Let's order strange things from NIST | Hacker News Reader

Let's order strange things from NIST

twitter.com

243 pointsggm54 comments

I used to work in a lab that bought mud from NIST. We were measuring soil contaminants, and needed a calibration point for the data collected. So, whenever we collected data from river mud, we also ran the NIST mud. That way, any drift in our experimental setup over time could be normalized out.

Ok, I gotta ask. How does NIST source a large supply of mud that's stable in composition over a long period of time, both in terms of supply chain and storage?

That seems to be the case for a lot of these items. Either they have a process for making these products with sourced ingredients that are somehow hyper stable or they buy up 1000lbs of breakfast cereal, grind it into powder, then vacuum seal and freeze it in a vault somewhere alongside powdered sewage, all the time assuming nobody in the future of the planet will need more than 1000lbs of reference powdered breakfast cereal total.

As I understand it, each and every sample isn't necessarily uniform bur rather each sample is characterized and qualified by a NIST lab and a certificate of analysis (COA) is supplied with each sample. This allows tier 2 and lower labs to verify that their equipment and analyses are within spec.

There are exact standards both for the accuracy of the reference material and the precision of the COA.

So NIST can work backward, using faster tests to confirm that the material has suitable qualities to represent, say, 2% milk, then take the precise measurements for the actual batch of sample, which of course will vary.

> Ok, I gotta ask. How does NIST source a large supply of mud that's stable in composition over a long period of time, both in terms of supply chain and storage?

Technically, they don't. Obviously, they have some storage (this is one of those things you don't want to use just-in-time manufacturing). Instead, NIST (and other analogous standards organisations) have batches of materials and really, really expensive machines (that are hard to mass-produce) that can measure out the precise quantities of whatever thing they need to measure backed by processes that intrinsically is just precise (and accurate, for pedants), for example, electron microscopes or atomic force microscopes, which is then used by others (like in this case for measuring mud) to machines that are easy to mass-produce but can measure badly if not checked regularly.

For some things though (like drug testing in urine), they instead have a pure substance laced with precise amounts of substances you've expect to find, but that's relatively easier than most SRMs where you're measuring things you've already have (whether it's mud, crude oil, whale blubber or peanut butter).

https://www.si.com/mlb/2019/08/07/baseball-mud-rawlings fun story about consistent mud collecting for the mlb

So, starting with an analogy. Suppose you want to know the distance from Grand Central Station in New York City to a cafe a block from the Eiffel Tower. You have two options of how to make the measurement. Option 1 is to directly measure the distance. Option 2 is to use a known distance between Grand Central Station and the Eiffel Tower, then add the distance from the Eiffel Tower to the cafe. Option 1 has fewer dependencies, but Option 2 is much easier and (assuming your reference measurement is correct) will give a much more accurate measurement.

Similarly, any time you're making a high-precision measurement, you want to make it relative to something that's already known to a high precision. It doesn't need to be the same high-precision measurement each time, so long as it's something that gets you into the same ballpark as the thing you're measuring. That way, you don't need to make an absolute measurement, but only need to make a relative measurement.

For the mud samples, we didn't actually care that it was the example same elemental composition in the handful of NIST samples we had, only that it had a known amount of the trace elements we were measuring. That way, instead of needing to explicitly measure detector efficiency, everything became a very simple scaling question of "(X counts in sample) / (Y counts in reference) * (Z known trace element in reference)".

I used to work extensively with the NIST radioisotope SRMs. I particularly remember working with SRM 4965a (https://shop.nist.gov/ccrz__ProductDetails?sku=4965a&cclcl=e...).

My wife was very worried when I told her that I'd be working with radioactive hydrochloric acid. Technically, that is what it is, but the acid is very dilute and the radioactivity is very low. It sure sounds dangerous though!

I had to repackage it into a new container so that I could contain and draw off the radon gas, but still have the capsule remain wet so that the emanation rate would stay constant. My low-tech solution was to put a wet sponge in the chamber, so that the air would remain saturated with water vapor but that the condensation would not dilute the acid. It worked great for about 10 years. I passed that source on to another company, it might be working still. Radium sources tend to be rather long-lasting.

My father actually put up that SRM :)

Dr. Ronald Colle

This guy has no idea of what he is talking about. r. Colle, radionuclidic metrologist for 43 years.l Lll

An economist article about Standard Reference Materials:

https://www.economist.com/science-and-technology/2022/08/17/...

The MSDS for that jar of peanut butter says it's not an asphyxiant... as someone with peanut allergy, I beg to differ ;)

Nice submission. Brings back memories of making alcoholic drinks from expensive spectroscopy grade ethanol. We never dared hitting the deuterated ethanol for NMR purposes...

Were the drinks good? Less hangover inducing?

Regarding the deuterated version, it's actually possible it would have very different pharmacokinetics to normal ethanol. The first step in metabolism involves breaking the -O-H bond to create the =O aldehyde, so replacing that with a heaver -O-D would have a kinetic effect.

For example, here's a Derek Lowe article about deuterated drugs (not ethanol):

- "So a well placed deuterium (or two, or three) can actually have a significant effect on how long a drug will circulate in the bloodstream, by slowing down the liver's clearance mechanism."

https://www.science.org/content/blog-post/first-deuterated-d... ("The First Deuterated Drug Arrives")

Something to keep in mind is that since normal hydrogen is about 0.1% deuterium, all hydrogen-containing molecules end up also 0.1% deuterated by default - so the effects of a deuterated molecule can only differ from the normal to the point that in 0.1% concentration it would not be noticed.

> The MSDS for that jar of peanut butter says it's not an asphyxiant... as someone with peanut allergy, I beg to differ ;)

Just in case the sarcasm wasn't abundantly obvious, the precise term used was simple asphyxiant with the following regulatory meaning[1]:

>> Simple asphyxiant means a substance or mixture that displaces oxygen in the ambient atmosphere, and can thus cause oxygen deprivation in those who are exposed, leading to unconsciousness and death.

[1] https://www.osha.gov/laws-regs/regulations/standardnumber/19...

That'd be one expensive cocktail.

Always remember to drink in (neutron) moderation!

i'd be curious how a deuterated cocktail would taste. a few people have consumed heavy water and it tastes sweet, i wonder if this carries over to ethanol. naturally no sugar added!

Tom Scott & Half as Interesting did a few related videos:

https://youtu.be/9mAcJRFfAL0

https://youtu.be/nAsrsMPftOI

https://youtu.be/jvJzi0BXcGI

In my youth I used to work in a laboratory that tested mattress flammability and we used the NIST standard cigarettes. I remember people used to pocket a few of them and smoke them on break, not sure how good an idea that was.

I'd like to imagine there's a cigarette reviewer out there who compares various manufacturers/products against 'NIST standard cigarette' as baseline. I've never smoked, I just think that'd be brilliant.

“We give American Spirits 8/10 Nicotines.”

I've been watching through Lost, with all the Dharma Initiative branded food in generic black and white boxes.

Hah, the aesthetic is indeed similar!

And that in turn influenced by generics[1], who have been raised to a yellow art form by Canada's "No Name" brand[3].

[1] No, the other kind.[2]

[2] http://gbnfgroceries.blogspot.com/2014/01/from-misc-foods-ai...

[3] https://www.noname.ca/

Threadreader: <https://threadreaderapp.com/thread/1335061382501261312.html>

Id highly encourage anyone who is baffled by this to explore the field of science known as Metrology.

link not requiring a logged in account and javascript:

https://nitter.net/tubetimeus/status/1335061382501261312

This is just because they haven't come up with a quantum standard for peanut butter yet.

Now we need most expensive restaurant cooking only with NIST-rated ingredients

https://jabde.com/2022/07/29/immaterials-and-methods-reagent...

A mere $30k for a laboratory grade cookie. I'm surprised coconut oil is the most expensive ingredient at over half the recipe cost.

I'm not, it would also be the most expensive ingredient at the supermarket instead of NIST.

Your tax dollars at work!

An army marches on its stomach. Without a standards organisation to ensure high-quality peanut butter, the nation will slowly fall to ruin!

Haha, definitely true, but of course the SRMs also help for more serious things (like metal strength for vehicles, which can be the difference between life and death).

Excatly. Keeping instruments calibrated and true.

Having standards is one of the more important functions of government and this work has broad implications for consumer protection. The price is mot even cheap so at least the largest beneficiaries pay to brunt of the cost. If you’re serious about government waste I would suggest looking at larger budget items like Lockheed or it’s partners in the military first.

In fact, it’s specifically called out in the US Constitution. Article I, Section VIII authorizes the government to “fix the standard of weights and measures”.

This is a really good way to spend tax dollars, actually.

They probably aren't losing money on $800 jars of peanut butter.

They probably are since the methods and requirements are very different to what you buy in the store.

So's the ARPANet.

Remember, kiddo: Microsoft-Orcale-Apple didn't invent the Internet. They just jumped on.

It looks like they charge for the materials to offset the cost. This whole program, to ensure Americans get what they pay for in consumer items, is a much better use of my tax dollars than, say, paying for rich kids college loans.

Rich kids tend not to have college debt.

https://www.propublica.org/article/university-of-illinois-fi...

Not the only example, simply one of the first that popped up in my brief (~30 seconds) search.

Luckily we can do both

Not without bringing on double-digit inflation, angering the majority of the population with its unfairness, and putting the nation further into debt.

Press x to doubt. I haven’t seen any evidence that student loan forgiveness will cause double digit inflation.

I have seen baseless claims that the current inflation is being exacerbated by the student loan relief which hasn’t occurred yet, but as I said, they are baseless

You would rather have industrial makers have nothing to test and calibrate their impurity detection equipment on? Probably nothing bad will happen if they just wing it

This but unironically.

NIST also runs the national time references and the broadcast stations WWV and WWVH that share them with everyone. It's a hell of a good use of tax dollars.

Standardization is one of the key enablers for mas production.

I wish my country did 10% of the work that the NIST and USDA make.

Reply on news.ycombinator.com