Can I analyze NO, N2O and NO2 via GC?

2011-jaap-pasfoto4This question pops up a number of times every year.  The joking answer my former colleague Dick Zwiep (when we both worked at Chrompack time) used to make was:  No and No-two!

The challenge is not the N2O, but the NO and NO2. NO and NO2 are involved in several reactions/equilibria, that makes the quantification challenging.

N2O  (Nitrous Oxide) , also known as “laughing gas”, is quite easy to analyze via GC. One can do that on a porous polymer type phase, like the Rt Q-BOND, see fig.1.  The N2O elutes quite fast, meaning that a splitted injection is preferred. In this application a 0.25mm PLOT column was used as the detector was an MS. N2O is also often measured via ECD.

Fig.1 Nitrous oxide elutes as a symmetrical peak from a Rt-Q-BOND porous polymer.

NO,  (Nitric Oxide), is more challenging to measure.  When exposed to oxygen, NO is converted into NO2 (Nitrogen Dioxide), which has a dark brown color.

2 NO + O2 → 2 NO2


This is also one of the challenges that is faced. NO will immediate react with O2. If O2 is absent, NO will elute from most columns, but it will show minimal retention. Even on a porous polymer it co-elutes with permanent gases.

It is retained on Molecular sieve type adsorbents and  can be measured, but peak shapes are usually not symmetrical, which challenges elution and detection  of low levels.


Some more models of chemical kinetics on NO/NO2 can be found here:


NO2, (Nitric dioxide), is most challenging to analyze via GC.   NO2 exists in equilibrium with dinitrogen tetroxide (N2O4):

2 NO2 <=>     N2O4

Fig.2 Reactivity of porous polymers based on vinylbenzene, with NO2 makes quantitation difficult

Few applications are out there. Most obvious would be to use porous polymers for this, but the Nitric dioxide reacts with the porous polymer forming an ethylene peak, see fig.2,. This maybe a way to measure NO2 indirectly by converting the NO2 into ethylene in a hot inlet, however, the conversion is believed not to be quantitative.

For low level detection one can use MS, ECD, PDD ands also Cheluminescence detection. TCD is also possible, but only at high ppm levels


58 Responses to “Can I analyze NO, N2O and NO2 via GC?”

  1. Dear Jaap –

    Yes, N2O is a little challenging, but indeed possible to determine, even in trace amounts!

    We are using HP5890-II, GSV 1.00 mL loop, direct injection on 3 m x 2.16 mm ID x 1/8″ OD Porapak Q 80/100 mesh column, packed in stainless steel and the ECD as detector. The contents of N2O in air (approx. 0.5 ppm-v N2O), can be used as “reference” and elute as a little but very nice peak (analysis time 3 min.).
    The TCD can be used as detector when the N2O conc. is higher than 40-50 ppm-v. CO2 elute just before N2O and can interfere on the N2O determination, if present in high amounts.

    With kind regards –
    Lars Kürstein, Copenhagen, Denmark

  2. Phung D says:

    Good day

    I’m a new research student so has little knowledge on GC, and as part of my project, want to analyse N2O emission from waste water treatment. Our lab has access to the the Varian 450-GC (which i found some projects used it to measure N2O), and hopefully can access the fund to buy the required Ponapak Q or Rt Q-BOND columns. However, based on what i was told, none of the lab technician is familiar with measure gas using GC (they have only used it for volatile liquid samples) and some even warned me about how it will take all my years of research just to calibrate and still may not be able to measure N2O . I can access some outsider help who has more expertise with GC, but it gonna be limited down to 1-2 months and several booking sessions.

    So i want to know whether it is that difficult to learn and measure N2O using GC? Can i quickly learn how to do by myself (+self research) with support? Can i just use normal air or do i need the standardised gas to calibrate ?

    Will really appreciate any advices or opinions on it

    Best Regards

    Phung D, Australia

  3. Hi Phung
    N2O can easily eb analyzed with GC. You need a column that shows sufficient retention. Use a Rt Q-BOND and get very nice separations. If you sent me an email, I can share some chromatograms with you. The problem-makers are NO and NO2, not N2O.
    You need detector that can “see” N2O: use micro TCD or a MS for that. we have done both.

    For calibration you do need a standard you have to make or purchase.


  4. Phung D says:

    Good day

    Thank you, Jaap for the swift response

    As mentioned before, my biggest problem is the lack of confidence when dealing with this unfamiliar ‘machine’ (especially when both of my supervisor and staff technician also are not confident with it), so it is a great boost to hear assurance/opinion from another independent expert.

    In coming weeks, I will invest more time on exploring the situation with my lab GCs to see how it’s going :)

    Thank you once again


    Phung D, Australia

    PS: if you have any data and chromatograms that you can share, can you please send it to Even the most fundamental, obvious piece of information from you, still may assist a newbie like me to understand more on GC

  5. Aleksandra says:

    Hi Jaap,

    I need to determine N2O (500 ppm) in presence of NO and NO2 (10%). Usually, I analyse N2O in other gas samples by GC with TCD detector, but I am affraid that now, the big amount of NO and NO2 could be damaged the TCD…
    Are you really sure, that we can analyse NO and NO2 by TCD ??
    Aleksandra, Poland

  6. Hi Aleksandra
    You can measure N2O using TCD, MS or ECD.

    Gases like NO and NO2 are extremely difficult to quantify via GC because of reactivity. I don’t think the TCD will have a problem with NO or NO2, the problem is to elute it into the detector from the Injector/column system.

  7. Aleksandra says:

    Hi Jaap,
    thanks for your answer and…sorry for my english…:(
    In my previous question I wanted to write…” the big amount of NO and NO2 could damage the TCD”
    I was affraid that so good oxidizer like NO and NO2 could oxidize the filaments in TCD and lead to their desactivation or even their damage.
    I don’t want to determine NO and NO2, only N2O.

  8. Daniele says:

    Hi All,
    I also need to determine N2O in water samples (head space technique) but I am facing problems with GC set up. I work with a Thermo finnigan GC equipped with an ECD and a Poraplot Q (0,32 30m). At the moment the ECD operates at 340°C, with a ref current of 1 nA, pulse amplitude 15 mV and pulse width of 0.5 uS. Make up is 35 ml/m, oven 35° C and carrier 150KPa. With this configuration and an injection of 500 uL of air (about 66 nmol N2O) I am not able to detect any N2O peak, but according to literarature I should do. Probably there is something wrong in my configuration.
    Do you have any suggestion?
    Any help would be very appreciated

    best regards

  9. this should be no problem. Make sure you do a split injection and first try higher levels. Also your syringe needs to be open, and not “leaking”. Note that 150kPa is a high pressure for a 0.32mm column, it may show some restriction. If you want to see traces you best use a 0.53mm Rt Q-BOND column. This way you can inject larger volume with lowest split (1:5).

  10. A. Daverey says:

    Dear All,

    I want to analyze N2O gas by GC-ECD. But gas samples also contain oxygen, which will interfere N2O analysis and even affect the ECD detector. So, could you please tell me how to solve this problem in order to analyze N2O gas by GC-ECD?



  11. Hi Daverey
    N2O can be analyzed very welll with GC. It’s just NO and NO2 that have issues.
    Use a Rt-Q-BOND and run this column around 50C. N2O is a nice symmetrical peak. A 30/0.53 or 0.32mm Use splitted injection.

    Detection limits for ECD are also quite low. This application is done quite frequenstly.
    If you have any questions, plse drop me a mail.

  12. also note that N2O is well separated from O2 on Q-BOND porous polymer materials. O2 has near no retention at 50C, N2O has a k factor of about 0.5.

  13. A. Daverey says:

    Dear Jaap,

    Thanks for your valuable suggestion. I have sent you an email. Please check.

  14. Sharon Billings says:


    For years, we’ve run mixed standards of N2O and CO2 to generate calibration curves for both gases, using an ECD for N2O and a TCD for CO2, with good luck. Our matrix for the standards has been N2. Our samples always have both gases in them (from soil efflux). Now we are running only N2O, and created standards accordingly (N2O of known concentration in a N2 matrix). Our calibration curve looks halfway decent, but pretty sloppy (instead of r^2 of 0.995 like we used to get, more like 0.96 or so). We see quite a spread of peak area values at the higher concentrations. Our range is atmospheric (~320 ppb/v) to ‘high’ by our standards (5000ppb). When we include CO2 in the standards, the N2O data are dead on again. Can anyone think of a mechanism that would explain this? Nutshell version: when CO2 is part of the N2O+N2 standard matrix, ECD-derived N2O values are fantastic for calibration. When N2O only is in the standard mix (with N2), the cal curve is sloppy.

    Thanks in advance,

  15. Roberta BF says:

    Dear Jaap,

    I am was hoping to measure N2O from my samples but from another trials I have seen really high concentration on NO (mainly) – my samples are kept anaerobic. The machine is a traditional HP 6890 with a ECD detector (flow 75 – 5% methane and 95% argon, temperature oven: 35oC ) and has a sample loop where the sample first pass through the column Hayesep T (1/16′ – 1 meter) and then through Hayesep Q (1/16′ – 2 meters).

    When I calibrate the machine I can see a beautiful peak at 2.71 for N2O. However, as I explained before, my samples have high amounts of NO (I was trying to measure N2O as result of the denitrification process) but I have a huge peak right before (at 2.54 min) that totally covers the peak that could be N2O (at 2.71 min).

    I am trying to solve this problem by myself since research money is difficult nowadays, but I can’t find a solution. Ah, another point I would like to add is that I tried putting 10.4 ppm of NO into the GC-ECD above and I don’t see the big peak that I was hoping to explain my problem (even though my samples probably have hundreds of ppm of NO).

    Could you please give me some advice or some idea on how I could eliminate this “interference peak” (that I think is due to high NO) for the measurement of N2O? Any idea or suggestion is really welcome.

    Thank you very much for your attention,


  16. Hi Roberta
    The most easy way is to get rid of the NO before the analysis;
    NO reacts immediately with O2. Is that an option? NO2 will generally not elute from GC columns.
    I do not understand why you use 2 columns in series. This analysis should be OK using the Q only.

    The separation between N2O and NO should be no problem, because NO elutes almost without retention. I wonder if your column is still efficient. Can you sent me a chromatogram what you get? You may also creat a big injection error. Reducing sample size will help also to reduce the injection band-width
    If you need more separation between the NO and N2O, you can run at a lower temperature? Retention time will increase, but it may work.
    There are several other ways to do this separation, I wil sent some options.


  17. Hi Sharon,
    What column are you using? It;s probably an adsorption column. When you look at the chromatograms, do you see a better “peak shape” when CO2 is present? CO2 can have a “deactivation” effect on adsorption columns that will result in better peak shapes for N2O. This normally should happen at lower concentrations.
    As ytou stated you get better results at higher concentartions is quite opposite of what one would expect.
    What level of CO2 are you talking about? Is it the same level in all the N2O standards, or is CO2 also higher in the higher N2O standards?

    Other then this, CO2 may have a kind of interaction (cleaning effect) on ECD foil, which make sit behaves more linear.
    You may check this by running N2O standards and in between each concentrations shoot a CO2 sample.


  18. Sharon Billings says:

    Hello Jaap,
    thanks for this response. Yes, [CO2] typically co-varies with [N2O], so higher CO2 with higher N2O. Our CO2 concentrations go up to about 5000 ppm (notice N2O only goes up to about 5000 ppb, or 1000X lower for N2O). The column is Porapak Q. I have not noticed improved peak shape for N2O when CO2 is present, but that might be the case – if so, it’s not shouting at us.

    Do you think it’s CO2 influencing the ECD foil, and I’ve just never noticed this before b/c we’ve always had CO2 in both our standards and our samples? And what in the world is CO2 actually *doing* to the ECD foil?

    Thanks for sharing your knowledge. So valuable!


  19. Michael says:

    You mentioned in your introduction Jaap that a TCD is only appropriate for high concentrations of N2O. Could you tell me how high please?

    I’m wondering if I can use this instead of buying an ECD.

    Many thanks

  20. It depends a little on the type of TCD and your matrix. Today’s micro-TCD’s for GC platforms can see about 10 ppm. The portable systems with micro TCD can see 1 ppm. The amount depends a lot on the retention of your analyte: if the peak elutes fast, it will be higher The challenge is also that TCD is a universal detector, and “sees” everything. That means, if your matrix is different from the carrier gas, it will produce also a huge peak. ECD is selective as it will not respond to gas-matrix.

  21. Hannah says:

    Hi Jaap, I came across this blog via your colleague Russell at Restek in the UK, who has been extremely helpful in recommending methods and columns for simultaneous analysis of N2O, CH4 and CO2 by GC.

    However, if I give you a little more detail about our requirements, I would be very grateful if you could clarify a few things for me.

    We have 3 GCs in our laboratory with the following detectors installed: 1) MS+FID; 2) MS+ECD and 3) TCD. The concentrations of gases in our samples are expected to be between <10 – 100 ppm for N2O and CH4 and 5,000-50,000 ppm for CO2, and the sample matrix is air (we will be measuring gases emitted from soils, and they will also contain moisture – not sure how much and whether this will be a problem). We would like to use a headspace autosampler to inject the samples to the GC. My question is, can we analyse the gases simultaneously with our current set-up OR if we need to measure the gases in separate samples, what combinations of gases should we analyse with our current configuration of detectors?

    Otherwise, we may need to install a new detector on one of our GCs, and from the information Russell provided it seems as though a TCD –FID in series would be the best option, although I’m not sure whether this would be sensitive enough for our expected levels of gases.

    Apologies for the long message – any help or comments would be greatly appreciated!!

    Many thanks, Hannah

  22. Hi Hannah
    I gave Russel already a recommendastion on this.
    The N2O can best be done using ECD to get the low sensitivity;
    The methane can be done with FID. I would use the TCD-FID in series for doing CO2 and methane. N2O may also give a signal on TCD, but only if higher levels are present.

    Temperature will be around 40 C. Need to use splitted injection, minimal 1:5. Use the 30m x 0.53mm Rt Q-BOND.

    If you can set it up in one GC, you may also use one column and split eluent to 2 detectors using a “Y” piece.
    Of course you can also have 2 injection port and run these analysis parallel.


  23. Gerardo says:

    Hi jaap!!

    if i using the “Y” separator for ECD-FID together, is necessary configure the flows “makeup” of a especial mode??

    The flow of the carrier, will be modificated?

    whats is the especification of separator “Y”?

  24. Hi
    make up gas is a separate flow taht is typically used in ECD. Its separate from carrier gas flow.

    If you use a “Y”, the dimensions of the fused silica tubing after the “Y” piece will determine the split of the flows.
    Usae deactivated fused silica tubing for that. You may also use a section from an older Rtx-1/5 coated column.


  25. Gerardo says:

    The 30m x 0.53mm Rt Q-BOND
    with this column i can analysing methane?

  26. Yes, methane is separated from air using the Rt-Q-BOND porous polymer columns. This separation is temperature dependent, but is possible. See:

    Additionally, if you use a FID, this detector will give no signal for inert gases.

  27. Gerardo says:

    and the dead time… ?

    if I use a FID… the First peak will be methane?

  28. jaap says:

    yes, First peak detected with FID will be methane on a Q-BOND porous polymer column.


  29. Gerardo says:

    this is correct for cuantification of methane?

  30. jaap says:

    If you use split injection, this for sure will give you reproducible methane rsponse.

  31. Gerardo says:


  32. Gerardo says:

    do you recomend me some books or text of the cuentification of first peaks?

    thanks for all!

  33. jaap says:

    sorry, GC is an analytical technique which is usually part of an education in chemistry. If you need to use GC and get correct data, I recommend you to follow some GC courses to understand the basics. You can also go on-line and find a lot of basics there.

  34. André Cruz says:

    Hello Jaap,

    I need to analyze N2O in O2 (gas) production in a TCD. I work with a GC Valco-3800. The range i need to measure is 0-100ppm. I will like to ask you about the method that i have to use, like time retention and others.

    Thanks for the support.

  35. Hi andre
    This should be possible although TCD has limited sensitivity. You may get to about 10ppm, not lower.

    That’s why they often choose for ECD or even MS. If you use Rt Q-BOND, retention factor for N2O is about 1 at 35C.
    so you know where to expect it.


  36. André Cruz says:

    Hi Japp,

    The TCD has a WCOT as a column. Can i use this column to analyze N2O in O2 (gas)?. if i can, can you tell me please about the column oven temperature, time retention, flow and the others parameters that i have to use for this analysis.

    I have another cuestion, do you have another method to analyze N2O in O2 (gas), it’s doesn’t matter if is quantitative or qualitative.

    Thanking your support and time.

    Best Regards

    A. Cruz

  37. Alex Sopilniak says:

    Hi Jaap,
    We would like to analyze N2O at low concentrations of about 1 ppm and less , by injecting the gas into a GC-MS with a diffusion pump. When you used a plot column Rt Q-BOND, did you protect the MS from flying particles? if so, what did you use to catch them? Is it safe to use this column with a turbo pump?

    Thank you in advance,

  38. Hi Alex
    the PLOT columns that we manufacture, in general do not need particle traps as they are quite stable. we do advise to “flush” a column with twice the gas velocity, when it has been transported. This because often transportation is affiliated with uncontrolled situations.
    In case of MS, there is another problem, and that is the high vacuum that can rip off PLOT sections. For MS work, we advise to couple any PLOT column with 5 m x 0.25mm deactivated fused silica (like hydroguard) and this will take care of the big pressure change at the end of the column.

  39. Umair says:

    Hello all,

    Can anybody let me know about how to calibrate the Agilent 6890 GC ECD for Nitrous oxide. What is limit of GC ECD for Nitrous oxide. I am working on wastewater nitrification.

    Thanks in advance,


  40. Blagoj M says:

    Dear Jaap,
    do you, or anyone reading this, have experience with Valco Pulsed Discharge Helium Ionization Detector in ECD mode? They claim it is as sensitive as the radioactive ECD, but it does not contain a radioactive foil. Any example chromatogram?

  41. The PDHID or PDD is very sensitive. It uses ultrapure helium and is basically universal detector, meaning it responds to almost all components. It is a very nice detector if you need to measure inert gases, CO, CO2 etc. It is 50-100 x more sensitive then the TCD. It needs some experience.
    I do not know what they mean with “ECD-Mode”. The ECD is very selective and sensitive for electronegative components and is often used for trace pesticides. It is quite a stable detector if you treat it well. I have not seen C’grams of such components with PDD. We have one we used for gas analysis, see

  42. with ECD you should see sub ppm levels. make sure you use 0.53mm columns, like Rt Q BOND. use splitted injection as this peak elutes fast and injection errors always impact the early eluting peaks.. calibration: need standards of N2O in range 0.1, 1, 10, 100ppm in a neutral gas. (N2)

  43. KIRAN KUMAR says:

    Good noon,
    I am kiran kumar i am new user of Thermofisher Trace 600 GC. We are using PorapakQ and 5% OV-17 columns. and FID, ECD, TCD detectors. I want toknow now what all can analysed by this configure and detection limits for different combinations please help…

  44. Kiran, You can do many applications, but before using, I would recommend you to follow some GC training, so you can understand basics of column and application. Restek company will be happy to help you with any application on GC or LC. For now, Porapak Q is for gases and volatiles. the OV-17 is a liquid phase used for less volatile materials. For general purpose application, Instead of the OV-17 phase, I would recommend to use a less polar phase, like the Rtx-1 type, which is also available as a bonded phase in packed column configurations. The OV-17 is not a very efficient phase and always shows some activity.
    Detection limits depend on many parameters, like type of component, retention, peak shape, matrix, amount injected, column efficiency, type of detection, system background, so I cannot give good answer. If all is optimal, you can detect around 1-5 ppm with FID.

  45. Jim Mitchell says:

    Hi Jaap,
    I have an Agilent 6890 and 7890 fitted with tcds and fids. I would like to analyze for NO and NO2 in air and CO2 and have tried N2O and not had a problem getting a detector response using tcd. Can I buy a column and use my existing detectors for this NO NO2 analysis or is the answer still no and no-two? At present I am unsure what levels of NO and NO2 will be present in the bulk gas.


  46. Hi Jim
    Thanks for asking. Indeed N2O is not a problem. This component elutes from most materials. Rt Q-BOND, Shincarbon and even a Molsieve 5A.
    NO will elute from a Rt Q-BOND, but it will come together with N2, O2 and CO. Also, as stated in the No- No-2 blog, NO is very reactive and will immediate be converted in NO2 with just a little O2. NO2 I have never seen as a decent peak on any type of GC-column.
    NO can elute from a Molsieve 5A, but as a strong-tailing peak.

    Regards, jaap

  47. Muqthar says:

    Hi I have GC-450 from Bruker I tried with application note 1303 to find the N2O-3.00 ppm in Oxygen in it, But it not support, but it support for 1.00 % of N2O in N2.. Please advice me the method.

  48. For columns, the porous polymers like Rt U BOND and/or Q-BOND are mostly used.. But also detection must be OK. Most used is ECD and MS.
    Make sure when you use MS, to use a restriction column to deal with the vacuum and high pressure drop at the end of the column.
    j de zeeuw

  49. Marcos Gaertner says:

    Hi! I was looking for a way to analise NOx by GC and found your blog.
    As you said, N2O is not big deal to analise, but NO and NO2 are. Regardless of NO reactions I´m worried about separation and detection.
    In air samples, using the Q column you´ve mentioned, do I have any contaminant that may interfere with NOx peaks using an ECD?
    With MSD may I use a SIM method to minimise interferences? Or is the NCD the best choice to obtain selectivity?


    Marcos (Rio de Janeiro – Brazil)

  50. Hi Marcos,
    N2O will be separated from these 2 components. NO elutes VERY fast, NO2 I have not seen elute as single peak.. If it elutes it’s later then N2O.
    Mostly CO2 can interfere if a lot is present

    Most use ECD; MS you can select the typical fragments for N2O. (28,30,44).
    NPD is not a very stable detector.


Leave a Reply