In principle, it is possible because FTIR spectrum is response of lattice vibration clacified by point group of materials, e.g. glass is not crystal. However it is diffcult to determine them quantitative concentration because of the poor sensitivity of FTIR for the quantitative determination.

As Infrared is a vibrational technique the spectrum contains information on both the chemical and physical state. The overall spectrum of crystalline material will have much sharper bands than the spectrum of amorphous material. In the "fingerprint" region of the spectrum crystalline nylon and long chain acid/esters can be identified this way. Good examples are the spectra of C13-C17 acid/esters. The melting point is about 35-40c. If you prepare a film between salt flats (or pressing on an ATR crystal) the finger pressure will melt the sample and give a broad liquid like spectrum. On standing the "film" crystallises and the resulting spectrum contains sharp bands from which the chain length can be identified. As far as your samples are concerned it is suck it and see. With some inorganics Raman spectroscopy is more sensitive to the crystalline state. However the previous comment on the difficulty of obtaining quantitative information is valid in many cases but not all, try it

Raman may also be applied..to distinguish between polymorphic forms since differences in conformations & inter-molecular interactions should result in peak shifts. But as far as quantification goes not sure.

I'd expect that it's possible. IR is very sensitive to intermolecular interactions, which largely determine the structure of material.

Since you're talking about glass-crystal phase transition, I recommend in-situ temperature-dependent IR spectroscopic experiment in transmission mode. This is extremely informative, especially with respect of phase transitions. Off course, if you need some help in analysis, feel free to contact me.

Each unit cell having n atoms can be considered as an n-atomic molecule, hence there are 3(n-1) vibrational degrees of freedom, and correspondingly there are 3(n-1) infrared peaks. However, in addition, in a crystal the unit cells are related by a fixed phase and a wavevector can be defined for the entire crystal, an acoustic phonon, which is actually a long wavelength modulation on these peaks, resulting in a broadening of these peaks into bands whose strength scales nonlinearly with crystal size. In a glass, these long wavelength modes are absent or weak. So the infrared peaks do not broaden into bands that are much weaker and whose intensity scales linearly with sample size.

FTIR measurement can help us to determine the bonding constant. If the Zachariasen hypothesis is used, it would not be possible to differentiate between the glassy or crystalline form. The best bet is that the crystalline form would give relatively sharper peak than those of glass. Quantitatively, the convolution area or volume of each peak can be compared.

In the case where the glass is very crystallized the FTIR bands will be significantly sharpened. Quantitative assessment is difficult with this method, however. We did some of this work on sodium aluminosilicate crystallization in nuclear waste glasses and published the results in [McCloy, J.S., C. Rodriguez, C. Windisch, C. Leslie, M.J. Schweiger, B.J. Riley, and J.D. Vienna, "Alkali/ Alkaline-Earth Content Effects of Properties of High-Alumina Nuclear Waste Glasses," In Ceramic Transactions, K.M. Fox, et al., ed., vol 222, Advances in Materials Science for Environmental and Nuclear Technology, John Wiley & Sons, Hoboken, New Jersey, 63-76 (2010).]

Yes, it is possible. The extreme case is pure SiO2 glass (amorphous) vs. quartz (crystalline). They can be easily identified and differentiated. Overall, like others have said, quartz has much sharper peaks. Glass will have a high broad peak around 1050 cm-1, and another broad low peak around 780 cm-1. Quartz has a main peak at 1084cm-1 and a very characteristic doublet at 797+779cm-1. The doublet will become broader and less sharp as disorder increases, till it becomes one broad peak (like in glass). I would try first the two extremes and a middle case (does not need to be exact middle, as it was mentioned before, it's difficult to quantify exactly), check the common peaks, and measure their FWHM (full width at half maximum). The FWHM should be your measured parameter, and it should increase the more disordered your sample is.

Iwamato et al. has published a method to characterize the crystallinity of PVA (Poly vinyl alcohol) using FTIR, I attached the paper for your revison

I have just made some exciting measurements in the Far-IR. As this region is associated with crystal lattices it should help with the characterisation of amorphous and crystalline forms.

In general, since vibrational spectroscopy (both Raman and infrared) gives information on how atoms are connected, in principle you should be able to distinguish between amorphous and crystaline materials. Any other problems may be sample and/or instrument related and you may not know until you try it, as Geoffrey Dent puts it, just try it and see.

As pointed out by previous researchers, the main difference between Raman spectra of vitreous and crystalline phases lies in the bandwidth of the peaks: rather broad for glass and sharp for crystal. Another difference is that several sharp peaks appear in crystal spectrum while glass usually exhibits one main peak corresponding to the fundamental vibration mode and overlaps with the corresponding peak of the crystal.

Frankly speaking, using vibrationnal spectroscopy to distinguish glass and crystal is extremely at risk for inorganic materials. You can draw any conclusion you wish! It is wiser to rely on other techniques: XRD and DSC. Raman may be added as a complement to confirm the conclusions arising from these measurements.

G.Saffarini

As other researchers pointed out, it is possible to differentiate between glass and its crystalline counterpart. However, it is always preferable to add another technique, such as XRD.

If you want your results to be well accepted, getting the same results using multiple techniques is advised. Using powder XRD you could see sharp crystalline peak (for crystalline material) and broad halos for glassy material. This is the method of choice. Using DSC you could detect the glass transition temp in glass material and sharp crystal melting point for the crystalline material. If there is a mixture of both you would see both types of peaks. This may be useful for quantification too. Using Visual microscope under cross polarised light the amorphous matrial would appear dark while the crystalline matrial would look bright (unless your crystal space group is cubic!), which may turn dark if the analyzer is rotated by some degree. If you have access to transmission electron microscope, one can do electron diffraction to see if you get any diffraction pattern for the particles which appear to be crystalline. Only diffraction pattern would confirm the presence of crystals. You may also try the FTIR techniques which others have suggested, but the results would alway be in doubt unless you corraborate it with other techniques too.

As has been mentioned by Prof Dent, FTIR is a vibrational technique. The position of the vibrational peaks is depend on the vibration of the atoms, regardless they are in the glassy form or crystalline. Unless they are organic based glass, for inorganic based, the best method is XRD. Using HRXRD could clearly differentiate between those materials. Even better if you can use SAED that can give you up to the atomic spacing distance of the atoms. This technique will give the results that surely will give you what you really looking for.

I think the problem may be even more complex..

Off course I agree in general about the band sharpens change parameter. But You have to remember that during crystallization process in glasses the phase separation may occur as a pre-stage before crystals are formed, this may cause molecule environment change which at the end may cause the molecule symmetry change giving a very different spectrum of your crystalline products. So it's not obvious to use FWHM as a crystallization degree parameter in this case (using FTIR) and results may strongly depend of the type of material you are synthesizing.

In my opinion XRD is obligatory to distinguish the amorphous or crystalline state. You can use FTIR once you calibrate the process, in different stages by XRD (for example comparing samples with different cooling rates) or trying to perform experiment in opposite way from amorphous phase you can try to crystallize the glass by heating procedure (the temperature should be above the glass transition). Most effective is to choose crystallization temperature determined for example by DSC - this way you can also check if the material crystallizes to one or more crystalline phase - you should see how many exothermic crystallization peaks you have in DSC thermogram and compare it with XRD on different stages.

We have been also struggling with the problem in oxide glasses you may see some of our results (perhaps it will be helpful) :

Mroczkowska, M., Nowinski, J. L., Zukowska, G. Z., Mroczkowska, a., Garbarczyk, J. E., Wasiucionek, M., & Gierlotka, S. (2007). Micro Raman, FT-IR/PAS, XRD and SEM studies on glassy and partly crystalline silver phosphate ionic conductors. Journal of Power Sources, 173(2), 729–733. doi:10.1016/j.jpowsour.2007.05.048

And also there is some other examples for different appearance of the FTIR spectra of crystalline and amorphous phases of the same compound where you need to choose which maximum is important for the crystalline state.

See CaCO3 example - more crystalline it is, (more like calcite), the bigger V4/V3 bands intensity ratio value (so you have relative intensity change effect also) see the reference:

Gueta, R., Natan, A., Addadi, L., Weiner, S., Refson, K., & Kronik, L. (2007). Local Atomic Order and Infrared Spectra of Biogenic Calcite. Angewandte Chemie, 119(1-2), 295–298. doi:10.1002/ange.200603327

But also you can observe CaCO3 in vaterite crystalline structure which presents different spectrum than calcite and aragonite polymorphs (obviously because of the different crystal structure, bonds length change, molecular symmetry change...)

see the spectra in the paper:

Rodriguez-Navarro, C., Jimenez-Lopez, C., Rodriguez-Navarro, A., Gonzalez-Muñoz, M. T., & Rodriguez-Gallego, M. (2007). Bacterially mediated mineralization of vaterite. Geochimica et Cosmochimica Acta, 71(5), 1197–1213. doi:10.1016/j.gca.2006.11.031

so it's good to know what are possible crystal structures at the end of crystallization process to decide if you can use FWHM or something else as a tool to determine crysstallinity of the sample by FTIR, (in case of glass crystallization the may be also get meta stable crystals but that's another story :))....)

sorry for so long post I suppose You already know a lots of mentioned above :) but I'm passionate of the subject :)

Regards Maja

Yes, it is possible. I would pay particular attention to a few things though:

1) Chose the best spectral resolution;

2) Check changes in both peaks positions and widths;

3) Ensure that your sample preparation does not induce a physical change (ie pressure of a golden gate for example);

4) Combine with PXRD if possible.

I agree with Luca Quaroni ..... its perfect solution ....

but in base if you want to get quantitative data of both sets, then XRD and HT-DSC is advisable techniques....

agree with @ Lior Regev

regards

Follow this link maybe helpful

http://www.andersonmaterials.com/ftir.html

regards

If you look a bit further toward the visible, where the hydroxyl bands for inorganic glasses are located, you will find that those bands are much sharper than for the same glass. We used this method to identify the phase crystallizing in common glass-ceramics (like the old Corningware and Visionware materials -- we did this a long time ago). The IR was completely confirmed by X-ray diffraction of the same samples. Each crystalline material exhibited a different pattern for the OH bands.

Some comments to previous posts:

- The crystalline equivalent to vitreous silica is not quartz but tridymite (see e.g. Article Determination of the crystallographic orientation of oriente...

- glass is not necessarily isotropic (same reference as above, cf. also polymers, even if it usually is. Furthermore, a polycrystalline material between crossed polarizers can appear black if the crystallites are randomly oriented and all are small compared to the wavelength: Article Modelling IR-spectra of single-phase polycrystalline materia...

- even if you have a glass which has the same composition as a crystalline phase it does not necessarily contain only the same structural units as in the crystal, see e.g. Article Interpretation and modeling of IR-reflectance spectra of gla...

Many of the questions I find on this site lack enough detail to allow for specific answers. The answers are often very different for inorganic vs organic glasses, for example. In the current case, the difference between silicate, borate, phosphate, and germinate glasses (all "oxide" glasses) can be significantly different. The answer for vitreous silica (which usually crystallizes to form either quartz or cristobalite -- I've done both) can be very different than that for soda-lime-silicate or aluminosilicate glasses. X-ray powder diffraction will, of course, differentiate glass from crystal (unless the crystals are very, very small), but many of the phases in non-silicate systems do not have known patterns.

As an example, in another question I saw today, the question specifically referred to vitreous silica vs. a thin slice of "Crystal", which I assumed meant single crystal z-cut quartz. If that is indeed the case, the answer is each. If the "slice" was polycrystalline, the answer is a little more difficult. Most of the answers were very generic and required far more effort than if the samples were just vitreous silica and single crystal quartz.

I just with the questions could be more detailed and specific about what the questioner wants to know.

Well i am not experienced with the inorganic molecules however in case of organic molecules i can provide you some references if requested.

Yes it is possible.

Please refer to Regev answer.

Regards,

Ghassan Saffarini