There a variety of glasses that can transmit in the IR spectrum. The choice of the most appropriate material is a matter of compromise. The various parameters to be taken into account are:

1- Maximum wavelength to be transmitted (here, presumably 25 µm)

2- Acceptable thickness of substrate (< 1mm, 1 to 4 mm, > 1cm)

3- Durability in ambient atmosphere

4- Transparency in the visible spectrum

5- Minimum Tg (glass transition temperature)

Fluoride glasses (especially fluoroindates) can be used between 200 nm and 8 to 9 µm

Sulphide glasses can be used up to 10 µm (e.g. As2S3, or GLS (gallium lanthanum sulfide)

Heavier chalcogenides (selenides, tellurides) may be used for transmission at longer wavelength. But they are not transprent in the visible spectrum.

A special case refers to glasses made from tellurium halides (Te-Cl-Br). Black in the visible, they trasmit below 400 cm-1. But they Tg is close to room temperature. 

Chloride and bromide glasses are transparent far beyond 10 µm. Unfortunately there are hygroscopic and not stable in current atmosphere.

Silica-based glass is not suited for IR optics. Please have a look at the below links:

http://www.edmundoptics.de/technical-resources-center/optics/the-correct-material-for-infrared-applications/

http://www.photonics.com/EDU/Handbook.aspx?AID=25495

http://www.newport.com/Optical-Materials/144943/1033/content.aspx

All of silica-based glasses have a absorption at 1000 - 1200 cm-1. ZnSe is good for IR spectroscopy (20 000-50 cm-1), NaCl (50 000-625 cm-1) and Si too (6667- ). 

Mr. Kollath,

Please pay attention to discussion: 'Can I test the FTIR spectrum of a thin film deposited on glass/quartz and take the glass/quartz as a background instead of using KBr?' (http://www.researchgate.net/post/Can_I_test_the_FTIR_spectrum_of_a_thin_film_deposited_on_glass_quartz_and_take_the_glass_quartz_as_a_background_instead_of_using_KBr)

Thank you all for your inputs. I will initially go for Si wafer with SiO2 on both sides. However, suggestion for a transparent substrate (economical) is much appreciated.

If it needs to be a glass you can use chalcogenide glasses. Due to the higher atom masses their vibrational modes are redshifted compared to silica glasses. They are commercially available, see e.g. http://www.schott.com/advanced_optics/english/products/optical-materials/ir-materials/infrared-chalcogenide-glasses/index.html

There a variety of glasses that can transmit in the IR spectrum. The choice of the most appropriate material is a matter of compromise. The various parameters to be taken into account are:

1- Maximum wavelength to be transmitted (here, presumably 25 µm)

2- Acceptable thickness of substrate (< 1mm, 1 to 4 mm, > 1cm)

3- Durability in ambient atmosphere

4- Transparency in the visible spectrum

5- Minimum Tg (glass transition temperature)

Fluoride glasses (especially fluoroindates) can be used between 200 nm and 8 to 9 µm

Sulphide glasses can be used up to 10 µm (e.g. As2S3, or GLS (gallium lanthanum sulfide)

Heavier chalcogenides (selenides, tellurides) may be used for transmission at longer wavelength. But they are not transprent in the visible spectrum.

A special case refers to glasses made from tellurium halides (Te-Cl-Br). Black in the visible, they trasmit below 400 cm-1. But they Tg is close to room temperature. 

Chloride and bromide glasses are transparent far beyond 10 µm. Unfortunately there are hygroscopic and not stable in current atmosphere.

Marcel's answer covers the topic. He is an expert on this subject. Listen to him.