Modern electrochemical energy conversion devices require more advanced proton conductors for their broad applications. Phosphonated polymers have been proposed as anhydrous proton conductors for fuel ...cells. However, the anhydride formation of phosphonic acid functional groups lowers proton conductivity and this prevents the use of phosphonated polymers in fuel cell applications. Here, we report a poly(2,3,5,6-tetrafluorostyrene-4-phosphonic acid) that does not undergo anhydride formation and thus maintains protonic conductivity above 200 °C. We use the phosphonated polymer in fuel cell electrodes with an ion-pair coordinated membrane in a membrane electrode assembly. This synergistically integrated fuel cell reached peak power densities of 1,130 mW cm
at 160 °C and 1,740 mW cm
at 240 °C under H
/O
conditions, substantially outperforming polybenzimidazole- and metal phosphate-based fuel cells. Our result indicates a pathway towards using phosphonated polymers in high-performance fuel cells under hot and dry operating conditions.
The operation of fuel cells under low relative humidity (RH) conditions gives substantial cost and performance benefits. Nonetheless, it is not currently feasible to operate anion exchange membrane ...fuel cells (AEMFCs) at low RH conditions because current materials for membrane electrode assembly cannot provide sufficient water for the oxygen reduction reaction. Here we synthesized polyfluorene ionomers with different ammonium concentrations for anode and cathode to control water management. We designed several asymmetric electrodes that enable high performance under low RH conditions
via
not only conventional backward water diffusion (anode to cathode) but also forward diffusion (cathode to anode). The AEMFCs using optimized asymmetric electrodes exhibited high H
2
/CO
2
-free air performance (rated power density of circa 540 mW cm
−2
at 90 °C under 75% (anode) and 50% RH (cathode) conditions), which is comparable to those of state-of-the-art AEMFCs under nearly water-saturated conditions. The durability of the AEMFCs is excellent, generating 0.6 A cm
−2
for >900 h at 80 °C under 50% RH (cathode) conditions. This study demonstrates that high-performance and durable AEMFCs under low RH and high current generating conditions are possible.
An asymmetric anion exchange membrane fuel cell enables high performance and operational durability under low RH conditions.
The operation of fuel cells under low relative humidity (RH) conditions gives substantial cost and performance benefits. Nonetheless, it is not currently feasible to operate anion exchange membrane ...fuel cells (AEMFCs) at low RH conditions because current materials for membrane electrode assembly cannot provide sufficient water for the oxygen reduction reaction. In this work, we synthesized polyfluorene ionomers with different ammonium concentrations for anode and cathode to control water management. We designed several asymmetric electrodes that enable high performance under low RH conditions via not only conventional backward water diffusion (anode to cathode) but also forward diffusion (cathode to anode). The AEMFCs using optimized asymmetric electrodes exhibited high H2/CO2-free air performance (rated power density of circa 540 mW cm-2 at 90 °C under 75% (anode) and 50% RH (cathode) conditions), which is comparable to those of state-of-the-art AEMFCs under nearly water-saturated conditions. The durability of the AEMFCs is excellent, generating 0.6 A cm-2 for >900 h at 80 °C under 50% RH (cathode) conditions. This study demonstrates that high-performance and durable AEMFCs under low RH and high current generating conditions are possible.
Modern electrochemical energy conversion devices require more advanced proton conductors for their broad applications. Phosphonated polymers have been proposed as anhydrous proton conductors for fuel ...cells. However, the anhydride formation of phosphonic acid functional groups lowers proton conductivity and this prevents the use of phosphonated polymers in fuel cell applications. Here, we report a poly(2,3,5,6-tetrafluorostyrene-4-phosphonic acid) that does not undergo anhydride formation and thus maintains protonic conductivity above 200 °C. We use the phosphonated polymer in fuel cell electrodes with an ion-pair coordinated membrane in a membrane electrode assembly. We find that this synergistically integrated fuel cell reached peak power densities of 1,130 mW cm-2 at 160 °C and 1,740 mW cm-2 at 240 °C under H2/O2 conditions, substantially outperforming polybenzimidazole- and metal phosphate-based fuel cells. Our result indicates a pathway towards using phosphonated polymers in high-performance fuel cells under hot and dry operating conditions.
Copper (Cu) electrodeposition (ECD) in through-silicon-vias (TSVs) is an essential technique required for high-density 3-D integration of complex semiconductor devices. Cu ECD most commonly utilizes ...an acid sulfate electrolyte containing a finely tuned three-additive combination of inhibiting and Cu
2+
aquo complex catalyzing species to produce void free deposits through the depth of high aspect ratio features. Despite their proven efficacy in filling micro- and nanoscale features, these three-additive electrolytes have not been shown to be viable for filling ICs at the mesoscale (> 100 µm). This is likely explained by the seminal curvature enhanced adsorbate coverage (CEAC) mechanism that describes the adsorption/desorption action of the additive species to an electrode surface.
1, 2
This mechanism theorizes that the interaction between the constituent additives and the rate of Cu
2+
reduction is highly dependent on the local curvature of the electrode surface. In addition, these systems often require finely tuned ECD techniques like the use of pulse plating regimes, tightly controlled additive concentrations, and relatively low applied current densities (< 10 mA/cm
2
).
To overcome the deficiencies of the three-additive acid sulfate chemistry that the CEAC mechanism describes, a single-additive chemistry was developed for the purposes of filling conformally conductive high aspect ratio features exclusively from the bottom-up by Moffat, T.P. and Josell, D.
3-5
This chemistry relies on a specific polyether suppressor additive, which induces voltammetric hysteresis and a positive feedback action as metal deposition disrupts Cl
-
-suppressor complexes at the cathode surface. The acid sulfate resistive Cu electrolyte developed by Moffat, T.P. and Josell, D. utilizes 1 M CuSO
4
+ 0.5 M H
2
SO
4
in addition to µM level concentrations of chloride and polyether suppressor species.
3-5
To investigate the functionality of a novel chemistry, the work outlined herein relies on the same mechanism and utilizes the same polyether suppressor and chloride concentration ranges, but instead uses a 1.25 M CH
3
SO
3
H + 0.25 M H
2
SO
4
electrolyte. This electrolyte exhibits a greater Cu solubility, requires a lower concentration of acid, thus increasing the Cu plating rate while still retaining the same S-NDR mechanics as developed by Moffat, T.P. and Josell.
3-5
Upon adoption of the above-described chemistry, several issues were explored. The importance of convection on plating profiles within mesoscale (600 μm depth, 5:1 aspect ratio) blind TSVs is examined. In addition, potentiostatic methods are used to explore the process window for achieving bottom-up feature filling in the mesoscale TSVs described above. However, these potentiostatic techniques rely on the use of a reference electrode in solution, which is not a viable setup in industrial wafer-scale production tools. Thus, a galvanostatic approach is outlined and shown to be a viable alternative based on resulting fill profiles. The values used for the current density were calculated by comparing the breakdown potential of the suppressor molecule with the current reading at that set potential. Using only the area of the feature bottoms, a current density was calculated and fixed during subsequent plating experiments. This method is potentially problematic as the non -referenced electrochemical potential will maintain some unknown hysteresis, but initial results are shown to be promising. This galvanostatic approach, using the novel single additive chemistry is currently being explored for subsequent implementation into full wafer plating apparatus.
Supported by the Laboratory Directed Research and Development program at Sandia National Laboratories, a multimission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC., a wholly owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA-0003525.
Andricacos, P.C., et al.,
Damascene copper electroplating for chip interconnections.
Ibm Journal of Research and Development, 1998. 42(5): p. 567-574.
Moffat, T.P., et al.,
Superconformal film growth: Mechanism and quantification.
Ibm Journal of Research and Development, 2005. 49(1): p. 19-36.
Moffat, T. P., and D. Josell. "Extreme bottom-up superfilling of through-silicon-vias by damascene processing: suppressor disruption, positive feedback and turing patterns."
Journal of The Electrochemical Society
159.4 (2012): D208-D216.
Josell, D., D. Wheeler, and T. P. Moffat. "Modeling extreme bottom-up filling of through silicon vias."
Journal of The Electrochemical Society
159.10 (2012): D570-D576.
Wheeler, Daniel, Thomas P. Moffat, and Daniel Josell. "Spatial-temporal modeling of extreme bottom-up filling of through-silicon-vias."
Journal of The Electrochemical Society
160.12 (2013): D3260-D3265.
The incorporation of copper-filled, mesoscale through-silicon-vias (TSVs) as a three-dimensional integration technique allows for increased input/output per unit volume compared to wire bonded ...microelectronic devices. Mesoscale TSVs that span the full thickness of a silicon wafer allow for preservation of a large radius of curvature during multi-layer device fabrication and for conservation of mass in the device, which is often a requirement in MEMS applications. The damascene process is commonly implemented for Cu electrodeposition in high aspect ratio features and typically involves a three-additive system in a CuSO
4
-H
2
SO
4
electrolyte. However, a single-additive system has recently been shown capable of achieving bottom-up superfilling in features with higher aspect ratios than those commonly used in damascene interconnects.i This resistive electrodeposition system utilizes CuSO
4
, methane sulfonic acid (MSA), chloride, and a poloxamine suppressor additive. Cyclic voltammetry (CV) can be used to characterize the plating solution and identify a potential range that exhibits hysteresis in the voltage-current relationship that is caused by suppressor disruption at the cathode surface during electrodeposition. This hysteretic region serves as an operating window where void-free Cu filling of high aspect ratio features may be achieved.
Using this chemistry, potentiostatic and galvanostatic plating conditions for void-free filling were developed for 100 μm diameter vias that are 600 μm deep, a 6:1 aspect ratio.ii,iii Cu electrodeposition in TSVs with a higher aspect ratio of 10:1 is currently under investigation. These TSVs have a 62.5 μm diameter etched into a 625 μm thick silicon-on-insulator (SOI) wafer. Conditions that produce void free, bottom-up filling for 100 μm diameter TSVs have not directly translated to fill 62.5 μm diameter TSVs. In this work, the poloxamine suppressor concentration, current, potential, and sample rotation rate were varied to analyze their effect on the fill profile in these 62.5 μm diameter TSVs. Fill profiles were analyzed after mechanically cross-sectioning and optically imaging the samples. Cu fill profiles were also obtained through X-ray computed tomography (CT) scans. This work details the experimental approach associated with determining galvanostatic and potentiostatic plating conditions for 62.5 μm diameter TSVs and presents the resultant fill profiles of Cu in these vias.
Sandia National Laboratories, a multimission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC., a wholly owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA-0003525.
This paper describes objective technical results and analysis. Any subjective views or opinions that might be expressed in the paper do not necessarily represent the views of the U.S. Department of Energy or the United States Government.
i. Moffat, T. P., and D. Josell. "Extreme bottom-up superfilling of through-silicon-vias by damascene processing: suppressor disruption, positive feedback and turing patterns."
Journal of the Electrochemical Society
159.4 (2012): D208-D216.
ii. Menk, L. A. et. al. “Bottom-Up Copper Filling of Large Scale Through Silicon Vias for MEMS Technology.”
Journal of the Electrochemical Society
166.1 (2019): D3066-D3071.
iii. Menk, L. A. et. al. “Galvanostatic Plating with a Single Additive Electrolyte for Bottom-Up Filling of Copper in Mesoscale TSVs”
Journal of the Electrochemical Society
166.1 (2019).
Developments in through silicon via (TSV) fabrication and filling have allowed for significant advancements in the microelectronics industry by enabling compact 3D integration. Copper, due to its ...desirable electrical properties, is the most commonly used metal for TSVs filling. TSVs are filled electrochemically by selectively reducing solvated metal ions in an electrolyte solution to confined TSV geometries via the addition of organic and inorganic additives (halides, suppressors, accelerators, levelers). Optimization of this mechanism can often be temperamental requiring alterations in the solution’s chemical makeup as well as electrical bias and fluid flow parameters. Therefore, consistent and reliable methods of characterizing electrochemical copper fill quality is critical to provide feedback necessary for process adjustments. Historically, characterizing TSV fill profile has relied on physical cross sectioning by mechanical grinding, planarization and polishing (Figure 1). These methods are effective in that they offer direct visualization of the TSV fill quality; however, they have certain inherent drawbacks. The most critical drawback is that mechanical cross sectioning is a destructive process and can only be used in process development and not on deliverable parts. Also, mechanical material removal can often be time consuming and inconsistent requiring significant effort by process technicians to generate desirable results. Implementing non-destructive methods for characterization is necessary for the advancement of TSV technology and has been explored in an effort to make TSV research more streamline and efficient. Recent work at Sandia National Laboratories has evaluated non-destructive TSV characterization processes through X-Ray Computed Tomography (CT), digital radiographs and X-ray Phase Contrast Imaging (XPCI). The benefit of X-ray evaluation of TSV filling is batch level processing of many TSVs at once (Figure 2A), digital model rendering for multidimensional analysis (Figure 2B), and complete preservation of the physical integrity of test pieces and deliverables. Each nondestructive technique for TSV characterization will be directly compared and contrasted to conventional grinding and polishing on the merits of defect detection, resolution, time consumption and reproducibility. This talk will outline the development of non-destructive methods for TSV characterization and evaluate the reliability and consistency of these approaches.
Sandia National Laboratories, a multimission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC., a wholly owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA-0003525.
This paper describes objective technical results and analysis. Any subjective views or opinions that might be expressed in the paper do not necessarily represent the views of the U.S. Department of Energy or the United States Government.
Figure 1
Multi project-wafers (MPWs) offer cost-effective options for fabrication of application specific integrated circuits (ASICs) produced for entities not requiring a large volume of die from a dedicated ...full wafer run at an advanced semiconductor foundry. MPWs consist of a mixture of different devices shared on a single wafer; thereby reducing cost of processing by allowing the purchase of real estate allotted to each device. This practice is advantageous for those in academia or government that often only require small quantities of devices for research or niche applications. With many teams investing in a single wafer run, strict design rules must be followed on these microelectronic chips. In particular, the metal used in the final routing layer must be AlCu. This AlCu metallization is not compatible with solder connections for flip chip bonding applications because solder is unable to wet AlCu. To allow MPW die integration with flip chip bonding applications, a solder compatible under bump metallization (UBM) must be deposited on top of the AlCu. This UBM deposition can be achieved through physical vapor deposition and subsequent electrolytic plating into a photolithography defined plating mold. Unfortunately, photolithographic processes are only compatible with full wafers and not with singulated die. This leaves electroless deposition methods such as immersion tin or zinc followed by electroless nickel, electroless palladium, immersion gold (ENEPIG) or electroless nickel and immersion gold (ENIG) as the only options for depositing a wettable UBM on MPW die. Electroless deposition can be performed with a single die at a time, however this process is not well suited for manufacturing. In this work, we have developed a scheme for batch level UBM deposition on singulated semiconductor die by coupling temporary bonding with electroless deposition.
A unique batch approach is proposed to electrolessly deposit UBM to terminal AlCu on MPW die. Preliminary Si sidewall passivation is used to eliminate metallization of unintended surfaces in subsequent steps. Temporary custom fixturing onto a carrier substrate is leveraged allowing a variety of devices to be processed like a full wafer, establishing feasibility for industrial scaling. The bulk aluminum oxide is then solubilized by chemical etching and replaced by reduction of Sn as a base layer for ENIG and ENEPIG processes as seen in Figure 1. After metal deposition, devices are liberated from the temporary mounting substrate making them ready for flip chip bonding. Due to creating an UBM through electroless processes, options for interposer bumps are limited and outlined in Figure 2. Solder bonds from Cu pillars topped with SnAg or electroplated with either AuSn or SnAg allow a reflow solder attachment to complete electrical contact between the two chips. Electroplated Au bumps allow for a thermocompression bond to be made between the two chips. Both methods have benefits based on circumstance. Reflow solder is advantageous for fragile chips that cannot withstand the high pressure associated with thermocompression. Cu pillars are also attractive for RF devices in that Cu offers reduced electrical parisitics. Thermocompression is advantageous because it does not require high level of uniformity in the final Au bump. In this presentation, we will provide details on the passivation, batch level handling process, uniformity, and yield statistics for the electroless deposition.
Supported by the Laboratory Directed Research and Development program at Sandia National Laboratories, a multimission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC., a wholly owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA-0003525.
Figure 1
Electroforming Nanostructured-Based Thermoelectrics Arrington, Christian L.; Sharma, Peter A.; Burleigh, T.D. ...
Meeting abstracts (Electrochemical Society),
08/2014, Letnik:
MA2014-02, Številka:
15
Journal Article
Odprti dostop
Thermoelectric (TE) materials offer engineering solutions for applications such as cooling without moving parts and conversion of waste heat to electricity. However, devices using existing ...thermoelectric materials suffer from poor efficiency and high cost. Electrochemical deposition (ECD) has the potential to make the synthesis of thermoelectrics more cost effective and potentially enhance thermoelectric properties. The best thermoelectrics to date are thin film super lattices, fabricated using vapor phase methods 1. However, these materials have a greatly limited application space due to their limited thickness (~ 10 µm). In this work, ECD techniques are used to nanostructure Bi
2
Te
3
based
thermoelectric materials over large areas and thicknesses, as well as electroforming it into a mold to enable future batch fabrication. Multiple experiments were performed to better understand scalability of thermoelectric film growth. Large grain boundaries were found to affect density and stoichiometry of the material impacting TE performance. A 7 liter chemistry make-up and ECD conditions are established targeting wafer scale batch fabrication. The highest Seebeck, 85 µV/K, was found when deposited in small electroformed posts with a resistivity of 5 mΩ cm.
1 R. Venkatasubramanian, E.S., T. Colpitts, B. O'Quinn,, Nature, 2001. 413: p. 597.
PDF
