Improved wetting of gold active braze alloy on diamond for use in medical implants

https://doi.org/10.1016/j.diamond.2020.108089Get rights and content

Highlights

  • New method enables biocompatible braze joints in diamond by improving wetting of gold based braze.

  • Niobium/molybdenum bilayers are required to promote wetting of molten gold on diamond

  • Method of generating complex gold structures embedded in diamond as alternative, biocompatible electronic components

  • Brazing with gold enables formation of diamond hermetic capsules.

Abstract

Medical implants containing active electronics must have a leak-proof encapsulation to be certified as safe for human use. Implantable devices made from diamond demonstrated exceptionally long implantation lifetimes due to the outstanding biostability and biocompatibility of the material. However, since diamond does not melt and is therefore not weldable, forming joints between diamond components or embedding metallic wires and bonding pads within diamond is challenging. One method consists of using active braze alloys to bond diamond surfaces together. These active brazes comprise a precious metal alloy containing a carbide forming element that chemically bonds to the diamond as the braze metal melts. Silver-based active braze alloys are used successfully for brazing diamond in industrial applications, but silver is toxic to living tissue and, therefore unsuitable for use in implants. Gold active braze alloys (Au-ABA) are biocompatible but exhibit very poor wetting on the diamond. Here we demonstrate the use of molybdenum (Mo) and niobium (Nb) interlayers including single layers of Mo, Nb or Mo/Nb bilayer thin films as a solution to improve the wetting of Au-ABA on diamond surfaces. Theses interlayers provide for excellent penetration of the braze into the grooves and crevices in the diamond surfaces. We report on optimum recipes for the interlayer, both for the fabrication of weld lines and for the formation of smaller complex micro-structures and hermetic electrical feedthroughs.

Introduction

Electronic technology is now sufficiently advanced that we can easily measure in vivo bioelectric signals and deliver biologically relevant stimuli. The challenge is now to develop the interface materials and technology to miniaturize devices so that they are all but invisible to the human immune system and last in vivo for many decades. Along with the desire to miniaturize devices, improvements in digital technology create the opportunity to strive for increasing complexity, and devices able to perform a wider range of functions [1,2]. For example, cochlear implants [[3], [4], [5], [6], [7]] stimulate the auditory nerve to synthesize the sensation of hearing, and their recording capability enables the devices to self-calibrate to the needs of an individual user. Closed-loop implants [[8], [9], [10], [11], [12], [13], [14]], with the ability to record and self-regulate a therapy, are becoming more and more common. Examples include implantable blood glucose autonomous basal insulin load management [[15], [16], [17], [18], [19]], pain management using loosed-loop autonomous regulation of stimulation intensity of the spinal cord [20,21], and control systems for prosthetic limbs' using touch sensation feedback for the user [9,22,23].

Such closed-loop systems require sophisticated electronics, often in the form of an Application Specific Integrated Circuit (ASIC) implanted into the body. Larger devices can be well accommodated in some locations within the body, for instance, cardiac pacemakers in the chest wall [24,25]. However, eye or brain implants would benefit from smaller devices to minimize risk by reducing the invasiveness of the surgery. Modern electronics and vertical integration techniques enable increasingly complex electronics to fit within a very small footprint. The search for safe and biocompatible materials for encapsulation of these millimetre-scale devices, in particular when many electrical feedthroughs are required, is ongoing. For many applications, the aim is to fabricate devices with hundreds of electrical feedthroughs connected to stimulation or recording electrodes. For example, Fig. 1 shows a diamond interposer component from a retinal stimulation device, designed to be implanted into the eyeball. This device is intended to restore a degree of vision to people with blindness caused by degenerative retinal diseases [7]. All exposed materials used in such a component must be non-cytotoxic and highly resistant to degradation by the harsh environment inside the body. There are several properties of next generation implants that would be greatly enhanced by long lived materials. Device is expected to become increasing small and complex. With this, comes increased sensitivity to moisture ingress. In addition, there is a strong drive to reduce clinical risk by minimising repeat surgeries. Therefore, longevity of the implant by ensuring hermeticity and choosing inert materials is highly desirable. The popularity of cochlear implants[3,4,6], for instance, is at least partly due to the fact that they are designed to last 50 plus years which minimizes the risks associated with repeat surgeries which become necessary if the device fails.

Diamond based materials come in a wide variety of ‘flavours’ from electrically insulating to conducting and transparent to opaque and possess several unique optical qualities with potential for a range of application in implantable bionics [[26], [27], [28], [29], [30], [31], [32], [33]]. In terms of longevity, diamond is a material without peer regarding chemical and mechanical stability. Furthermore, a wide range of studies has confirmed the non-cytotoxicity of diamond in many of its forms including boron-doped electrically conducting diamond [26,27,[34], [35], [36]]. Whilst diamond is extremely robust and long-lasting, it is also difficult to machine into desired shapes and forming hermetic joints between diamond components is still challenging.

It has been previously shown that diamond can be laser milled into complex shapes, integrated with electronics in applications designed for neural stimulation and neural recording, and employed as a material for high-density electrical feedthrough arrays [7,28,30,31,33]. However, since diamond does not melt and is therefore not weldable, forming joints between diamond components or embedding metallic wires and bonding pads within diamonds requires the development of dedicated technologies. One method consists of using active braze alloys which comprise a precious metal alloy containing a carbide forming element that chemically bonds to the diamond as the braze metal melts. This paper describes how, by the use of wetting layers, we were able to overcome previous limitations of brazing technologies in forming hermetic joints.

To evaluate adhesion wettability studies of liquid metals on diamond are essential, as they allow the work of adhesion (Wadh) to be determined. Wadh is defined as the difference between the surface energy of two separated surfaces and the energy of the interface at equilibrium [37]. The dependence of Wadh on the contact angle (θ) and the liquid–vapor surface energy (σLV) can be obtained by the Young-Duprè equation [37]:Wadh=σLV·1+cosθ

It can be deduced from Eq. (1) that with decreasing the contact angle, i.e. improving the wetting, Wadh increases, i.e. the more energy is required to separate the two surfaces in the solid state. Nevertheless, the thermodynamic work of adhesion derived from wetting experiments is not sufficient to determine the work of adhesion; indeed, other factors contribute to the fracture energy of a deformable solid [37]. Table 1 compares the contact angles of gold (Au) on the different surfaces including diamond.

In previous work, methods to form a hermetic encapsulation using diamond are described [30]. In that work, a study showing that a gold-based active braze alloy (Au-ABA) was well tolerated in vivo, eliciting a similar (minimal) immune response to medical-grade silicone when implanted into the back muscle of a Guinea pig over a period of 12 weeks. However, that work also highlighted that molten Au-ABA exhibits extremely poor wetting on the diamond and hence is unusable as a brazing material when directly deposited.

Fig. 2 illustrates the problem to be solved in the present work. Fig. 2 shows laser milled polycrystalline diamond (PCD) samples with a series of 200 × 200 μm2 squares and 200 μm wide strips, brazed at 1100 °C for 30 min. Panel (a) shows the sample coated with Au-ABA paste before brazing, panel (b) the Au-ABA paste after brazing, and panel (c) after excess braze has been polished away. Clearly, the molten Au-ABA during brazing does not wet the diamond substrate, instead of beading and retracting to a small area. Following polishing, only a small subset of features was filled with Au-ABA braze. Although Lichter et al. [30] found that silver was successful as a metallic wetting layer for Au-ABA on diamond, its cytotoxicity limits its use in a permanent implant.

In the present work, we use molybdenum (Mo) and niobium (Nb), two metals with good biocompatibility profiles [[40], [41], [42], [43], [44], [45], [46], [47]], as materials for improving the wetting of Au-ABA on diamond. The choice of Mo and Nb was motivated by reports that Mo has good resistance to corrosion by mineral acids, its toxicity is considered small and there is an increasing trend in using alloys containing Mo for manufacture of surgical implants [47]. Nb has been reported that has a relatively low electrical resistivity, good corrosion resistance [48] and significantly good biocompatibility for using as biomaterials and in vivo studies [43]. Furthermore, not only are Mo and Nb both carbide forming elements, but they also exhibit good wetting with gold [40,41]. Au-ABA contains titanium (Ti), which forms carbides easily; but Au does not wet TiC well, as it can be seen from Table 1. Frage et al [38], however, describe that the contact angle of Au with 3.7% Ni content on TiC is dramatically decreased compare to pure gold (not shown in Table 1). The Au-ABA used in the work described in this article has 3% Ni content. This finding supports our proposal that TiC layers between Au-ABA and diamond do not form quickly enough to promote effective wetting. An interlayer is required to maintain wetting long enough for Ti to diffuse to the surface from the braze bulk and for TiC to form.

The challenge is to find the optimum point in the competition between carbide formers of Mo or/and Nb with Ti as a transient bond to the diamond surface which can lead to the strongest wetting and bonding. Here we describe optimized parameters for generating small (<200 μm) inlaid features in diamond for a generation of electrical interconnects, pads, and wiring. We describe a different set of parameters for larger features (>200 μm) employed as weldable inserts in diamond components or metallic electrical feedthroughs. The methods described are shown to be very suitable for the fabrication of miniature, hermetic, bio-permanent, bio-compatible components for complex neuromodulation devices.

Section snippets

Materials and fabrication process

TM100 Polycrystalline CVD diamond (PCD) plates sourced from Element Six Ltd. (either 0.3 or 0.5 mm in thickness, single side polished (<50 nm RMS)), were used as the base material. Gold Active Braze Alloy (Au ABA), in paste form, was purchased from Morgan Advanced Materials [Hayward, CA, USA], with a nominal manufacturer specified composition of Au: 96.4%, Ni: 3%, and Ti: 0.6%. Diamond samples were patterned using a 2.5 W Nd:YAG, 532 nm wavelength, nanosecond pulsed laser micromachining system

Characterization of braze adhesion with Mo and Nb

Initial results demonstrated that optimal wetting layer thickness and composition were different depending on the size of the features milled into the diamond, therefore we described the results based on the feature size of samples. The composition of optimized wetting layers for different feature sizes is summarised in Table 2.

Depositing single layers of Nb and Mo on the separate LP samples, partially improved the wetting of Au-ABA on the PCD plates. The wetting was also increased by

Discussion

There were observed cracks on the back side of the single Nb-deposited samples as well as bilayer samples. The cracks probably originated from mechanical stresses due to i) a good adhesion, ii) a thermal expansion mismatch between dissimilar compounds, such as Nb, Au-ABA, diamond and Ti, iii) heating/cooling rate, iv) excess load of braze, or v) laser micromachining of a symbol on the back side of sample, weakening the diamond. The amount of braze is controlled by hand and we did not have any

Conclusion

Forming joints between diamond components or embedding metallic wires and bonding pads within diamond is challenging due to the lack of surface wetting between carbon and most liquid metals. We introduced metallic interlayers which can improve and maintain wetting of Au-ABA on the diamond surfaces. We found that the best recipe of the wetting layer depends on the size of milled features. For small features, the best recipe is Mo/Nb bilayer (Mo deposited first) and for larger features, the best

CRediT authorship contribution statement

Experimental Design: Khatereh Edalati, Melanie Stamp, Kumaravelu Ganesan, Alastair Stacey, David Garrett, Supervision and Funding: David Garrett, Steven Prawer. Diamond/Gold Component Integration: Gabriel Martin-Hardy Manuscript Preparation: Khatereh Edalati, Réjean Fontaine Melanie Stamp, David Garrett, Steven Prawer.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

SP is a shareholder in iBIONICS, a company developing a diamond based retinal prosthesis. SP and DJG are shareholders and public officers of Carbon Cybernetics Pty Ltd., a company developing diamond and carbon-based medical device components.

Acknowledgments

The work was performed in part at the Melbourne Centre for Nanofabrication (MCN) in the Victorian Node of the Australian National Fabrication Facility (ANFF). DJG is supported by NHMRC Project Grant GNT1101717 and by an Australian National Fabrication Facility (ANFF)/Melbourne Centre for Nanofabrication (MCN) Technology Ambassador Fellowship. SP is a shareholder in iBIONICS, a company developing a diamond based retinal prosthesis. SP and DJG are shareholders and public officers of Carbon

References (57)

  • L. Karumbaiah et al.

    Relationship between intracortical electrode design and chronic recording function, Biomaterials

    (2013)
  • F.G. Zeng et al.

    Cochlear implants: system design, integration, and evaluation

    IEEE Rev. Biomed. Eng.

    (2008)
  • T. Stöver et al.

    Biomaterials in cochlear implants, Laryngorhinootologie

    (2009)
  • A. Kral et al.

    Developmental neuroplasticity after cochlear implantation, Trends Neurosci

    (2012)
  • O. Macherey et al.

    Cochlear implants

    Curr. Biol.

    (2014)
  • K. Fox et al.

    The bionic eye: a review of multielectrode arrays, Handb

    Bioelectron. Directly Interfacing Electron. Biol. Syst.

    (2015)
  • K.M. Silay et al.

    A closed-loop remote powering link for wireless cortical implants

    IEEE Sens. J.

    (2013)
  • R.S. Armiger, F. V. Tenore, K.D. Katyal, M.S. Johannes, A. Makhlin, M.L. Natter, J.E. Colgate, S.J. Bensmaia, R.J....
  • J. Wright et al.

    A review of control strategies in closed-loop neuroprosthetic systems

    (2016)
  • F. Kohler et al.

    Closed-loop interaction with the cerebral cortex: a review of wireless implant technology§, Brain-Computer Interfaces

    (2017)
  • S. Wendelken et al.

    Restoration of motor control and proprioceptive and cutaneous sensation in humans with prior upper-limb amputation via multiple Utah Slanted Electrode Arrays (USEAs) implanted in residual peripheral arm nerves

    (2017)
  • G. Shin, A.M. Gomez, R. Al-Hasani, Y.R. Jeong, J. Kim, Z. Xie, A. Banks, S.M. Lee, S.Y. Han, C.J. Yoo, J.L. Lee, S.H....
  • S. Nicolaidis

    Neurosurgery of the future: deep brain stimulations and manipulations, Metabolism

    (2017)
  • M.M. Ahmadi et al.

    A wireless-implantable microsystem for continuous blood glucose monitoring

    IEEE Trans. Biomed. Circuits Syst.

    (2009)
  • M.J. Taylor et al.

    In vivo study of a polymeric glucose-sensitive insulin delivery system using a rat model

    J. Pharm. Sci.

    (2010)
  • O. Veiseh et al.

    Managing diabetes with nanomedicine: challenges and opportunities

    Nat. Rev. Drug Discov.

    (2014)
  • C. Chen et al.

    Current and emerging technology for continuous glucose monitoring, Sensors (Switzerland)

    (2017)
  • H. Lee et al.

    Enzyme-based glucose sensor: from invasive to wearable device

    (2018)
  • Cited by (7)

    View all citing articles on Scopus
    View full text