-based decomposition. The instability can be mitigated through the use of additives such as ethylenediamine, and/or electrolyte modification such as the use of thiosulfate, to enable operation at pH values as low as 4.0.
(160) Similar to cyanide electrolytes, metal deposition from sulfite solutions is constrained by the formation of inhibiting species on the deposit surface. This native inhibition can be lifted by the addition of Tl.sup.+ or other metal ions to obtain accelerated Au deposition rates and surface smoothing. Alternatively, the native suppression can be enhanced through the use of other inhibiting additives. Such accelerating and suppressing additives underlie a number of processes for creating high aspect Au features through superconformal filling of high aspect ratio trenches and vias in patterned surfaces. Thallium is particularly interesting because of reports of coverage dependent behavior whereby Au electrodeposition is accelerated for low Tl coverage while significant suppression occurs at higher coverages. As will be shown in this comparative example, the non-monotonic behavior leads to a transition between two distinct behaviors indicated schematically in
Experimental
(161) The depositions detailed in this work utilized a commercial sodium gold sulfite source containing 2 troy ounces of gold per liter, equivalent to 0.32 mol/L Na.sub.3Au(SO.sub.3).sub.2. Gold electrodeposition experiments were conducted in a cell containing 60 mL of 0.080 mol/L Na.sub.3Au(SO.sub.3).sub.2+0.32 mol/L Na.sub.2SO.sub.3 with and without addition of aliquots drawn from a 1 mmol/L Tl.sub.2SO.sub.4 solution. The electrolyte pH was 9, as determined using a pH electrode calibrated using pH 7.0 and pH 10.0 buffer solutions. The electrolyte was sparged with nitrogen prior to use to remove dissolved oxygen. Thallium adsorption was also examined by cyclic voltammetry in the 0.32 mol/L Na.sub.2SO.sub.3 supporting electrolyte.
(162) Electroanalytical and deposit morphology studies were performed on Si wafers coated with 100 nm thick planar Au films by electron beam evaporation. The specimens were masked with plater's tape to expose a circular area of 0.54 cm.sup.2. The substrates were held in a vertical orientation under quiescent conditions, the only convection being that arising from the density-gradient due to deposition derived Au(SO.sub.3).sub.2.sup.3 depletion in the electrode boundary.
(163) The experiments were performed at room temperature (23 C.) while the electrolyte was sparged with nitrogen between experiments. A Hg/Hg.sub.2SO.sub.4/saturated K.sub.2SO.sub.4 reference electrode (SSE) was used for all experiments. The reference electrode was connected to the working electrode compartment via a fritted salt bridge filled with saturated solution of potassium sulfate while a platinum counter electrode was held in a separate fritted compartment filled with the working electrolyte.
(164) Electrochemical Measurements
(165) Cyclic voltammetry was used to examine the impact of Tl.sup.+ concentration on Au deposition through sequential addition of Tl.sub.2SO.sub.4 aliquots (
(166) Chronoamperometry was used to further examine the metal deposition dynamics. The potential was stepped from open circuit conditions to representative potentials in the various electrolytes (
(167) The chronoamperometric data were used to construct sampled current voltammograms and compared to the slow scan voltammetry data for the electrolyte containing 20 mol/L Tl.sup.+ (
(168) In the case of gold cyanide electrolytes, inhibition of the deposition reaction is associated with adsorbed CN and the related metal complexes. However, less is known about the interactions between Au and SO.sub.3.sup.2 and thus the electrochemistry of Au and Tl in the Na.sub.2SO.sub.3 supporting electrolyte was briefly examined. Cyclic voltammetry experiments performed on freshly prepared Au thin film electrodes in the Na.sub.2SO.sub.3 electrolyte are shown in
(169) To further examine the influence of Tl on the Au deposition reaction a voltammetric study in 10 mol/L Tl.sub.2SO.sub.4+0.32 mol/L Na.sub.2SO.sub.3 was performed. A voltammogram collected after holding the electrode at 1.4 V.sub.SSE for 20 min reveals the reversible potential for deposition and dissolution of bulk Tl at 1.3 V.sub.SSE followed by oxidation of underpotential deposited (upd) Tl near 0.93 V.sub.SSE (
(170) A second oxidation wave is evident between 0.7 V.sub.SSE and 0.4 V.sub.SSE in
(171) Since oxidation of Tl upd leads to the formation of its hydroxide and oxide, the adsorption of Tl.sup.+ on a fresh Au surface is an irreversible process. Nevertheless, the adsorbed Tl species still gives rise to surface confined voltammetric waves associated with the respective structural changes during cycling, albeit accompanied by some limited dissolution and readsorption of the Tl.sub.ads species. The net accumulation of Tl.sub.ads/TOH.sub.ads species was followed by monitoring the evolution of a fresh Au surface during rapid voltammetric cycling, 0.4 V/s, in the presence of 20 mol/L Tl.sup.+ (
(172) The potential dependence of the Tl upd adsorption was examined by holding fresh Au electrodes at different fixed potentials for 5 minutes followed by rapid voltammetric oxidation to assay the Tl surface coverage. As shown in
(173) The integrated oxidation charge between 1.3 V.sub.SSE and 0.35 V.sub.SSE is shown in
(174) Analogous stripping (more precisely, oxidation) experiments after holding the potential at 1.0 V.sub.SSE for various times were used to examine the time dependence of Tl accumulation. The time dependence evident in
(175) Upon oxidation a wide variety of behaviors are evident. The peak near 1.2 V.sub.SSE is suggestive of bulk Tl dissolution. The oxidation peak centered near 0.9 V.sub.SSE is consistent with that observed in
(176) Electrodeposit Morphology
(177) A series of Au films were electrodeposited from an electrolyte containing 0.080 mol/L Na.sub.3Au(SO.sub.3).sub.2+0.32 mol/L Na.sub.2SO.sub.3. The deposit surfaces were imaged by scanning electron microscope after 20 min of growth at the potentials used for the chronoamperometry in
(178) In contrast to the additive-free electrolyte, Au deposits from the Tl.sup.+ containing electrolytes exhibit optically bright, smooth surfaces at 0.9 V.sub.SSE (
(179) Further increase of adsorbate coverage through 1.05 V.sub.SSE is associated with decreased deposition current despite a deposit surface area that is orders of magnitude larger. It has been suggested that the threshold coverage beyond which adsorbed Tl no longer accelerates Au deposition is approximately 30% of the saturation coverage F=1.610.sup.9 mol/cm.sup.2, the deposition rate decreasing both above and below this value. The fractional coverage stripped after holding at 0.9 V.sub.SSE in the electrolyte containing 20 mol/L Tl.sup.+ (
(180) A transition to faceted dendrite-like structures analogous to those observed with the additive-free electrolyte is seen at 1.1 V.sub.SSE (
(181) Three regimes of behavior are thus identified for Au deposition from the Tl containing electrolyte. At small overpotentials deposition is characterized by smooth surfaces and rising current transients indicating that Tl adsorption accelerates the deposition rate and, by virtue of the CEAC mechanism, leads to smoothing of the deposit. Deposition at more negative potentials falling within the NDR region yields nanowires of well-defined dimension. As per
(182) The correlation of nanowire deposition and potentials within the NDR regime was probed further by planar and cross-section examination of deposits grown in electrolyte containing 20 mol/L Tl.sup.+. As shown in
(183)
(184) The microstructures of the nanowires were examined using transmission electron microscopy. Dark field imaging and selected area diffraction of wires grown at 1.0 V.sub.SSE for 10 min in electrolyte containing 20 mol/L Tl.sup.+ shows that they are single crystals, with growth axes directed along 110
directions, and free of twin boundaries throughout their length (
(185) The single crystal 110
oriented nanowires differ from free standing Au nanowires fabricated using colloidal surfactant-based processes. These include single crystal nanowires that are elongated in the
100
direction and
112
oriented nanowires that contain a (111) twin plane along the growth axis, the latter microstructure also seen with Ag nanowires produced in an electroless process. Gold nanowires that do grow in the
110
direction are 5-fold twinned rather than single crystal as obtained here, although the origins of internal microstructure and nanowire growth direction remain a topic of discussion. That said,
110
growth habits are observed with lower aspect ratio protrusions on branched Au structures such as stars or spikes. Given the branched geometry at the 3-fold rotationally symmetric junctions from which the wires originate this is not unreasonable.
(186) The epitaxial growth of the nanowires on the thin film substrate indicates that control of the substrate crystallography should yield uniform nanowire orientations. As shown in
(187) Au deposition from the sulfite electrolyte exhibits a range of potential dependent growth morphologies, and additional geometries arise upon inclusion of Tl.sup.+ in the electrolyte. Highly faceted, stacking fault-rich, vertically oriented nanowires can be grown in the additive-free Au electrolyte at small overpotentials (i.e., at 0.9 V.sub.SSE) according to prior literature; analogous structures are seen here in
(188) Addition of Tl.sup.+ to the plating electrolyte results in significant depolarization of the electrodeposition reaction. The acceleration of Au deposition at 0.9 V.sub.SSE and positive of it is associated with a dilute coverage of upd Tl species. The impact of Tl.sub.ads on the metal deposition reaction at these potentials is captured by the rising chronoamperometric transient; for deposition at 0.9 V.sub.SSE the rise time in the electrolyte with 10 mol/L Tl.sup.+ (
(189) At more negative potentials the Tl.sub.ads changes to an inhibitor of the Au deposition process. In combination, the potential dependent Au deposition kinetics (
(190) The most interesting finding in the present study is the bifurcation of the electrode into active and passive regions associated with deposition at potentials within the NDR voltammetric regime. On a planar electrode the growth velocity, v, surfactant coverage, , relationship exhibits a peak deposition rate just prior to the onset of the NDR. Beyond this coverage the surfactant becomes an inhibitor as indicated schematically in
(191) Nanowire development at 1.0 V.sub.SSE begins with a population of 130 nuclei/m.sup.2 (estimated from
(192) Extension of a nanowire leads to an increase in electrode area, A, associated with the newly created cylindrical sidewalls. Maintenance of the nanowire growth geometry involves passivation of the newly created sidewall by Tl.sup.+ accumulation from the electrolyte to form a blocking monolayer. The reasonableness of this mechanism is assessed through comparison of the diffusion limited flux of Tl.sup.+ in the electrolyte to the growing Au nanowire tip and the adsorbate coverage required for passivation of the surface area as it is created.
(193) Beyond mass balance, the mechanistic basis for Tl segregation to the sidewalls deserves some discussion. Segregation of the Tl co-deposited with Au via underpotential deposition onto the nanowire surface is supported by the intrinsic immiscibility of Au and Tl. Accelerated growth on the nanowire surface is proposed to occur through dilution of the suppressing Tl coverage by area expansion on the higher (compound) curvature tip. While the positive curvature of the growing tip will also lead to a negative shift of the reversible potential for a given equilibrium surface coverage of Tl, the effect is expected to be small given the (relatively) large 25 nm radius of curvature.
(194) Because the nanowires grow with a fixed orientation, anisotropy of the thermodynamics and kinetics of Tl upd, alloying and surface segregation and the associated impact on Au(SO.sub.3).sub.2.sup.3 reduction necessarily plays a role in shape evolution. Indeed, the substantial suppression of Au deposition evident in the voltammetry and the anisotropic crystallite growth obtained in the Tl-free sulfite electrolyte at less negative potentials (i.e., smaller overpotentials) suggest that the mechanism of nanowire growth by facet-dependent suppression or reaction kinetics applies to this system. Such anisotropy has been previously noted in a study of Tl upd on the three primary low index surface of Au in alkaline electrolyte. As such, the proposed mechanism of sidewall suppression by Tl adsorption has direct analogy in the commonly used Ag(I) assisted approach for growth of single crystal 100
oriented nanowires although the Ag.sup.+ in that approach is at much higher concentrations (similar to that of the Au.sup.+ this study); as with Tl here, Ag upd on the sides of the growing Au nanowires retards deposition on those surfaces. The mechanism by which the Tl.sup.+ initiates and maintains the nanowire growth along the
110
growth axes calls for further exploration.
(195) Deposition of Au from a Na.sub.3Au(SO.sub.3).sub.2Na.sub.2SO.sub.3 electrolyte has been studied as a function of potential and Tl.sub.2SO.sub.4 additions. Gold deposition from Au(SO.sub.3).sub.2.sup.3 is intrinsically slow at least partly due to SO.sub.3.sup.2 adsorption at potentials positive of 0.8 V.sub.SSE. The addition of Tl.sup.+ to the base electrolyte results in depolarization of the Au deposition and development of hysteretic voltammetry and an NDR regime. Electroanalytical studies of Tl underpotential and overpotential deposition reveal a close connection between Tl coverage and the resulting morphology of the Au deposits. Electrodeposition in the presence of limited Tl coverage at modest overpotentials for Au deposition, 0.8 V.sub.SSE and 0.9 V.sub.SSE, is accelerated and exhibits optically bright surfaces consistent with the CEAC mechanism of surface smoothing. Deposition at potentials that lie within the voltammetric NDR region, e.g. 1.0 V.sub.SSE, yields a dense array of high aspect ratio, single crystal gold nanowires with diameters of approximately 50 nm and axes along 110
directions. The nanowires are epitaxially oriented on the {111} textured substrates and grow from 111 oriented nuclei with three-fold symmetry reflecting anisotropic deposition along the
110
directions. The transition from CEAC based compact film growth to nanowire deposition is correlated with, and explained by, the surfactant Tl coverage and its effect on the rate of Au deposition.
Comparative Example D. Superconformal Bottom-Up Gold Deposition in High Aspect Ratio Through Silicon Vias
(196) This comparative example provides superconformal, bottom-up Au filling of high aspect ratio through silicon vias (TSVs) along with a predictive framework based on the coupling of suppression breakdown and surface topography. The work extends a previous study of superconformal Au deposition in lower aspect ratio TSVs. Deposition was performed in a Na.sub.3Au(SO.sub.3).sub.2 electrolyte containing a branched polyethyleneimine (PEI) deposition-rate suppressing additive, dissimilar from the superconformal filling composition described in the Detailed Description and Example.
(197) Voltammetric measurements using a rotating disk electrode (RDE) were used to assess the impact of the PEI suppressor concentration and transport on the rate of metal deposition, enabling the interplay between metal deposition and suppressor adsorption to be quantified. The positive feedback associated with suppression breakdown gives rise to an S-shaped negative differential resistance (S-NDR). The derived kinetics for suppressor adsorption and consumption were used in a mass conservation model to account for bottom-up filling of patterned features. Predictions, including the impact of deposition potential and additive concentration on feature filling, are shown to match experimental results for filling of TSVs. This further generalizes the utility of the additive derived S-NDR model as a predictive formalism for identifying additives capable of generating localized, void-free filling of TSVs by electrodeposition.
(198) While Cu is now the dominant metal used for on-chip interconnects, Au is still the metal of choice for contacts to wide bandgap semiconductors. Through-mask plating is widely used for forming these simple Au structures, with deposition proceeding from a common back contact with a layer of patterned photoresist defining the plated area. However, processes that achieve superconformal filling of damascene patterned features, such as those that dominate manufacture of on-chip Cu interconnects from the nanoscale to the microscale, are desirable because of the versatility offered by the Damascene process to build dense 3-D metallization networks of arbitrary complexity. Superconformal Au filling of micrometer size trenches has, in fact, been demonstrated using processes based on the use of a rate suppressing additive guided by the linear theory of leveling as well as the curvature enhanced accelerator coverage (CEAC) mechanism that involves competition between rate suppressing and accelerating additives. More recently, an effort to fill much larger TSVs using an electrolyte that exhibited non-linear S-shaped Negative Differential Resistance (S-NDR) was reported. The system gives rise to distinctive partitioning of the work piece into active and passive regions whence, for 3-D patterned structures, deposition on the free surface is blocked while Au is actively deposited within the recessed features, the transition from passive to active behavior within the filling features being a sensitive function of the process control parameters. Previously, the S-NDR mechanism has been shown to explain extreme bottom-up filling of Cu as well as the sidewall passive-to-active transition in Zn, Co, and Ni, including accurate quantitative prediction. However, no quantitative modeling has been applied to the Au system.
(199) Transport, adsorption, deactivation and surface area change all can impact the coverage of adsorbed additives that more broadly underlie superconformal feature filling. However, S-NDR based models implemented thus far only required the first three effects to accurately predict the temporal and spatial evolution of deposition. The superconformal filling profiles associated with S-NDR systems arise from competition between metal passivation by suppressor adsorption and surface activation through suppressor deactivation. The resulting nonlinear dependence of suppressor coverage on suppressor concentration couples with a gradient of suppressor concentration down filling features to localize deposition at, or toward, the bottoms of filling features. Both the side-wall passive-to-active transition and exclusive bottom-up geometries may be captured by S-NDR models, with critical behavior associated with the suppression breakdown underlying the highly nonlinear dependence of deposition rate on position in all cases. Importantly, such highly localized, sharp transitions cannot be derived from older leveling models based on linear relationships between deposition rates, suppressor coverage and suppressor concentration.
(200) The additive branched polyethyleneimine (PEI) is used as the suppressor; however, in the present case a five-fold larger molecular weight is used. Electroanalytical measurements are used to parameterize an S-NDR model that enables quantitative prediction of the onset geometry of significant Au deposition within TSVs. This understanding is then combined with a programmed potential step wave form to achieve nearly void-free Au filling of the TSVs. Based on previous work with the Na.sub.3Au(SO.sub.3).sub.2-PEI electrolyte system, significant porosity evident in the microstructure can negatively impact the mechanical properties, including crack formation and propagation. Thus, close attention will be given to the impact of PEI on the microstructure of the deposits.
Experimental
(201) Depositions were conducted at room temperature in a cell containing 35 mL of 0.32 mol/L Na.sub.3Au(SO.sub.3).sub.2 electrolyte of pH 9.0. Additive was introduced to the electrolyte from a master solution of 10 mmol/L PEI with a molecular weight of 10,000 g/mol in 18 M.Math.cm water. The electrolyte was sparged with argon between electrochemical measurements to reduce the impact of dissolved oxygen. A Hg/Hg.sub.2SO.sub.4/saturated K.sub.2SO.sub.4 reference electrode (SSE) was connected to the working electrode compartment via a fritted salt bridge and all quoted potentials are relative to this electrode. All experiments were conducted without iR compensation. A platinum counter electrode was held in a fritted compartment immersed within the main cell.
(202) Voltammetry was conducted on Au rotating disk electrode (RDE) of 1.0 cm diameter (area 0.78 cm.sup.2). The RDE was polished with 1200 grade SiC paper and rinsed with 18 M.Math.cm water prior to each experiment. Voltammetry was performed at 2 mV/s at different RDE rotation rates: 25 rpm, 100 rpm, 400 rpm and 1600 rpm. Currents in electroanalytical measurements are converted to current densities using the projected RDE geometric area. However, as will be seen, the deposit area is generally ill-defined beyond suppression breakdown. Specifically, deposition in the additive-containing electrolyte is not uniform so that the current density reported under such circumstance represents a lower bound estimate. Conversely, the measured currents can also be amplified by deposition on asperities that project beyond the boundary layer.
(203) Feature filling was performed using fragments of wafers patterned with 56 m deep TSVs of annular cross-section (courtesy of IBM) having a 1 m thick Cu seed in the field and a lesser amount on the side walls. To give definition to the metal ion and additive transport, the patterned substrates rotated about one end from a Pt spindle during deposition, like a helicopter blade, the patterned surface facing upwards. Based on the 1 cm distance between the features and rotational axis for most of the imaged TSVs, the 100 rpm (200 rad/min) rotation rate corresponds to an estimated 10 cm/s flow rate over the surface. Pre-wetting with ethyl alcohol was used to displace air bubbles that were otherwise trapped in the TSVs during immersion in the electrolyte for Au deposition. Following immersion, the specimens were rotated at open circuit for 2 min prior to starting the metal deposition in order to displace the alcohol from the TSV by mixing with the bulk electrolyte. In most cases a 5 s voltage pulse at 1.5 V was then applied, just prior to feature filling, to improve nucleation on the Cu seeded TSVs.
(204) The TSVs were imaged optically after embedding them in epoxy and then cross-sectioning and polishing them on diamond lapping films down to 0.1 m grit size using standard techniques and equipment. A subset of specimens was also examined by scanning electron microscopy. These specimens were subjected to an additional cleaning of the surface using oblique 4 keV Ar.sup.+ to remove residual surface damage from the mechanical polishing prior to imaging.
(205) Deposition on Planar and Patterned Substrates
(206)
(207)
(208) Fitted curves are overlaid on the experimental data in
(209) Parasitic processes such as water reduction that can reduce the current density associated with the metal deposition below the measured current density are taken to be negligible in this system. The values of j.sup.o.sub.=1 and .sub.=1 were therefore obtained directly from the fit to the negative-going sweeps up to the critical potential. The difference of .sub.=1 from .sub.=0 is not expected from a simple blocking mechanism. Rather, the change of the charge transfer coefficient from 0.45 to 0.35 between the free and fully suppressed surfaces speaks to a change in the deposition mechanism.
(210) Fitting of the CVs to obtain the remaining kinetic parameters k.sub.+ and k.sub. involves integrating for (t) and mass balance between suppressor diffusion across the boundary layer and its adsorption onto the surface. Values for the areal density of sites , suppressor diffusion coefficient D.sub.s and boundary layer thickness (scaling with the rotation rate as .sup.0.5) were estimated. Metal ion depletion across the boundary layer, more significant here than in our recent Ni and Co studies with higher metal ions concentrations, was accounted for by balancing the molar volume weighted ion flux across the boundary layer with the deposition rate. This balance is most relevant when deposition approaches the metal ion transport limit.
(211) Evolution of current density and adsorbate coverage depends on a number of parameters. However, with values for the other parameters either known or otherwise estimated, k.sub.+ and k.sub. were adjusted to capture the dependence of the suppressor breakdown potential on the suppressor flux to the interface, as manifest in the concentration and RDE rotation rate dependences. Significantly, simulations over a wide range of k.sub.+ values can exhibit similar suppression breakdown potential if k.sub. is scaled appropriately. This is clearest for the case of near ideal suppression (i.e., v.sub.=0>>v.sub.=1.fwdarw.0) for which the unsaturated steady state suppressor coverage so that k.sub.+ and k.sub. enter explicitly only in the ratio k+kk+k. However, the coverage also depends on the suppressor and metal ion concentrations at the electrolyte/metal interface (C.sub.S and v.sub.=0, respectively) and thus the boundary layer thickness. Simultaneous fitting of the suppression breakdown potentials at the four rotation rates therefore further restricts the range of k.sub.+ and k.sub. values.
(212) Fitting of the suppression breakdown observed in the cyclic voltammetry for two different suppressor concentrations and four different rotation rates is shown in
(213) TSV Filling Simulations
(214) The distinct breakdown of inhibition in the voltammetric curves reflects the emergence of a two state active-passive system, the critical potential representing the limiting condition for balance between the concentration-dependent rate of suppressor adsorption onto the uniform, planar surface and the potential-dependent rate of adsorbate deactivation. Decrease of the suppressor concentration within recessed features effectively introduces a position-dependent positive shift of the critical potential from that at the free-surface. For experiments of most interest the passive-to-active transition occurs at the location within the feature where the critical potential equals the actual applied potential.
(215) Previous models of Cu bottom-up TSV filling through the additive derived S-NDR mechanism evaluated both the metal ion and the suppressor concentration distributions within the boundary layer and the TSV, capturing experimental observations. These entailed full spatial and temporal analysis or a pseudo steady state, one-dimensional analysis that took advantage of experimental TSV fill times that were substantially longer than diffusional relaxation times. This comparative example, like the earlier Co and Ni studies, uses the pseudo steady state approach, neglecting the impact of the transient and spatial evolution of the actual current distribution. Likewise, although this is a less robust approximation due to the lower metal ion concentration in the present Au electrolyte, metal ion depletion is again neglected in the simulations yielding the location of the passive-to-active transition. The general uniformity of the metal deposited within the TSVs at less negative potentials (
(216)
(217) The differential equation, boundary conditions, suppressor adsorption, and metal deposition interactions permit simulation of deposition profiles using the kinetic parameters obtained from the RDE experiments. Suppressed deposition is predicted in the upper portion of the TSVs with a micrometer scale transition to unsuppressed deposition in the lower portion of the TSV at depth d.sub.s down the TSV that is consistent with experimental observations (and analogous to model predictions detailed previously with Ni and Co filling).
(218)
(219)
(220) Potential Mediated TSV Filling Processes
(221) Because the applied potential determines the location where suppression fails within the filling feature, metal deposition can be localized toward the bottom of the TSV and run at the transport limit without risk of pinch-off from deposition higher up. As demonstrated previously with Ni and Co, void-free filling can be accomplished by adjusting the applied potential over time in order to progressively actuate deposition higher in the TSV as appropriate for the evolution of filling farther down. As in the ferrous metal studies, both stepped and ramped potentials are examined here.
(222)
(223) These three differences have significant impact on the nature of filling and the deposit itself. The deposit roughness is a source of nanoscale voids within the deposit and where the sidewall deposits impinge (
(224)
(225)
(226) The superconformal Au deposition in the PEI-containing sulfite electrolyte shows localized deposition exhibiting a passive-to-active transition within the TSVs. Unsuppressed active metal deposition starts at a distance down the filling feature that is defined by the suppressor concentration and applied potential, metal deposition associated with a slower leakage through the suppressor layer occurring at the slower passive rate above this location. The geometry is analogous to that previously noted during Au deposition in lower aspect ratio square TSVs as well as Ni and Co deposition in similar annular TSVs, all using a smaller molecular weight version of the PEI suppressor. The S-NDR model previously applied to both the Ni and Co systems once again provides reasonably accurate prediction of the position of the passive-to-active transition as a function of the applied potential. Also as previously demonstrated with those systems, waveforms that progressively move the active region further up the feature enable filling of the TSVs without the keyhole voids that would otherwise result from metal ion depletion effects. It is expected that filling times can be reduced below those shown for both the stepped and ramped cases by further optimization of the waveforms.
(227) Regarding the quality of the Au deposits, growth in the neat sulfite electrolyte is generally rough and dendritic. While potential control of the passive-to-active transition associated with the S-NDR mechanism of superconformal filling prevents pinch-off and macroscale void formation, the intrinsically rough active deposits in the essentially suppressor-free lower portion of the TSV do incorporate nanoscale porosity. Furthermore, as seen previously with the Co S-NDR filling system, the material deposited on passivated surfaces is highly porous. The not insignificant passive deposition rate in the PEI-containing electrolyte yields a significant thickness of porous Au on the sidewalls higher in the TSV. Nonetheless, this work further demonstrates the broad applicability of superconformal S-NDR based feature filling processes, the power of S-NDR based models for predicting the feature filling evolution based on electroanalytical measurements that can be used to assess systems for improved feature filling and deposit quality.
(228) Superconformal Au deposition in annular TSVs has been experimentally demonstrated and predicted using a simple S-NDR model. These results extend previously demonstrated superconformal deposition in the PEI-containing sulfite electrolyte to higher aspect ratio features and demonstrate that the filling can be quantitatively predicted using kinetics acquired from simple electroanalytical measurements using a rotating disk electrode. The cyclic voltammetry, was used to assess the impact of PEI concentration and transport on the suppression breakdown potential. The results were analyzed to obtain kinetics used in an S-NDR based model to simulate superconformal Au deposition in the annular TSVs. The model predicts the experimentally observed passive-to-active transition within the TSVs, including its location as a function of both suppressor concentration and applied potential. Void-free feature filling, not possible for fixed potential deposition, was achieved using potential waveforms that take advantage of the potential-dependent suppression depth and geometric leveling to produce superconformal filling. These results, combined with previously detailed superconformal Cu, Zn, Ni and Au filling of TSVs, demonstrate the broad applicability of the S-NDR based mechanism for superconformal feature filling and S-NDR based models for prediction of feature filling.
(229) While one or more embodiments have been shown and described, modifications and substitutions may be made thereto without departing from the spirit and scope of the invention. Accordingly, it is to be understood that the present invention has been described by way of illustrations and not limitation. Embodiments herein can be used independently or can be combined.
(230) Reference throughout this specification to one embodiment, particular embodiment, certain embodiment, an embodiment, or the like means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. Thus, appearances of these phrases (e.g., in one embodiment or in an embodiment) throughout this specification are not necessarily all referring to the same embodiment, but may. Furthermore, particular features, structures, or characteristics may be combined in any suitable manner, as would be apparent to one of ordinary skill in the art from this disclosure, in one or more embodiments.
(231) All ranges disclosed herein are inclusive of the endpoints, and the endpoints are independently combinable with each other. The ranges are continuous and thus contain every value and subset thereof in the range. Unless otherwise stated or contextually inapplicable, all percentages, when expressing a quantity, are weight percentages. The suffix (s) as used herein is intended to include both the singular and the plural of the term that it modifies, thereby including at least one of that term (e.g., the colorant(s) includes at least one colorants). Optional or optionally means that the subsequently described event or circumstance can or cannot occur, and that the description includes instances where the event occurs and instances where it does not. As used herein, combination is inclusive of blends, mixtures, alloys, reaction products, and the like.
(232) As used herein, a combination thereof refers to a combination comprising at least one of the named constituents, components, compounds, or elements, optionally together with one or more of the same class of constituents, components, compounds, or elements.
(233) All references are incorporated herein by reference.
(234) The use of the terms a and an and the and similar referents in the context of describing the invention (especially in the context of the following claims) are to be construed to cover both the singular and the plural, unless otherwise indicated herein or clearly contradicted by context. Or means and/or. Further, the conjunction or is used to link objects of a list or alternatives and is not disjunctive; rather the elements can be used separately or can be combined together under appropriate circumstances. It should further be noted that the terms first, second, primary, secondary, and the like herein do not denote any order, quantity, or importance, but rather are used to distinguish one element from another. The modifier about used in connection with a quantity is inclusive of the stated value and has the meaning dictated by the context (e.g., it includes the degree of error associated with measurement of the particular quantity).