Spectrometry: The Marvel of the Lab!
By Jim Stark

We occasionally get questions about how oil analysis works. You send your oil to us and you get a report back, but what happens in the meantime? Is it magic? Some sort of voodoo? What happens to the oil that allows us to determine what's in it?

At the heart of most oil analysis businesses is a spectrometer. It is the machine that allows us to quantify wear metals, additives, and contaminants in oils, making oil analysis a useful service in predicting potential problems in machines of all types (though it's most useful for internal combustion engines).

The plasma in the process
A spectrometer can be aimed at a star to determine what elements may exist in the star, if all the star's light is being generated by the star (rather than reflected off the star). Spectrometry works on the same principle, but we have to first create the light. We do this by converting the actual oil into light energy. This is done by injecting the oil into something called plasma. You can think of plasma as a flame, since it looks like a green flame. But plasma is much hotter than a normal flame, and it needs to be in order to do its work. The plasma we use has a temperature of about 10,000° C. Plasma is actually the highest state of energy (the states of energy are solid, liquid, gas, and plasma).

Different types of plasma have been used over the last several decades that oil analysis has been commercially available. Early on, plasma was electrically generated as an arc. The drawback of an electric arc is that as it is generated, it can vary in intensity because the electrical part generating the arc erodes. The erosion causes changes in system resistance, resulting in variable plasma intensity. When using plasma to read the intensity of light from elements, it's best if the plasma's light is constant. Otherwise, errors can be introduced into the process.

Inductive coupled plasma, known in the trade as ICP, works by converting argon gas into plasma. So long as the argon pressures and flow rates don't change, and the power causing the plasma's generation is steady, the intensity of the plasma stays the same. This gives ICP spectrometry the industry gold star for incredible accuracy.

The rainbow connection
To understand what happens next, think of a rainbow. When you see a rainbow, what you're really seeing is moisture droplets in the air acting as prisms to refract (to separate) the various wavelengths of light into individual colors that can be seen by the human eye.

A spectrometer uses this same principle. A prism inside the machine takes the "light" that's generated by injecting the oil through the plasma and separating it into the different light frequencies of the elements. Each beam of light is then directed to a tiny slit on what is called an aperture plate. The aperture plate is a thick metal device, about 10 inches wide by 18 inches long, and the slits engraved in it are finer than a human hair. The aperture plate allows us to measure the intensity of each beam, using an ingenious device known as a photomultiplier tube.

A photomultiplier tube senses light and reacts to its intensity by vibrating faster as the light intensifies. Voila! By placing a photomultiplier tube at one of the slits on the aperture plate, we can get a digital readout of the intensity of light for any particular element in an oil sample. However, as amazing as this process is, the spectrometer is just as dumb as your home computer if the operator doesn't install a program that will let the computer strut its stuff.

Let's recap what we've learned so far. We know that argon is turned into extremely hot plasma, which burns the oil completely, turning it into light waves. The spectrometer refracts this light with a prism and then optically directs the distinct light frequencies of each of the elements to a slit in an aperture plate. A photomultiplier tube travels to each of the light slits and "reads" the amount of light there by vibrating. This marvelous arrangement still can't tell us what we want to know without further instructions.

Setting the standard
The next step in determining what is in the oil (and in what quantities) comes in the form of "standards." You can think of standards like your daily vitamin. Just as you can buy vitamins that contain a certain amount of iron, the iron standard (which is a liquid) contains a certain, "standard" amount of iron. You can buy standards that contain however much of any element you need.

Each standard has a certain amount of a particular element in it. If we want to know, for example, how much iron is in an oil sample, we need to give the spectrometer something to measure against. This allows it to know how many vibrations to count to determine how much iron is present. The first standard we use is a blank — that is, a zero standard — that has no iron in it. At the iron slit in the aperture plate, the photomultiplier tube vibrates at a certain rate per second. Then it remembers that rate as zero. Then, for example, a 100 ppm iron standard is fed into the machine, and again the photomultiplier tube vibrates, but this time at a faster rate. The machine remembers this rate is equal to 100 ppm. Setting the standards in the spectrometer is a process is known as calibration, and it's something we do two to three times each day. It allows the spectrometer to know what standards it should be measuring against.

The spectrometer records each element's information into a chart and uses the chart to determine how much of each element is an in actual oil sample. This process, where the photomultiplier tube travels to each slit and vibrates, repeats for each element we want to measure in an oil sample. The vibrations are translated to ppm (parts per million) readouts using the charts that were set up by the standards. Suddenly the spectrometer looks like a genius! It burns and oil and tells us how much of each element is in the oil.

There are 72 elements on the periodic chart that make enough light, when injected into the plasma, to be read on a spectrometer. Some elements make lots of light and are easy to analyze accurately. Others, like tin, make very little light and are more difficult to accurately gauge. This, along with differences in standards, calibration, and the set-up of different spectrometers, is the reason that you may find differences in the results coming from various laboratories.

A spectrometer is like your television or your car — you don't have to understand how it works to use it. There is only one answer to how much iron, copper, or any other element may exist in an oil sample. We think ICP spectrometry has the best shot at giving you the correct answer. It is accurate and repeatable, which is a requirement for giving you an accurate appraisal of how your engine is doing mechanically based on its wear properties.

2003 BMW 325i
Dinan CAI, Throttle body
Dinan Exhaust
Dinan stage 3
Dinan Strut Braces

Ipquick,

Thanks for the information. I found it very useful.

Is this the ‘free’ test that a lot of vendors supply to there customers?

I have heard that if wear partials are larger than 8 microns the spectrometer will not see them. Is this true? If it is, then I would assume that the PPM would be a little off if there happened to be large wear particles from, say, a gear fault. (Industrial use)

Thanks again for your input,

Red

Ipquick,

Can you help me out? I was really hoping you would answer my questions.

Or anyone , for that matter.

Red

Red,
Elemental spectrometry does have size limitations and does not generally detect particles larger than 8-10 microns because they are not fully consumed in the plasma described by Ipquick. This shoull not elininate it from your test slate as it does provide usefull data on wear rates of machine componants. Wear debris analysis utilizing Analytical ferography or Direct Reading Ferography will help identify particles larger than can be detected with the spectrometer. As for what is included in your "Free" lab package you would have to contact the lab or your lubricant provider to answer those questions. Any usefull lab test package should include Elemental Spectroscopy, Analytical Ferography or Wear Debris Analysis, Viscosity, Particle Count, TAN/TBN, Water Content, and FTIR Spectroscopy.
The Lubrication Excellence Conference in Nashville is only a few weeks away there you can find a wealth of knowledge and is worth the trip for any Lubrication professional. For information contact Noria.

Clyde Hughes, MLT I, MLA I&II

Noria Field Services

Clyde,

Your response is much appreciated.

Red

Red

Some labs don't accept samples if they are in transit much more than 24 or 48 hours. Is this something you've heard? They guy I heard this from gets his samples to the airport while still warm.

RH

I've never heard that. I wonder why that would matter?

Red

Red,

ICP analysis is not possible if the particles size is ( over 10 -12 microns) large. Alternatively you should do wear metal analysis of used oil through Ferrography. Predict Technologies in US is a big name in ferrography.

Hussam Adeni

Jim,

I agree with you in your article, this was the first tool to raise current level of oil analysis. In the case of new oils, generally free of wear particles, ICP work better. Normally RDE is better for used oil because sampler method can capture greater particles than ICP.

But, we need to know, as all thing in the world, this method has limitation. The limit is the size of particles contaminating used oils and samples. No over 3 -5 microns particles concentration can be meansure by ICP. Then, to increase the features of oil analysis, we can use Automatic Particle Counters. This methods can detect concentration and distribution of particles in used oil samples, from 2 microns and up. Better accurancy is between 5 to 75 microns or 6 to 70, agree with new ISO 11171-99 standard.
Wear particle analysis is an exception methode to be used when you find alter trend in Particle Count meansures. This method show microscopics details on size, quantities, shape and origin of particles in used oil samples.

Normally, as we say in Noria Seminars, is a good idea to work with basic analysis package, and then, when you find alter trend or over passed limits from you alarm group, you can use exception methodes as Analitical Ferrography. Basic analysis could be 3 or 4 critical as Viscosity and others and Exception analysis could be Analitical Ferrography because its bigger cost.

hi i find this article very informative,thanks.
however i have a further queries
a) is there any way,heaper way of testing the contaminants-limited to iron,brass,particularly for diesel engines?
b) for running engines,how can this be done online-and is there a laid down periodicity.what are the general standards followed?
thanks

I think you got some infowhich is quite similar to the one I shared with you
regards
Arupanjan

quote:

---

Originally posted by Clyde Hughes:
Red,
Elemental spectrometry does have size limitations and does not generally detect particles larger than 8-10 microns because they are not fully consumed in the plasma described by Ipquick. This shoull not elininate it from your test slate as it does provide usefull data on wear rates of machine componants. Wear debris analysis utilizing Analytical ferography or Direct Reading Ferography will help identify particles larger than can be detected with the spectrometer. As for what is included in your "Free" lab package you would have to contact the lab or your lubricant provider to answer those questions. Any usefull lab test package should include Elemental Spectroscopy, Analytical Ferography or Wear Debris Analysis, Viscosity, Particle Count, TAN/TBN, Water Content, and FTIR Spectroscopy.
The Lubrication Excellence Conference in Nashville is only a few weeks away there you can find a wealth of knowledge and is worth the trip for any Lubrication professional. For information contact Noria.

Clyde Hughes, MLT I, MLA I&II

Noria Field Services

---

Dear Hughes,

I am happy that you are a certified member of ICML. I am a Nigerian interested in ICML certification and I hope you gain a lot of knowledge from you.

Please help me with any material you think will be of help to me towards this certification. Though my area of interest in Lubricant Analysis.

My email address is caleen98@yahoo.com

I hope to hear from you soon.

Thanks and best regards

quote:

---

Originally posted by Clyde Hughes:
Red,
Elemental spectrometry does have size limitations and does not generally detect particles larger than 8-10 microns because they are not fully consumed in the plasma described by Ipquick. This shoull not elininate it from your test slate as it does provide usefull data on wear rates of machine componants. Wear debris analysis utilizing Analytical ferography or Direct Reading Ferography will help identify particles larger than can be detected with the spectrometer. As for what is included in your "Free" lab package you would have to contact the lab or your lubricant provider to answer those questions. Any usefull lab test package should include Elemental Spectroscopy, Analytical Ferography or Wear Debris Analysis, Viscosity, Particle Count, TAN/TBN, Water Content, and FTIR Spectroscopy.
The Lubrication Excellence Conference in Nashville is only a few weeks away there you can find a wealth of knowledge and is worth the trip for any Lubrication professional. For information contact Noria.

Clyde Hughes, MLT I, MLA I&II

Noria Field Services

---

Dear Sirs,
You often dispute about size of particles which can or not can measure by one or another analytical method. Don’t forget please that Used Oil Analysis has preventive, prophylactic and stop-all nature such as fire-control system.
If we have NORMAL wear process we can measure the metals by inexpensive and quick method very well, for example ICP. But if we have ABNORMAL wear process, when there are appear more and more of wear debris than at normal conditions and these debris has different sizes (maybe the big particles will be more than small particles), ICP would show INCREASED values of wear metals. It would be alarm signal for actions. In this moment more important that WE CAN FIX with ICP the increased wear, but it is not important what sizes these wear particles. We now hear the signal of fire-control system and we must to decide what actions needed turn on. Then later we can find the cause of fire - cigarette end or short circuit. Now we must suppress fire for non-admission of catastrophic damage. Analogously, we will find the cause of increased wear with ferrography, microscopy and others expensive and laborious methods.
In Used Oil Analysis we control the baseline. By the highest standards it is not important what accuracy these readings. The deviation from base line on some values would be alarm. Note that this deviation has the same accuracy. We saw only that normal process is broken, and we can see it by using ICP - old and tested expedient! Thanks the peoples which were develop and introduce this analytical method.