14

I am about to try my first "reflow skillet" soldering job, and as I look at the available types of solder paste I see there are lead-free pastes with much lower melting temperatures than others.

For example, this one from ChipQuik.

The advantages seem obvious, but somehow the marketing literature does not mention any drawbacks to this type of solder paste. In the quantities I would order the price seems about the same. Is there a reason this Sn42Bi58 formula hasn't become standard?

Kaelin Colclasure
  • 2,669
  • 7
  • 27
  • 37
  • 1
    Too expensive for production perhaps? – Olin Lathrop Jan 07 '12 at 14:04
  • @OlinLathrop FYI 7 years on (due to a reader's recent 'questions'). The alloy is sold mainly as a desoldering solution because, when used to resolder an existing lead based solder joint, the resultant alloy can then be mechanically picked apart easily with a sharp tool. The solder CAN be used in very controlled situations by competent assemblers when lead poisoning can be guaranteed not to occur - IBM used it at one stage. – Russell McMahon Jan 07 '19 at 04:24
  • @OlinLathrop Many examples of sale for desoldering use [**here**](https://www.google.com/search?num=40&ei=ytIyXLCNKcTHwASBnJvQAw&q=bismuth+chip+quik+solder+removal+-paint&oq=bismuth+chip+quik+solder+removal+-paint&gs_l=psy-ab.3...64624.66749..67767...0.0..0.259.1598.2-7......0....1..gws-wiz.......0i71j33i160.Owz-ui_tERI) – Russell McMahon Jan 07 '19 at 04:26
  • @RussellMcMahon the OPs link is a dead link, but they mention "solder paste", not "desoldering". Your link pulls up a desoldering product, not solder paste. "Chip Quik� Surface Mount Desoldering Kit" – johny why Jan 08 '19 at 19:06

2 Answers2

20

42/58 Tin / Bismuth is not unknown as a low temperature solder but has issues.

While widely used for some very serious applications (see below) it is not a mainstream industry contender for general use. It is not obvious why not given its substantial use by eg IBM.

Identical to the Bi58Sn42 solder you cite is:

  • Indalloy 281, Indalloy 138, Cerrothru.

    Reasonable shear strength and fatigue properties.

    Combination with lead-tin solder may dramatically lower melting point and lead to joint failure.

    Low-temperature eutectic solder with high strength.

    Particularly strong, very brittle.

    Used extensively in through-hole technology assemblies in IBM mainframe computers where low soldering temperature was required.

    Can be used as a coating of copper particles to facilitate their bonding under pressure/heat and creating a conductive metallurgical joint.

    Sensitive to shear rate.

    Good for electronics. Used in thermoelectric applications.

    Good thermal fatigue performance.

    Established history of use.

    Expands slightly on casting, then undergoes very low further shrinkage or expansion, unlike many other low-temperature alloys which continue changing dimensions for some hours after solidification.

Above attributes from the fabulous Wikipedia - link below.

According to other references it has low thermal conductivity, low electrical conductivity, thermal embrittlement issues and potential for mechanical embrittlement.

SO - it MAY work for you, but I'd be very very very cautious about relying on it without very substantial testing in a wide range of applications.

It is well enough known, has obvious low temperature advantages, has been widely used in some niche applications (eg IBM mainframes) and yet has not been welcomed with open arms by industry in general, suggesting that it's disadvantages outweigh advantages except perhaps in areas where the low temperature aspect is overwhelmingly valuable.

Note that the chart below suggests that flux cored versions seem to be specifically unavailable either as wire or as preforms.

Comparison chart:

enter image description here

The above chart is from this superb report which however does not provide detailed comment on the above issues.

Wikipedia notes

  • Bismuth significantly lowers the melting point and improves wettability. In presence of sufficient lead and tin, bismuth forms crystals of Sn16Pb32Bi52 with melting point of only 95 °C, which diffuses along the grain boundaries and may cause a joint failure at relatively low temperatures. A high-power part pre-tinned with an alloy of lead can therefore desolder under load when soldered with a bismuth-containing solder. Such joints are also prone to cracking. Alloys with more than 47% Bi expand upon cooling, which may be used to offset thermal expansion mismatch stresses. Retards growth of tin whiskers. Relatively expensive, limited availability.

Motorola's patented Indalloy 282 is Bi57Sn42Ag1 . Wikipedia says

  • Indalloy 282. Addition of silver improves mechanical strength. Established history of use. Good thermal fatigue performance. Patented by Motorola.

Useful lead free solder report - 1995 - nothing to add on above subject.

Russell McMahon
  • 147,325
  • 18
  • 210
  • 386
  • This answer is hard for me to read, cuz the OP asked about lead-free, and you keep mentioning leaded alloys. Created confusion for me. – johny why Dec 27 '18 at 21:08
  • 1
    @johnywhy Maybe your comment was placed on the wrong answer. It does not match my answer in any way. My ONLY references to lead solder are 1. A warning that the solder asked about can cause joint failures when used on joints that have already been soldered with lead-based solders. An important warning. 2. The same warning in more detail in a note. Note that this relates to parts that have been PRE-TINNED with lead solder. If you use such with this solder the joint will probably fail, ... – Russell McMahon Jan 06 '19 at 18:16
  • 1
    @johnywhy ... viz "A high-power part pre-tinned with an alloy of lead can therefore desolder under load when soldered with a bismuth-containing solder." – Russell McMahon Jan 06 '19 at 18:17
  • your warnings are only "important" if the original poster asked about that scenario. Which they didn't and no reason to assume they plan on doing that. Otherwise just creates confusion. – johny why Jan 07 '19 at 01:23
  • @johnywhy It seems unlikely that we'll be able to bridge our communication gap :-). - The OP says: "... the marketing literature does not mention any drawbacks ... Is there a reason this Sn42Bi58 formula hasn't become standard?..." -> My answer directly addresses this core question with an extremely important answer. This may be summarised as "In tightly controlled environments under expert management the product may be useful (eg IBM use). However, if lead "poisoning" is possible, as is commonly and semi randomly the real world case, then the joints are **likely** to fail mechanically". – Russell McMahon Jan 07 '19 at 04:10
  • @johnywhy Starting 6 years ago, 12 users thought this answer good and the OP thought it the best one. I think I can observe factually that your confusion appears to be based on a lack of understanding of the question, the answer and the technology. I did not frequently refer to leaded solders, my reference to lead relates directly to what is the most probably answer to the OP's question, and the question and answer correlate well. I suggest that you having a close look at what was asked and what was said would be useful to you. | – Russell McMahon Jan 07 '19 at 04:21
  • @johnywhy IMPORTANT: Note that the cited chipquik "solder" is designed and sold essentially as a solder REMOVAL system based on the mechanical frangibility of the resultant when applied to an existing leaded solder joint. https://www.google.com/search?num=40&ei=ytIyXLCNKcTHwASBnJvQAw&q=bismuth+chip+quik+solder+removal+-paint&oq=bismuth+chip+quik+solder+removal+-paint&gs_l=psy-ab.3...64624.66749..67767...0.0..0.259.1598.2-7......0....1..gws-wiz.......0i71j33i160.Owz-ui_tERI – Russell McMahon Jan 07 '19 at 04:21
  • the OPs link is a dead link, but they mention "solder paste", not "desoldering". Your link pulls up a desoldering product, not solder paste. "Chip Quik� Surface Mount Desoldering Kit" – johny why Jan 08 '19 at 19:08
  • How is Indalloy 281, Indalloy 138, Cerrothru Identical to the Bi58Sn42 ? – johny why Jan 08 '19 at 19:10
  • @johnywhy I admire your persistence, I think :-). Garglabet will tell you the answers to "How is ..." if the links above are not enough. In the above table the 58/42 Bi/Sn alloy melts at 138C, and [This Indalloy FAQ page](https://www.indium.com/solders/bismuth/) says - "Indalloy® #281 (58Bi 42Sn) which melts at 138°C" - Garglabet knew that. [138C ~= 281F). Indalloy 138 seems to have vanished BUT allows which melt at 138C having 58Bi/42Sn concentrations are covcered in detail [in the "solder alloys" table](https://wiki2.org/en/Solder#cite_ref-emph_20-10) in this wikipedia page. – Russell McMahon Jan 11 '19 at 06:17
  • @johnywhy [CS alloys Tru 281](http://www.csalloys.com/products-cerrotu-alloy.html) - no prizes for guessing compostion of leting point, was, they say, formerly known as Cerrotru. [Presumably "previously" includes 7 years ago when I gave the original answer]. ||| They say: ... CS Alloys Tru 281 ...... melting temperature is 281 °F (or 138.0 °C), ... Table: Tru 281 Bi- 58.00 Sn - 42.00 – Russell McMahon Jan 11 '19 at 06:32
  • @Johnywhy I'd send this privately if I had contact details. || I'm happy to answer sensible questions and even to a degree, inexcusably uninformed questions. But, repeated comments which seem more like trolling | lack of desire to understand | & reflect zero effort to check why facts stated may be facts, then I may be roused from torpor to action. Maybe some combo of comments to admins| raising a meta question | sulking in a corner. | You have managed all of the above. I've copied the above thread for reference. Really want to know? By all means ask. Want to stir? - please don't. – Russell McMahon Jan 11 '19 at 06:41
  • My confusion about your answer likely reflects my own ignorance. My comments are ONLY an effort to learn and improve my understanding, and to help SO become an ever-better tool for disseminating clear, accurate, educational information. My confusion with your answer was completely sincere. I assumed by your first reply that you want engage in enlightening discussion. It's clear from your subsequent replies, you DON'T want engage in enlightening discussion, you just want me to "go away". I am contemplating your new comments about metals, but i shall respect your desire to avoid discussion. – johny why Jan 13 '19 at 01:32
  • @johnywhy As I noted - we don't seem to be managing to mutually communicate :-(. We could perhaps try again? | I was with you until " ...it's clear from..." -> I looked at all my comments above again. I **GENUINELY** don't know how you came to that conclusion - and it's not what I intended. "I'm here to serve" - fwiw. The following is not to 'count coup' but to demonstrate that what you think of me is inconsistent with the norm - we can no doubt both do better. ... – Russell McMahon Jan 14 '19 at 04:37
  • @johnywhy ... The system tells me that I've answered 2466 questions. [**121 [about 5%} relate to solder or soldering in some way.**](https://electronics.stackexchange.com/search?q=user%3A3288+solder) A substantial portion of my answers are the "accepted" one. Hopefully you'll suspect that we are managing to talk at cross purposes. – Russell McMahon Jan 14 '19 at 04:39
4

The only thing that springs to mind is that some components may get hotter than the solder and melt it?

It'd be quite rare for that to happen, but supposing you had a component which used some pins as a heatsink (some use ground pins as this), and it got hotter than the solder could cope with - the solder would melt, the connection would break down, the heat sink would fail, and the component would fry.

- This is just my thoughts, so is probably completely wrong ;)

Majenko
  • 55,955
  • 9
  • 105
  • 187