I have recorded TG of ammonia borane in air and N2 atmosphere at 10 K min-1 and 2 K min-1. In all cases the resulting TGs are very strange (see attached picture). Namely, after 100 °C a decrease of the mass is observed, as expected. However, above 110 °C, an unexpected behaviour, i.e. steep increase of the mass, is evident. This happens up to 150 °C. After that, mass slowly decreases again. I could imagine that oxydation occurs in air atmosphere, which would explain the strange shape of TG curve. Thus I'm surprised when the same happened in N2. Also, variable temperature IR spectra, as recorded in air show no evidence of BO stretching modes, which would arise in the case of oxydation.
The similar behaviour is mentioned in Frueh et al. Inorg. Chem. 50 (2011) 783. They also repported a "normal" TG when recorded in Ar. But I didn't find any explanation of this behaviour in that paper.
Do anybody have an idea what happens?
Nikola
If this was done in an open system/container, I would guess that ammonia borane adsorbs moisture from your atmosphere. Gas is never 100% clean and flowing atmosphere only makes things worse.
The other thing that comes to mind is nitrogen replacing boron and creating ammonia. Here the mass goes up to 107% and theoretically it would be around 110%, but there should be overlapping continuous mass loss, like the one you get above 150C. The mass gain in nitrogen looks about right for this. TG-MS would help a lot here, or you could stop the heating around 120C and analyze the product.
The plot suggests that the initial weight loss that occurs up to around 110 ºC is seemingly favoured by the N2 atmosphere, while the weight regain that starts just after and continues to about 165 ºC is apparently favoured in air. Following your own words, a 'normal’ TG was found reported when Ar was used. Notice that we may reasonably expect that the humidity content of air is possibly greater than that of N2, while for Ar would probably be the lowest. Therefore, unless the air and N2 used was submitted to efficient drying treatment at the laboratory, we may have to consider the apparent possibility of humidity interference.
Let us consider a possible humidity uptake prior to TG analysis, what seemingly might translate as:
H3NBH3 + (2+x)H2O → 3H2↑ + (NH4)BO2·xH2O
Being this the case, humidity loss might occur upon heating, possibly according to:
(H3NBH3)y·(NH4)BO2·xH2O → (H3NBH3)(y-1)·((NH4)BO2)2 + 3H2↑ + (x-2)H2O↑
That might be possibly followed by further humidity uptake from the selected gas ─ according to first equation ─ with net weight gain. Finally, thermal decomposition, possibly accompanied by NH3 loss, could easily explain final weight loss. Complementary DSC data could be certainly quite helpful to check the validity of the proposed tentative explanation, which would require also further analysis with dry air and anhydrous sample.
2) Environmental factors during the experiment. What species can we expect to possibly react with melting decomposing ammonia borane and the products of the decomposition? Water definitley. Oxygen. Maybe nitrogen. Maybe carbon dioxide. Water is probably the most reactive with AB, with its solid state dehydrogenation productcts, and its gas phase decomposition products H2B=NH2 iminoborane, IB) and borazine. As you heat that furnace up, you are going to desorb water from the walls of the furnace, so no matter how dry your gas flow is, you are going to expose your sample to some water. Not one to rule anything out, it may be that if your gasses are dry and pure, you have the beginnings of evidence that nitrogen reacts with ammonia borne ... and though I doubt that is the case, an ultra high purity argon environment is you're best bet.
I am the author of the Frueh et. al. paper in inorg. chem. that you mentioned. I do not have clear explanation for the behavior you are observing, because there are many possible behaviors and combinations of behaviors that could produce the results you are observing. But from my experience I can offer three general areas of consideration which may help.
1) Sample Quality. What are you actually starting with? I suggest running an IR of your source material, and check it against other published spectra, with an eye towards O-H stretch. If you see O-H stretch, i'd say look for a better sample. Of course there are other possible ways you're sample may be compromised, and they are likely not to all be found in the literature. Purchasing a small quantity of high purity ammonia borne from a reliable source with a quality control workup available might be a good step.
3) Physical and Instrumental factors. If you heat AB up in a clear glass under ultra high purity flowing argon, first you will see it melt, and within seconds it will bubble vigorously releasing hydrogen, (and a low mol fraction but high mass fraction of borazine and IB). In the exhaust away from the heat the IB will condense and polymerize to a white powder. Then this bubbling churning hydrogen evolving ammonia borne liquid polymerizes to a condensed solid, which is like a foam, and has expanded in volume 10-fold or so. My first attempts at TGA saw dramatic weight increases. Then it dawned on me this expanded foam was pushing against the thermocouple directly above the sample pan, effectively pushing the pan downward, which the machine registered as a weight increase. Therafter I used tweezers to load just a couple milligrams of sample into the sample pan to avoid foaming effects. Even then I saw seemingly anomalous behavior, such as weight in increases above a certain temperature, that did not correlate with data I was getting from RGA mass-spectroscopy. Eventually I just used the TGA insturment housed entirely within a lab-built argon purged glove box with an argon purged antechamber to pass samples and tools through, passed ultra high purity argon through the furnace and monitored the exit gasses with an RGA to detect evolved gasses from the sample and ensure the absence of atmospheric gasses. Ran the furnace hot and empty to remove absorbed moisture before running any samples. There was of course no molecules that included argon detected by the RGA nor was argon detected, of course, in elemental analysis of the solid residue. The TGA curves, then represented true pyrolysis, rather than thermolytic reactions with trace atmospheric gasses.
To put all this simply, you just have to troubleshoot. The possible set of explanations are large and in part specific to you're instrumentation and how it is condigured... no two TGA's are alike. ... and it might take a hundred runs or so of trial and error to get something sensible and reproducible.
One thing is clear, if you have mass increase, that mass should be absorbed from the environment. This "environment" was not found in Ar, but must be in N2 or air. My first idea is also traces of H2O, but there is no B-O bands. Susbtituting of B with N seems to be improbable due to strong N-N triple bond, however, hydrogenation of N-N bodns cannot be completely excluded without study. The suggestion about TG-MS is correct, but TG-MS and TG-IR should also been done. If you want, we can help in it. Furthermore, analysis of the product formed with IR/Raman, or 11B NMR could also be an important step to assign the possible product(s). USing wet Ar or 20 %O2/80%Ar gas also would be informative, to assign whether H2O, O2, or N2 has importance in this process.
I wrote erroneously hydrogenation, it is hydroboration, of course. Additon of B-H bond to N-N triple bond.
Thank you for answers. Just a few clarifications..
1. I have purchased technical grade AB (Aldrich, 90 %), but I recrystallized it 3x from diethyl ether before use.
2. TG / DTA was recorded only to support the experiment of central interest, which in this case was in-situ temperature dependent IR spectroscopy of AB. This is also not so straightforward, but for this moment details are out of scope of this discussion.
3. IR spectra of AB at ambient conditions show that at start I have pure AB, without any traces of water (detectable by IR). The features due to BO bonds are also completely absent from spectra. The interesting things start to happen above 100 °C. Comparison of variable-temperature transmission and ATR measurements indicates considerable differences in processes occurring on the surface (ATR) and in the bulk (transmission). However, for this moment, I only want to say that only at surface of the sample some (not extensive) evidence of oxide formation can be visible. So, IR shows that oxide formation cannot satisfactorily explain so big increase of the mass evident from TG. Although it is not so clear, at higher temperatures some features arise, which could be attributed to hydrogen bonded v(OH) (broad feature between 3300 and 3000 cm-1) with remaining NH and BH groups (broad feature between 2050 and 1800 cm-1, observed only in ATR spectra - thus, it occurs at surface).
The other features can be explained by regular mechanism of dehydrogenation of AB through [H2N=BH2]x and [HN☰BH]x.
Off course, this "mysterious" TG results clearly show that this system should be investigated in much more details. You mentioned here some of techniques which could be applied and I would like to do it. I could measure in-situ TG-IR (gaseous products), but I'd be very happy if we could do additional measurements together.
Dear Nikolai, We can help with TG-MS measurement in He , N2 and air atmosphere.
After preparing some mg of sample with increased mass, XPS or perhaps 11B NMR could also be done. You can take contact with me directly, [email protected]. I think, it will be possibility in january. If you reques them, I will arrange a possibility to do them.
Dear Nikola, I am sorry I gave the e-mail.erroneously.
it is [email protected]
or
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude