AuSn

Two categories of solder are available to choose from when the in-service environment for a device reaches above 125°C either in continuous operation or thermal cycling accelerated life testing. These categories are those comprised primarily of lead, and those of gold. From the electronics industry’s drive to eliminate lead, many manufacturers who have traditionally used lead solders are treading cautiously, looking now at the gold solders, primarily at Indalloy 182 (80Au20Sn).

The most common concern regarding this switch relates to the strength of AuSn, which is much higher than the lead solders. The degree that this should be of concern however, should be realized within the scope of the application.

For instance, review this case scenario:

Indalloy 159 (90Pb10Sn) was used in a device for years to adhere high temperature sensors to a calibration probe that is slowly cycled in operation from 350K (~75°C) to 500K (~225°C). The solder joins a nickel and gold plated Kovar™, or platinum or platinum coated, nickel lead to a tinned copper lead. The solder joint is not placed under tension or shocked.

Considering the high temperature solder options in this scenario, the AuSn would be mechanically preferred.

Why?

Well, tin-bearing soft solders will leach gold from gold metallizations during soldering, creating a brittle Au-Sn intermetallic layer within the solder joint. The more gold available, the more consumed, and the greater the thickness of the resultant intermetallic layer. The brittle nature of this layer, situated intimately next to the relatively soft PbSn solder layer, creates differential stresses that promote crack propagation upon thermal cycling.

AuSn was not considered previously because the engineers were familiar with its hardness and, given the cracking failure described using a softer solder, they did not anticipate improvement. It was a pleasant surprise to them to find that switching to a lead-free solder would not sacrifice the quality of their device. AuSn is a brittle alloy but, unlike the description above, no differential stresses are involved. 

Note: Eutectic gold solders have been used for many years to solder nickel plated Kovar™ lids to high reliability ceramic packages and have a good history of fatigue performance.

Recent test results show that solder fluxes handle high reflow temperatures (>450°C), providing better-than-expected visual results of flux residue!  

Solder fluxes have not traditionally been used with AuSn, AuGe, or AuSi eutectic solder, because their peak reflow temperatures were very close to or above the flux activation range. An average flux activates at approximately 125°C and is not recommended for temperatures in excess of 350°C. Although AuSn solder melts at 280°C, peak reflow temperatures are recommended to be >300°C, nearing the maximum suggested temperature of flux. AuGe and AuSi alloys melt at 356°C and 363°C respectively, which exceed the documented flux activation range.

Since these alloys contain ≥80% gold and are resistant to oxidation, flux is not always necessary. Other methods have conventionally replaced the flux function, such as mechanical scrubbing, or forming gas purging. If these technologies are not available, or assembly speed is priority though, a flux may be required.

So I, along with my fellow engineer, Brandon Judd, sought out to test some of our best fluxes with Au alloys at these high temperatures. The result: Not all, but a few of these fluxes work extremely well up to temperatures as high as 450°C!!!

The reflow profiles used tested the extreme abilities of our fluxes:

·         Peak temperature 410°C

·         Nitrogen Purge

·         Time above liquidus: 137 seconds

·         80AuSn solder preforms 0.249” square x 0.002”

Some fluxes did what we expected- they charred and burned. They simply were not designed for this environment. 

Others, such as our TacFlux010® were very resilient at these temperatures. 

For more information about these test results, please contact myself or Brandon.

Happy Testing!!

Amanda

• 96.5%Sn 3.5%Ag (221C)
• 80%Au 20%Sn (280C).

在这里给大家拜年了：祝愿大家身体健康，万事如意，财源滚滚，阖家幸福！

有人告诉我2011年是金兔年(Golden Rabbit Year)。我上网查了一下，还是得不到考证。但是却让我想到了，Indium公司和金gold (Au) 这种金属，还是有一定相关性的。

金gold (Au)自身的熔点是⁰C，但是如果和锡tin（Sn）在一起，做成80%Au20%Sn的共晶合金，熔点就只有280 ⁰C了。金锡共晶金属有很高的焊点强度(joint strength)，抗腐蚀能力，导热性能好(thermal conductivity)，能够与各种贵金属兼容，还符合无铅的要求；可靠性很高。

金锡共晶金属在SMT里或是电子焊接中的用途很广泛。在SMT中，如果需要分温度层焊接(step soldering)，金锡的280度正好作为第一梯度的焊接；第二梯度的焊接合金可以选择锡银铜SAC或是锡铅SnPb；如果有第三梯度，可以选择锡铋SnBi。 在IGBT，automotive， 和 Radio Frequency (Power Amplifier) 的第一层焊接中，金锡常常是首选。 在电子焊接中，金锡的用途就更广泛了，特别是在医疗器械、仪器中。比如说在catheter导管的应用中，就可以用金锡做成微细的像小弹簧形状的物体，放入心导管中，帮助心肌梗塞的病人…随着全球人群的老龄化，在医疗器械、仪器方面的金锡应用，应该会越来越广泛。(当然，我可不希望有一天自己要靠它来救命哦。)  

当然，金的价格不菲，特别是这几年的疯长。所以选用这种合金，出于经济的考虑因素，也是要很谨慎的。 如果几年前我也买了金存着，现在也是一笔增值可观的小财富了：-）

Cheers!

PS: 我们家的张“小兔”宝宝五月底就要出生了，他现在已经在我肚子了老是练功夫了，踢来滚去的，肯定是我怀孕以来突然多看了功夫片和武侠小说的缘故。如果今年真的是金兔年(Golden Rabbit Year)，只希望小兔以后能够健康和财运亨通（lots of gold!），并且有一颗金子般的心(a golden heart).

Pic:
1. Baidu image
2. www.indium.com Indium Corporation

剛剛看見一個關於如何使用AuSn預成型焊片 （金錫，主要是Au80Sn20的共晶材料）來做laser bar bonding的短文和錄像，頗受啓發。在此與大家共享。短文和錄像轉載在Finetech:  http://www.finetech.de/micro-assembly/applications/laser-bar-bonding.html

Laser Bar Bonding
In comparison to single lasers, laser bars consist of several single edge-emitting lasers comprised in a silicon bar. Bundling the light results in significantly increased output power as needed in a wide range of applications.

Laser bars are high power products, applied wherever small and efficient light-emitting units are needed. They mainly serve as pumping sources for optical resonators of high power lasers. Another field of use can be found in medical appliances.

What are the challenges?

• Sensitive components with brittle materials and optically active areas
• Delicate edges and facettes
• Perfect coplanarity required to ensure proper thermal management
• Highest placement accuracy needed to ensure defined overhang
• Stacked mounting (lasers to submount) requires different solder materials with their appropriate profiles
• Void free bonding
• Solder materials are prone to oxidation, process integration required
• High magnification of large components
• Indium bonding under process gas atmosphere

First image & English content source: FineTech; AuSn image from Indium Corporation.

PS: 最近在做某大客戶的生意，其中有一個性格爽朗的活躍女工程師K引起了我的注意。她看樣子就像剛PhD畢業參加工作的女孩子。後來客戶中了解這位工程師的好友L告訴我，K其實已經30多嵗了；18嵗時在某囯嫁給了36嵗的男人，漂亡來美國這個異國它鄉，先後生下兩個孩子。因爲前夫對她不好，K忍無可忍終于離婚了，但是前夫一直以來不讓她有探望孩子的權利，更別説照料了。K在美國，先後讀完了本科，碩士和博士；現在在好公司有份好工作，並且每天都努力地工作著。現在K也有了一個相處4年多的穩定男朋友了……聼完K的故事，讓我肅然起敬。從K開懷的笑聲中，外人全然看不出來她有這種痛苦的過去! 一個不對困境地頭，對生活充滿了熱情，對未來充滿期望並為之奮鬥的可貴靈魂！

This blog has been in existence for a little over two years now, and we would like to thank our readers for the feedback and inquiries you have provided. I welcome your comments on what you would like from us. Leave a comment below, or email me at jhisert@indium.com.

And now a look back on past topics of interest:
 

Grid Ink, Silver Ink, Conductive Ink

Bismuth/Tin Tabbing Ribbon, A Low Temperature Pb-Free Alternative

Plated Metallization for C-Si Solar Cells

Increase Packing Density for Evaporation Crucibles

Photon’s 5th PV Tech Show 2010 USA

IPC Solar Standards Update

Solder Shelf Life as Explained by Eric Bastow

Tips to Speed Your Solder and Flux Selection

What's Happening in the Technical Service Department 

A Day in the Life of a Tech Guy

A Clean Laboratory

CIGS for Beginners

3rd Renewable Energy Expo 2009 in New Delhi, India

Solar Products and Representatives

Kleenex®, Google™, FedX®, CIGs?

Indium Solar Products Reunited

Trade Show Visitors Love the Ground Floor

Solar Product Data Sheets

Intersolar 2009 – What Barrier to CIGS Technology?

Concentrator Photovoltaic Systems - Will they reach 50% Efficiency?

Standards for Solar Panel Manufacturing

Solar Panel Certification: “Barrier and Benefit” Reviewed by Eric Bastow

Low Temperature Metallization Paste

What Will Your Interest Be At InterSolar? Meet the Bloggers And Let Us Know.

Share Your Solar Images

SAC vs. Sn/Ag for Solar Soldering

Solder Thickness for PV Interconnect

What is Bus Ribbon?

Standard PV Interconnect Ribbon Sizes

No-Clean Flux

Photovoltaics in EMS Sector

PV Interconnect Products

Eric Bastow - East Coast Technical Support

Mario Scalzo - West Coast Technical Support

Au/Sn Sputtering Targets

SMT Goes Solar

A Trip Down Memory Lane 

More Information About Metallization Paste

A year in the Life of a Solar Blog

CIG Target

23rd European Photovoltaic Solar Energy Conference and Exhibition

TCO choices for CIGS manufacturing 

CIGS Absorber Layer Electroplating

No Slump Metallization Paste

Meet the Bloggers

CIGS - Can sputtering make a breakthrough?

Fluxes for Soldering Tabbing Ribbon

Computer Brain vs. Solar Photovoltaic

Beam it down from space

Selection of the Optimum Lead-Free Solder for Solar Tabbing Ribbon

Record Makes Thin-Film Solar Cell Competitive with Silicon Efficiency

Why Thin-Film Solar Cells are Here to Stay

Hot Rooftops to Flashy Digital Cameras

Synchronize Your Solar Cell

Solar Conversion Efficiencies  

Government Support is the Key

It's Just a Beginning ...

I have been asked on numerous occasions to calculate the potential for galvanic corrosion between metals. Most times, when I am approached with this, the concern stems from an application in which the bonding metals will be mated in a corrosive environment, such as a salt solution.

When the potential is to be calculated for two elemental metals bonding, the potential for galvanic corrosion is simple to calculate. Simply look up the anodic potential difference between the two metals under the galvanic series in a general chemistry handbook and if the value is less than 0.15V (the maximum recommended for a salt solution), galvanic corrosion should not be a concern. For normal environments, such as storage in warehouses or non-temperature and humidity controlled environments there should not be more than 0.25 V difference in the Anodic Index. For controlled environments, such that are temperature and humidity controlled, 0.50 V can be tolerated. 

This value is much more difficult to calculate, however if the bonding metals are alloys rather than elemental metals. 

For instance, I cannot easily supply the anodic potential difference between 80Au20Sn and a pure Au plating to prove that it is less that 0.15V. This is because I cannot calculate the anodic potential theoretically for the AuSn alloy. Data is readily available for pure metals, but the potential for individual solder alloys must be determined experimentally because the voltage potential is not linear and as you begin to add a second metal to a pure metal, the rate of voltage change is different between different alloys. 

For this exact situation, I can speak practically however. We have tested gold plated Kovar lids for corrosion that were sealed to semiconductor packages that had a gold seal ring using a preform of AuSn. They were tested for corrosion in a salt spray chamber per MIL STD 883. Corrosion, when it occurred, always was on the lid where the porous gold allowed underlying nickel corrosion. There was never an instance of corrosion at the Au/Sn and Au interface region.

• 焊接点牢固 
• 十分抗腐蚀
• 特别高的热传导性能
• 与别的贵金属相兼容
• 出色的抗热疲劳特性
• 优异的润湿性
• 高熔点（2800 C）

Eutectic Gold Tin (AuSn) with a composition of 80Au20Sn is a unique material.  This particular alloy of gold tin (AuSn) is considered a solder because it has a melting temperature of 280ºC, which is lower than the 350ºC transition temperature into braze materials.  Still, there are some similarities between this solder alloy and braze alloys. The most obvious is the hardness of the gold tin (AuSn) alloy. With a tensile strength of 40,000PSI, this solder is much more rigid than the tin solders most are familiar with. The strength is more closely compared to the silver brazes which melt above 500ºC. 

With that strength has come some unique manufacturing difficulties. For many years, one obstacle for implementation of gold tin (AuSn) as a solder preform or wire, was its availability in thin forms or fine diameters. The gold tin (AuSn) is extremely hard and it became brittle as it was handled through manufacturing and would crack if it was pressed too thin or fine.  

Luckily, in the 40+ years since eutectic gold tin (AuSn) was first used in electronics manufacturing, processing techniques have come a long way.  Today, gold tin (AuSn) solder can be made into dimensions much smaller than the soft solders, allowing it to be used in applications which require the highest level of precision.

Typical dimensions and tolerances of gold tin (AuSn) can be found in the below chart.

This chart as well as more detail on gold tin (AuSn) applications are available in the paper titled, “Process and Reliability Advantages of AuSn Eutectic Die-Attach,” presented at IMAPS 2009.    

Greenpeace has just updated their "Guide to Greener Electronics."  There are a couple of interesting tibdits that I took from their report:

1. They are really focusing on the phase out of BFR's and PVC from electronics.  They dropped HP and Dell down because they are loosening their timeline of BFR and PVC phase-out.  I find it interesting that they make no note of what the replacements should be.  This is concerning that the replacements could potentially be MORE toxic than what they are replacing.  It took years to fully characterize the situations where BFR's and PVC are of concern (dioxin formation and bioaccumulation).  In addition, all BFR's are not the same.  If companies are phasing out these materials, how can they do a full risk assessment of the replacements in one to two years?
2. They have added Antimony (Sb) to their list of materials that need to be phased out.  This can be challenging for a number of soldering applications.  Component manufacturers have been using Sn/Sb alloys inside their components.  Sn/Sb is the highest melting point Pb-Free alloy that actually solders reasonably well (other than Au/Sn which is 1000x as expensive).  The component guys are using this so that those alloys are not remelting when that component is assembled in a SAC SMT process.  Eliminating Sb will create a number of assembly challenges as well as potentially significant reliability issues.
3. By reading the summary of the report, they praise Apple for phasing out virtually all BFR's and PVC.  However, their ranking is still in the bottom half.  I will write more about this one in a future entry.

I am all for designing electronics for the environment, but I think there needs to be more focus on the consequences of making those design changes.  Are the alternatives actually any better?  What is the impact on product reliability and functionality?

A passage from Metals Handbook, Volume 6, 1983Ӕ

Fact: To achieve a strong, reliable solder joint, it is important to remove all oxides from the substrate as well as the solder.  Generally this is done using a compatible flux.

Reality: Sometimes flux cannot be used, so it is important that the starting materials be as oxide-free as possible.  Au/Sn is a popular solder in flux-less applications. The preform is a popular form for Au/Sn because it delivers a consistent, controlled amount of solder every time.  However, unless properly controlled, the manufacturing process can introduce oxides that become imbedded in the material and cannot be easily cleaned.

Solution: Indium Corporation has perfected its process used to manufacture Au/Sn preforms.  This is no small task.  Our experienced Metallurgists and Process Engineers have taken a good thing and made it even better.  If you're looking for proof, talk with one of our Application Engineers about evaluating Indium Corp. Au/Sn preforms in your process!

Break down of a radio frequency

终于，终于发3G牌照了。基站设备厂商们拿到电信，移动，和新联通的订单后，也已经在紧锣密鼓地建设新基站或是给旧的基站更新换代了。

在3G基站的建设中，会用到Radio Frequency (RF)射频设备。一般来说，每一个RF中有两个功率放大器Power Amplifier(PA)。无论是PA的制造，还是RF的组装，Indium公司都提供了系列可靠性能强的精确焊接材料。比如说PA的制造，我们有AuSn预成型焊片(preform)或是焊带(ribbon)；RF的组装，我们可以提供有铅无铅的焊锡膏(solder paste)，含驻焊剂的焊片(flux coated preform)，还有散热界面材料Thermal Interface Material (TIM).

好友王说，移动在全国大概要建40万台3G基站；从家人朋友那里也听闻到电信，新联通的一些数据，粗略估计，分别建(或是硬件更新)30万台基站吧（如果我的估算偏差大了，欢迎随时联系azhang@indium.com ）。那么，全国就要建100万台3G基站了。平均每一个基站有6个RF，每一个RF有两个PA，每一个PA的制造/封装，可以用到4个含驻焊剂的焊片，每一个RF可以用到一个散热界面材料(TIM)，我们把这些乘的乘，加的加，也大概可以估算出中国的3G基站建设用到多少焊接材料了。

Cheers!

PS: 1. 好久没有回中国了，这次回来亲身感受到各方面的"冲击"，其中热门的3G应该是最大的冲击之一吧。 2. 好久没有正面接触曾经的老本行通信了；这些年来学习和工作的内容，与通信直接相联系的较少。这次当被一个同事问起3G的一些知识时，我哑口无言了…当时倍感对不起北邮，对不起亲戚朋友们。哈哈，看来我这"烂船"也要修修补补了，不然连"三斤铁"都没有了。

Pic: From Jordan Ross with Indium Corp.

In addition to Au/Sn sputtering pucks, Indium Corporation can also manufacture full size Au/Sn sputtering targets.  80Au/20Sn is a versatile solder that is used in a wide variety of specialty applications within (and outside of) the photovoltaic world.

Au/Sn is commonly used as a Pb-free solder to attach Au surfaces.  The 80/20 alloy melts at 280°C, and can be used with a mild flux or a reducing atmosphere.  Another key point of 80Au/20Sn solder is that it has a tensile strength of 40,000psi.  This material's high melting point, combined with the non-hazardous nature of Au and Sn have made it very popular for Pb-free assemblies that must endure subsequent Pb-free reflow temperatures (~245°C).

Once you have an understanding of the type of solder needed for an application, (a hard solder such as AuSn for reflow temperatures above 125°C or a soft solder such as SnPb or SnAgCu for lower temperatures) it is time to consider the impact of the substrate metallization on your solder choice. 

Common surface finishes for soldering include gold (ENIG), copper and OSP, immersion silver, tin, and nickel. Each element reacts differently with these finishes and solders should be carefully chosen to match the finish to prevent issues such as brittle intermetallics and excessive scavenging.

The data sheet titled flux and solder compatibility found here recommends solder choices for various substrate finishes as well as incompatible solders.

For more assistance in choosing an appropriate solder for your particular application, please feel free to contact me directly.

A set of wires prepared for connection

Properly joined thermocouple wires

Thermocouples do a lot for us.  We use them for profiling reflow ovens, checking material temperatures, and a host of other temperature related measurements.  They see a lot of abuse and they are bound to break with enough rough use.  So how do we fix broken thermocouple wires?  I have a method that works very well.  It may not be the cheapest way to fix the wires but it has the following advantages:

• Pb-free joint
• Highest thermal conductivity of any current method
• Good for use up to 280°C
• Strongest connection (40,000psi tensile strength)
• Requires no specialized equipment

This method can also be used to convert leaded thermocouple wires to pb-free.

Here’s how:

1) Clip the ends of the thermocouple wires so they are even

2) Strip the sheathing back as shown in the picture (1/4inch)

3) Use a razor blade or emery paper to scrape the oxide layer off the wires, then twist the ends together

4) Put a very thin (~.001”) layer of NC 506 flux on the surface of a ceramic coupon and the exposed thermocouple wires

5) Place an 80Au/20Sn preform or a sphere(s) of the correct volume on the flux layer

6) Place the coupon onto a hotplate set to 400°C

7) Bring the wires over to the Au/Sn (which should now be molten)

8) Dip the wires into the solder

9) The solder should wick onto the wires, when it does – remove the wires.

You can leave the no-clean flux residue on the wires, or wipe it off using a solvent and rag.  You now have a high-temp pb-free thermocouple.

If you'd like to discuss this with me, click here or just give me a call @ (315) 853-4900 x-7592. 

http://upload.wikimedia.org/wikipedia/commons/b/ba/Thermometer.jpg

Operational temperature

Metallizations
Flux usage
Reflow method
Temperature limitations
Special conditions

Whether the market is communications, automotive, energy, military, medical or aerospace, choosing the correct solder is essential to making a quality product. A solder joint plays a very important role in an electrical or mechanical assembly. How this solder is selected for this joint is extremely important to the application, so following a guideline will insure that this is accomplished.

Before choosing a solder for an application, the operational or service temperature must be known; i.e. the minimum and maximum temperatures the assembly will see, during testing and in service. Continuous operational temperatures above 125°C will require the use of precious metal solders or hard solders such as Au/Sn, Au/Ge, or Au/Si.

Generally, it is best with soft solders to choose one that will melt at least 50°C higher than the expected service temperature of the assembly. (This includes temperature cycling during test and burn-in). If the difference in temperature (Delta T) between operational and the solidus temperature of the solder is too small, this could result in excessive scavenging, leaching and intermetallic growth. These unwanted conditions could lead to problems and eventual joint failure. 

After the correct temperature range for the solder is determined, consideration needs to be given to what metallizations in the joint will interact with the solder.

Metallizations will be addressed next in our series of steps to consider when designing a trouble-free solder joint.