How to Interpret POSS Spectra: 29Si NMR, FTIR, and MS Guide

An intact, fully substituted T₈ silsesquioxane cage leaves a characteristic fingerprint in every standard spectroscopic method. In ²⁹Si NMR it gives a single sharp T³ resonance, in ¹H and ¹³C NMR a single set of substituent signals with integrals matching eight identical arms, in FTIR a dominant asymmetric Si–O–Si stretching band near 1100 cm⁻¹, and in high resolution mass spectrometry a molecular ion whose isotope envelope matches the simulated pattern of an Si₈ species. This guide follows the chemistry of octa(3-aminopropyl)silsesquioxane (OAS-POSS), one of the most versatile POSS building blocks. It begins with the ammonium salts of OAS-POSS, continues with the amides obtained by acylation, and then turns to the imines obtained by Schiff base condensation. For each family it explains which signals change, which stay the same, and how to recognize an incomplete reaction, an opened cage, or a mixture of cage sizes, using complete spectra from published work.

Contents

1. Silicon environments in POSS
2. Reading POSS spectra: the general toolkit
3. OAS-POSS salts: the starting point
4. From amines to amides
5. From amines to imines
6. Comparing ammonium, amide, and imine POSS
7. Common pitfalls in interpreting POSS spectra
8. Quick reference tables
9. Frequently asked questions
10. Sources and related articles

Silicon Environments in POSS: T Units and the T₈ Cage

Polyhedral oligomeric silsesquioxanes are built from T units, silicon atoms that carry one organic substituent and three oxygen atoms. A superscript gives the number of oxygen atoms that bridge to another silicon atom. In a closed cage every silicon atom is a T³ unit, bonded to one carbon atom and to three siloxane oxygens. A T² unit carries only two bridging oxygens and one terminal OH or OR group, which is the hallmark of incompletely condensed species such as open cages, trisilanols, and resin fragments. The full notation, including M, D, and Q units, is explained in the article on silsesquioxane nomenclature, and the geometries of the common cage sizes are described in the overview of polyhedral silsesquioxane structures.

The practical consequence is simple. In a perfectly symmetric octafunctional T₈ cage all eight silicon atoms are chemically equivalent T³ units, and all eight organic arms are equivalent as well. Spectroscopic evidence of symmetry is therefore the most direct evidence of a clean product, and every deviation from that symmetry, whether an extra ²⁹Si resonance, a second set of methylene signals, or an unexpected mass envelope, deserves an explanation before the compound is used further.

Reading POSS Spectra: The General Toolkit

Each spectroscopic method answers a different question about a POSS derivative. ²⁹Si NMR and mass spectrometry report on the cage, ¹H and ¹³C NMR report on the organic arms and on the degree of substitution, and infrared spectroscopy provides a quick check of both the framework and the functional groups. The following sections summarize what each method can and cannot show, before the same principles are applied to the individual compound families.

²⁹Si NMR Spectroscopy: Is the Cage Intact?

²⁹Si NMR is the most decisive single measurement for a new POSS derivative, because it reports directly on the silicon framework rather than on the periphery. A single narrow line establishes that all silicon atoms are equivalent T³ units of an intact cage. The absence of any additional resonance at lower field is equally informative, since T² silicon atoms bearing silanol groups resonate several ppm downfield of the corresponding T³ signal.

The ²⁹Si chemical shift depends mainly on the group bonded directly to silicon. Cages with aryl, vinyl, or hydrogen substituents resonate considerably further upfield than cages with alkyl substituents, and larger T₁₀ and T₁₂ cages resonate a few ppm upfield of the corresponding T₈ cage. Table 1 lists approximate literature values that are useful for orientation. Because shifts depend on solvent, concentration, and referencing, a new compound should always be compared with a reference measured under the same conditions.

Table 1. Approximate ²⁹Si chemical shift ranges of common silsesquioxane environments (general literature values given for orientation).

Silicon environment	Typical example	Approximate δ(29Si) / ppm
T3, alkyl T8 cage	octaisobutyl or octa(3-aminopropyl) POSS	−66 to −68
T3, phenyl T8 cage	octaphenyl POSS	about −78
T3, vinyl T8 cage	octavinyl POSS	about −80
T3, hydrido T8 cage	octahydridosilsesquioxane	about −84
T2, silanol bearing unit	trisilanol heptaisobutyl POSS	about −58
glass of the NMR tube and probe	background, not the sample	broad hump centered near −110

The ²⁹Si nucleus has a low natural abundance of about 4.7 percent, a negative gyromagnetic ratio, and long longitudinal relaxation times in rigid cage compounds. Proton decoupling should therefore be applied with an inverse gated sequence, which suppresses the negative nuclear Overhauser effect that can otherwise attenuate or even null the signal. A small amount of a paramagnetic relaxation agent such as chromium(III) acetylacetonate shortens the relaxation delay considerably and is recommended whenever relative intensities of several silicon environments must be compared. A broad background hump from the borosilicate glass of the tube and from probe components commonly appears near −110 ppm and should not be mistaken for Q units of the sample. The spectra on this page were referenced either to tetramethylsilane at 0 ppm or to the water soluble standard DSS (4,4-dimethyl-4-silapentane-1-sulfonic acid) at 1.31 to 1.32 ppm.

¹H and ¹³C NMR: Have All Eight Arms Reacted?

For derivatives of OAS-POSS the 3-aminopropyl arm acts as a built in reporter. Its three methylene groups respond differently to chemistry at the nitrogen atom. The SiCH₂ group next to the cage barely changes and confirms that the arm is still attached to silicon, the central CH₂ group shifts slightly, and the NCH₂ group reports directly on the nature of the terminal group. Comparing these three signals before and after a reaction is therefore the fastest way to follow the conversion.

Integration provides the second check. In a fully substituted cage every methylene group integrates to sixteen protons, and a signal unique to the new group, such as an amide NH or an imine CH=N proton, integrates to eight. A single set of arm signals with these integrals shows that all eight arms are equivalent on the NMR time scale, while a second, smaller set of methylene signals points to arms that did not react.

FTIR Spectroscopy: Framework and Functional Groups

In infrared spectra the silsesquioxane framework is represented by the asymmetric Si–O–Si stretching vibration, which is usually the most intense band of the spectrum and appears near 1100 cm⁻¹. The terminal groups add their own characteristic bands, discussed separately for each family below. FTIR alone cannot prove that a cage is intact. According to the classical work of Brown, Vogt, and Prescott, cage-like phenylsilsesquioxanes show a single strong Si–O–Si band near 1120 to 1130 cm⁻¹, whereas ladder polymers show two bands near 1135 to 1150 and 1045 to 1060 cm⁻¹, as discussed on the introduction to silsesquioxanes page and in the article on phenylsilsesquioxane equilibration from cage to ladder. Band positions and shapes, however, are also influenced by the substituents and by cage distortion in the solid state, and infrared data should always be interpreted together with ²⁹Si NMR and mass spectrometry.

Mass Spectrometry: Reading the Isotope Pattern

High resolution electrospray ionization mass spectrometry (ESI-MS) provides the molecular formula and at the same time excludes cage sizes other than T₈. A molecule with eight silicon atoms does not give a single molecular ion peak but an isotope envelope, because the heavier isotopes ²⁹Si (about 4.7 percent), ³⁰Si (about 3.1 percent), and ¹³C (about 1.1 percent) are statistically present in a large fraction of molecules. The larger the molecule, the broader the envelope, and for molecules with many carbon atoms the most intense peak lies one or more mass units above the monoisotopic ion. Heteroatoms such as bromine and sulfur reshape the envelope further. The decisive criterion is therefore not a single m/z value but the agreement between the measured envelope and the envelope simulated for the proposed formula, in both peak spacing and relative intensities. A T₁₀ or T₁₂ analogue, or a partially substituted cage, would give a separate envelope at a clearly different mass.

OAS-POSS Salts: The Starting Point

Octa(3-aminopropyl)silsesquioxane is usually prepared and stored as an ammonium salt rather than as the free amine. The salts are crystalline, stable, and, as discussed in the post on base choice and cage integrity in amide POSS synthesis, resist the cage opening that affects the free octaamine in the presence of water. Synthetic routes to OAS-POSS are described in the article on octa(3-aminopropyl)silsesquioxane. This section presents the hydrochloride and the triflate, two salts with the same cation but different counterions.

NMR Signatures of the Ammonium Salts

In DMSO-d₆ the eight ammonium groups give a single broad signal integrating to 24 protons, and the NCH₂, central CH₂, and SiCH₂ groups give signals of 16 protons each (Table 2). The position of the NH₃⁺ signal depends on the counterion. For the hydrochloride it appears at 8.23 to 8.28 ppm in two independent measurements, whereas for the triflate it appears at 7.57 ppm, a difference that reflects the stronger hydrogen bonding of chloride compared with the weakly coordinating triflate anion. The NCH₂ protons resonate at 2.75 to 2.79 ppm and the SiCH₂ protons at 0.66 to 0.75 ppm. The ²⁹Si spectra of both salts show a single resonance at −66.5 ppm, as expected for an intact T₈ cage.

Table 2. NMR signals of the OAS-POSS salts in DMSO-d₆ (chemical shifts in ppm, ¹H at 500 MHz, ¹³C at 126 MHz, ²⁹Si at 59.6 MHz). ^a See Table 3. ^b Referenced to DSS.

Compound (source)	NH3+, δH	NCH2, δH	CH2CH2CH2, δH	SiCH2, δH	NCH2, CH2, SiCH2, δC	δSi
OAS-POSS · HCl (2025 study)	8.28	2.76 to 2.79	1.70 to 1.77	0.72 to 0.75	41.0, 20.6, 8.4	−66.5
OAS-POSS · HCl (2014 study)	8.23	2.79	1.74	0.75	recorded in D2Oa	−66.52
OAS-POSS · CF3SO3H (2014 study)	7.57	2.75	1.56	0.66	recorded in D2Oa	−66.53b

The ¹³C spectra of the salts from the 2014 study were recorded in D₂O (Table 3). In the triflate the anion itself is visible as a quartet centered near 122 ppm with a ¹J_CF coupling constant of about 316 Hz, which provides a direct spectroscopic proof of the counterion. The propyl carbon signals of the two salts differ by about 2.6 ppm, which illustrates how strongly carbon shifts measured in D₂O depend on the referencing method, since no internal TMS can be used in this solvent.

Table 3. ¹³C NMR signals (126 MHz) of the OAS-POSS salts in D₂O from the 2014 study (chemical shifts in ppm).

Compound	NCH2, δC	CH2CH2CH2, δC	SiCH2, δC	CF3SO3−, δC
OAS-POSS · HCl	41.73	20.62	8.67	absent
OAS-POSS · CF3SO3H	44.35	23.26	11.28	about 122 (q, 1JCF about 316 Hz)

Infrared and Mass Spectra of the Ammonium Salts

The infrared spectra of both salts, recorded in KBr pellets, show the dominant Si–O–Si band near 1100 cm⁻¹ together with a very broad N–H stretching absorption of the ammonium groups between about 3200 and 2800 cm⁻¹. The triflate additionally shows the strong bands of its anion, assigned in the figure to CF₃, SO₃, and C–S vibrations near 1250, 1030, and 640 cm⁻¹. These anion bands partly overlap the Si–O–Si band and must not be confused with it.

Ammonium salts do not appear with their counterions in positive ion ESI. Both salts give exactly the same ion at m/z 881.29, which corresponds to the protonated free octaamine [M−8HX+H]⁺ with the formula C₂₄H₆₄N₈O₁₂Si₈. Mass spectrometry therefore confirms the cage but cannot identify the salt, which must be established by NMR, IR, or elemental analysis. Because this ion contains only 24 carbon atoms, its monoisotopic peak is also the most intense peak of the envelope.

Spectra of OAS-POSS · HCl

The NMR spectra shown here were recorded in the 2025 study and the infrared and mass spectra in the 2014 study. Water and residual DMSO appear in the ¹H spectrum at 3.38 and 2.50 ppm.

Figure 1. Structure of OAS-POSS · HCl, [(H₃NCH₂CH₂CH₂)₈Si₈O₁₂]Cl₈.

Figure 2. ¹H NMR spectrum (500 MHz) of OAS-POSS · HCl in DMSO-d₆.

Figure 3. ¹³C NMR spectrum (126 MHz) of OAS-POSS · HCl in DMSO-d₆.

Figure 4. ²⁹Si NMR spectrum (59.6 MHz) of OAS-POSS · HCl in DMSO-d₆.

Figure 5. FTIR (KBr pellet) spectrum of OAS-POSS · HCl. The broad N–H stretching absorption of the ammonium groups and the strong Si–O–Si band are labeled in the figure.

Figure 6. Measured (top) and simulated (bottom) ESI-MS isotope pattern of OAS-POSS · HCl. The ion at m/z 881.29 is the protonated free octaamine [M−8HCl+H]⁺.

Table. Assignment of the NMR signals of OAS-POSS · HCl (DMSO-d₆).

Group	δH / ppm	δC / ppm
NH3+	8.28 (24H)	absent
NCH2	2.76 to 2.79 (16H)	41.0
CH2CH2CH2	1.70 to 1.77 (16H)	20.6
SiCH2	0.72 to 0.75 (16H)	8.4
29Si, T3	δSi −66.5

Spectra of OAS-POSS · CF₃SO₃H

All spectra of the triflate were recorded in the 2014 study. The ¹³C spectrum in D₂O includes an enlarged view of the CF₃ quartet of the anion, and the ²⁹Si spectrum was referenced to DSS.

Figure 7. Structure of OAS-POSS · CF₃SO₃H, [(H₃NCH₂CH₂CH₂)₈Si₈O₁₂](CF₃SO₃)₈.

Figure 8. ¹H NMR spectrum (500 MHz) of OAS-POSS · CF₃SO₃H in DMSO-d₆.

Figure 9. ¹³C NMR spectrum (126 MHz) of OAS-POSS · CF₃SO₃H in D₂O.

Figure 10. ²⁹Si NMR spectrum (59.6 MHz) of OAS-POSS · CF₃SO₃H in DMSO-d₆, referenced to DSS.

Figure 11. FTIR (KBr pellet) spectrum of OAS-POSS · CF₃SO₃H. Bands of the triflate anion (CF₃, SO₃, C–S) appear next to the Si–O–Si band and are labeled in the figure.

Figure 12. Measured (top) and simulated (bottom) ESI-MS isotope pattern of OAS-POSS · CF₃SO₃H. The same ion at m/z 881.29 as for the hydrochloride is observed, [M−8CF₃SO₃H+H]⁺.

Table. Assignment of the NMR signals of OAS-POSS · CF₃SO₃H (DMSO-d₆ for ¹H, D₂O for ¹³C).

Group	δH / ppm	δC / ppm
NH3+	7.57 (24H)	absent
NCH2	2.75 (16H)	44.35 (D2O)
CH2CH2CH2	1.56 (16H)	23.26 (D2O)
SiCH2	0.66 (16H)	11.28 (D2O)
CF3SO3−		about 122, quartet (D2O)
29Si, T3	δSi −66.53

From Amines to Amides

Acylation of OAS-POSS salts with acyl chlorides converts all eight ammonium groups into secondary amides in high yield while leaving the cage intact. The method and its scope are described in the post on high yield amide POSS synthesis from OAS salts. This section uses two representative examples, an aromatic amide obtained with 4-fluorobenzoyl chloride and an aliphatic amide obtained with hexanoyl chloride, to show how amide formation appears in each spectroscopic method.

What Changes in the NMR Spectra

Amide formation moves the NCH₂ protons from 2.75 to 2.79 ppm in the ammonium salts to 3.15 to 3.19 ppm, but leaves the NCH₂ carbon near 41.7 to 42.9 ppm, close to its position in the salts (Table 4). The ¹³C shift of this carbon therefore does not distinguish an amide from the starting ammonium salt, and the proof of conversion rests on the ¹H spectrum and on the carbonyl carbon, which appears at 165.2 ppm for the aromatic amide and at 176.2 ppm for the aliphatic amide. The amide NH proton is a useful but solvent dependent marker. In DMSO-d₆ it appears as a broad signal at 8.44 ppm integrating to eight protons, whereas in methanol-d₄ it is not observed at all, because it exchanges with the deuterium of the solvent. Fluorine substituents split the carbon signals of the ring they are attached to, as the doublet of the C–F carbon of the 4-fluorobenzamide at 163.7 ppm with ¹J_CF of about 249 Hz illustrates. The ²⁹Si spectra of both amides show a single resonance at −66.1 and about −66.4 ppm, which confirms that acylation leaves the T₈ cage intact.

Table 4. NMR signals of the amide group and the propyl spacer of octa-amide POSS (chemical shifts in ppm, ¹H at 500 MHz, ¹³C at 126 MHz). ^a In the hexanamide this signal overlaps with the methylene signals of the hexanoyl chain.

Compound	Solvent	NH, δH	C=O, δC	NCH2, δH	NCH2, δC	CH2CH2CH2, δH	CH2CH2CH2, δC	SiCH2, δH	SiCH2, δC
4-fluorobenzamide POSS	DMSO-d6	8.44	165.17	3.19	41.70	1.58	22.39	0.64	8.88
hexanamide POSS	methanol-d4	exchanged	176.24	3.15	42.92	1.60 to 1.62a	within 23.5 to 32.6a	0.64	10.69

Infrared and Mass Spectra of Amide POSS

In the infrared spectra the broad ammonium absorption of the starting salt is replaced by the N–H stretching band of the amide near 3300 cm⁻¹, and two new strong bands appear, the amide I (C=O stretching) band near 1640 cm⁻¹ and the amide II band that follows it at lower wavenumbers. The Si–O–Si band remains the strongest band of the spectrum. In ESI-MS the amides give protonated molecular ions [M+H]⁺, and because they contain 72 to 80 carbon atoms the maximum of the envelope lies one mass unit above the monoisotopic ion.

Spectra of Octa(4-fluorobenzamide) POSS

In DMSO-d₆ the amide NH proton of this aromatic amide is clearly visible, the two sets of aromatic protons of the para substituted ring appear as separate signals, and fluorine couples to the ring carbons, most strongly to the C–F carbon itself.

Figure 13. Structure of 4-fluorobenzamide POSS, C₈₀H₈₈F₈N₈O₂₀Si₈.

Figure 14. ¹H NMR spectrum (500 MHz) of 4-fluorobenzamide POSS in DMSO-d₆.

Figure 15. ¹³C NMR spectrum (126 MHz) of 4-fluorobenzamide POSS in DMSO-d₆.

Figure 16. ²⁹Si NMR spectrum (59.6 MHz) of 4-fluorobenzamide POSS in DMSO-d₆, referenced to DSS and TMS.

Figure 17. FTIR (KBr pellet) spectrum of 4-fluorobenzamide POSS. The N–H, amide I (C=O), amide II, C–F, and Si–O–Si bands are labeled in the figure.

Figure 18. Measured (top) and simulated (bottom) ESI-MS isotope pattern of 4-fluorobenzamide POSS. The [M+H]⁺ envelope begins at m/z 1857.4 and reaches its maximum one mass unit higher.

Table. Assignment of the NMR signals of 4-fluorobenzamide POSS (DMSO-d₆).

Group	δH / ppm	δC / ppm
NH	8.44 (8H)
C=O		165.17
aromatic C–F		163.7 (d, 1JCF about 249 Hz)
aromatic CH ortho to C=O	7.85 (16H)	129.72
aromatic CH ortho to F	7.20 (16H)	114.92
aromatic C ipso to C=O		131.03
NCH2	3.19 (16H)	41.70
CH2CH2CH2	1.58 (16H)	22.39
SiCH2	0.64 (16H)	8.88
29Si, T3	δSi −66.12

Spectra of Octa(hexanamide) POSS

The NMR spectra of this aliphatic amide were recorded in methanol-d₄, in which the amide NH proton exchanges with deuterium and is therefore absent. The long alkyl chains produce overlapping methylene signals that are assigned with the help of two dimensional spectra.

Figure 19. Structure of hexanamide POSS, C₇₂H₁₄₄N₈O₂₀Si₈.

Figure 20. ¹H NMR spectrum (500 MHz) of hexanamide POSS in methanol-d₄.

Figure 21. ¹³C NMR spectrum (126 MHz) of hexanamide POSS in methanol-d₄.

Figure 22. ²⁹Si NMR spectrum (59.6 MHz) of hexanamide POSS in DMSO-d₆, referenced to DSS.

Figure 23. FTIR (KBr pellet) spectrum of hexanamide POSS. The N–H, C–H, amide I (C=O), amide II, and Si–O–Si bands are labeled in the figure.

Figure 24. Measured (top) and simulated (bottom) ESI-MS isotope pattern of hexanamide POSS. The [M+H]⁺ envelope begins at m/z 1665.87 and reaches its maximum at m/z 1666.87.

Table. Assignment of the NMR signals of hexanamide POSS (methanol-d₄ for ¹H and ¹³C).

Group	δH / ppm	δC / ppm
NH	exchanged
C=O		176.24
NCH2	3.15 (16H)	42.92
COCH2	2.19 (16H)	37.21
CH2 of propyl and hexanoyl chains	1.60 to 1.62 (32H), 1.31 to 1.36 (32H)	32.61, 26.89, 24.13, 23.52
CH3	0.92 (24H)	14.45
SiCH2	0.64 (16H)	10.69
29Si, T3	δSi about −66.4

From Amines to Imines

Condensation of OAS-POSS salts with aromatic aldehydes in the presence of a base gives octa-imine POSS, also called Schiff base POSS. The synthesis, crystal structures, and thermal properties of the examples shown here are described in the article on octa-imine POSS crystal structures and thermal properties, and related compounds in the post on imine POSS crystal structures and supramolecular networks. Imine POSS also serve as precursors of porous silsesquioxane imine frameworks and of zinc imine POSS catalysts.

What Changes in the NMR Spectra

Imine formation changes the propyl arm more strongly than amide formation. The NCH₂ protons move to 3.54 to 3.73 ppm and the NCH₂ carbon to 61.7 to 64.7 ppm, about 20 ppm downfield of its position in the salts and amides, because the nitrogen atom becomes part of a C=N double bond (Table 5). The SiCH₂ group again stays almost unchanged, at 0.65 to 0.83 ppm and 9.6 to 9.9 ppm. A complete shift of the NCH₂ resonance, with no residual signal near 2.7 ppm, is the clearest proof that no unreacted aminopropyl arms remain.

The imine group itself gives a sharp singlet between 8.16 and 8.83 ppm in ¹H NMR and a carbon signal between 154.4 and 165.1 ppm, with the exact position depending on the electronic character of the aromatic substituent. Its integral provides an internal check of the degree of substitution. In a fully substituted cage the CH=N signal integrates to eight protons against sixteen protons for each methylene group, a ratio of 1:2:2:2. Substituents can add their own diagnostic features. When the aromatic ring carries a hydroxyl group next to the imine, as in POSS-6, the OH proton appears far downfield at 13.60 ppm, a position typical of a hydroxyl group engaged in a strong O–H···N hydrogen bond. The ²⁹Si spectra show a single resonance between −66.5 and −66.74 ppm, which confirms that the condensation leaves the cage intact.

Table 5. ¹H (500 MHz) and ¹³C (126 MHz) NMR signals of the imine group and the propyl spacer of octa-imine POSS in CDCl₃ (chemical shifts in ppm). ^a The imine carbon of POSS-9 lies within a group of three signals at 155.0, 155.5, and 156.3 ppm that also contains the two aromatic C–O carbons.

Compound	CH=N, δH	CH=N, δC	NCH2, δH	NCH2, δC	CH2CH2CH2, δH	CH2CH2CH2, δC	SiCH2, δH	SiCH2, δC
POSS-6	8.20	165.1	3.55	61.7	1.76 to 1.82	24.6	0.71 to 0.75	9.6
POSS-7	8.16	159.5	3.58	64.0	1.78 to 1.84	24.4	0.67 to 0.71	9.8
POSS-8	8.33	154.4	3.54	63.8	1.76 to 1.82	24.3	0.65	9.8
POSS-9	8.83	155.0 to 156.3a	3.73	64.7	1.96	24.6	0.83	9.9

Infrared and Mass Spectra of Imine POSS

In the diffuse reflectance (DRIFT) spectra of the imines, recorded in Kubelka–Munk units, the C=N stretching band appears between 1633 and 1647 cm⁻¹ and the Si–O–Si band between 1108 and 1131 cm⁻¹. Complete conversion of the aldehyde is indicated by the absence of the carbonyl stretching band, which typically lies near 1700 cm⁻¹ for aromatic aldehydes. POSS-6 shows a second strong band at 1035 cm⁻¹ next to its main band at 1131 cm⁻¹, although its ²⁹Si NMR spectrum and mass spectrum unambiguously confirm the intact T₈ cage, a reminder that a split band in this region is not by itself evidence of a ladder structure. Hydrogen bonding also leaves a clear trace in the spectrum of POSS-6. Instead of a sharp O–H stretching band above 3200 cm⁻¹, the spectrum shows a very broad absorption extending from about 3000 to 2500 cm⁻¹ with a shoulder at 2696 cm⁻¹, in agreement with the downfield OH signal in the ¹H NMR spectrum and with the hydrogen bonds observed in the crystal structure of POSS-6.

The imines were detected by ESI-MS as protonated molecular ions [M+H]⁺. With 64 to 128 carbon atoms their envelopes are broad, and the maximum lies one or two mass units above the monoisotopic ion. Heteroatoms reshape the envelope in characteristic ways. The eight bromine atoms of POSS-7 produce a very wide envelope spanning roughly 16 mass units, with its maximum near m/z 2218.8, because ⁷⁹Br and ⁸¹Br occur in almost equal amounts, and the eight sulfur atoms of POSS-8 enhance the M+2 contribution through ³⁴S.

Spectra of POSS-6, an Octa(hydroxyphenyl imine) POSS

POSS-6 carries a dimethyl substituted 2-hydroxyphenyl group on each imine. Its spectra illustrate how an intramolecular O–H···N hydrogen bond appears in NMR and IR spectra, with the OH proton at 13.60 ppm and a very broad OH stretching absorption below 3000 cm⁻¹.

Figure 25. Structure of POSS-6, C₉₆H₁₂₈N₈O₂₀Si₈.

Figure 26. ¹H NMR spectrum (500 MHz) of POSS-6 in CDCl₃.

Figure 27. ¹³C NMR spectrum (126 MHz) of POSS-6 in CDCl₃.

Figure 28. ²⁹Si NMR spectrum (59.6 MHz) of POSS-6 in CDCl₃.

Figure 29. DRIFT spectrum of POSS-6. Kubelka–Munk units. Main bands at 1633 cm⁻¹ (C=N), 1131 and 1035 cm⁻¹ (Si–O–Si region), and a broad hydrogen bonded OH absorption with a shoulder at 2696 cm⁻¹.

Figure 30. Measured (top) and simulated (bottom) ESI-MS isotope pattern of POSS-6. The envelope begins at m/z 1937.75 and reaches its maximum at m/z 1938.76.

Table. Assignment of the NMR signals of POSS-6 (CDCl₃).

Group	δH / ppm	δC / ppm
OH	13.60 (8H)
CH=N	8.20 (8H)	165.1
aromatic C–OH		157.5
aromatic CH and C	6.97 (8H), 6.81 (8H)	134.1, 128.9, 126.9, 125.7, 117.8
NCH2	3.54 to 3.56 (16H)	61.7
aromatic CH3	2.22, 2.23 (48H)	20.4, 15.6
CH2CH2CH2	1.76 to 1.82 (16H)	24.6
SiCH2	0.71 to 0.75 (16H)	9.6
29Si, T3	δSi −66.66

Spectra of POSS-7, an Octa(3-bromophenyl imine) POSS

POSS-7 carries a 3-bromophenyl group on each imine. Its aromatic region shows the four distinct protons of a meta substituted ring, and its mass spectrum shows the characteristic isotope pattern of a molecule containing eight bromine atoms.

Figure 31. Structure of POSS-7, C₈₀H₈₈Br₈N₈O₁₂Si₈.

Figure 32. ¹H NMR spectrum (500 MHz) of POSS-7 in CDCl₃.

Figure 33. ¹³C NMR spectrum (126 MHz) of POSS-7 in CDCl₃.

Figure 34. ²⁹Si NMR spectrum (59.6 MHz) of POSS-7 in CDCl₃.

Figure 35. DRIFT spectrum of POSS-7. Kubelka–Munk units. Main bands at 1647 cm⁻¹ (C=N), 1565 cm⁻¹ (aromatic ring), and 1116 cm⁻¹ (Si–O–Si).

Figure 36. Measured (top) and simulated (bottom) ESI-MS isotope pattern of POSS-7. The eight bromine atoms produce an envelope of about 16 mass units with its maximum near m/z 2218.8.

Table. Assignment of the NMR signals of POSS-7 (CDCl₃).

Group	δH / ppm	δC / ppm
CH=N	8.16	159.5
aromatic H2	7.86
aromatic H4, H6	7.49 to 7.57
aromatic H5	7.20 to 7.23
aromatic C1		138.4
aromatic CH		133.5, 130.7, 130.2, 126.9
aromatic C–Br		123.0
NCH2	3.57 to 3.60	64.0
CH2CH2CH2	1.78 to 1.84	24.4
SiCH2	0.67 to 0.71	9.8
29Si, T3	δSi −66.6

Spectra of POSS-8, an Octa(thiophene imine) POSS

POSS-8 carries a thiophene ring on each imine. The three thiophene protons appear as well separated signals between 7.0 and 7.4 ppm, and the imine carbon resonates at 154.4 ppm.

Figure 37. Structure of POSS-8, C₆₄H₈₀N₈O₁₂S₈Si₈.

Figure 38. ¹H NMR spectrum (500 MHz) of POSS-8 in CDCl₃.

Figure 39. ¹³C NMR spectrum (126 MHz) of POSS-8 in CDCl₃.

Figure 40. ²⁹Si NMR spectrum (59.6 MHz) of POSS-8 in CDCl₃.

Figure 41. DRIFT spectrum of POSS-8. Kubelka–Munk units. Main bands at 1635 cm⁻¹ (C=N) and 1108 cm⁻¹ (Si–O–Si).

Figure 42. Measured (top) and simulated (bottom) ESI-MS isotope pattern of POSS-8. The envelope begins at m/z 1633 and reaches its maximum two mass units higher, reflecting the combined contribution of silicon, carbon, and sulfur isotopes.

Table. Assignment of the NMR signals of POSS-8 (CDCl₃).

Group	δH / ppm	δC / ppm
CH=N	8.33 (8H)	154.4
thiophene CH	7.39 (8H), 7.25 (8H), 7.02 (8H)	130.3, 128.6, 127.4
thiophene C2		142.8
NCH2	3.54 (16H)	63.8
CH2CH2CH2	1.76 to 1.82 (16H)	24.3
SiCH2	0.65 (16H)	9.8
29Si, T3	δSi −66.74

Spectra of POSS-9, an Octa(dibenzofuran imine) POSS

POSS-9 carries a dibenzofuran ring system on each imine. Its imine proton resonates at 8.83 ppm, and its aromatic region contains seven distinct proton environments per arm.

Figure 43. Structure of POSS-9, C₁₂₈H₁₁₂N₈O₂₀Si₈.

Figure 44. ¹H NMR spectrum (500 MHz) of POSS-9 in CDCl₃.

Figure 45. ¹³C NMR spectrum (126 MHz) of POSS-9 in CDCl₃.

Figure 46. ²⁹Si NMR spectrum (59.6 MHz) of POSS-9 in CDCl₃.

Figure 47. DRIFT spectrum of POSS-9. Kubelka–Munk units. Main bands at 1647 cm⁻¹ (C=N), 1188 cm⁻¹, and 1111 cm⁻¹ (Si–O–Si), with aromatic C–H stretching at 3057 cm⁻¹.

Figure 48. Measured (top) and simulated (bottom) ESI-MS isotope pattern of POSS-9. The envelope begins at m/z 2305.7 and reaches its maximum near m/z 2307.7.

Table. Assignment of the NMR signals of POSS-9 (CDCl₃).

Group	δH / ppm	δC / ppm
CH=N	8.83 (8H)	within 155.0 to 156.3
aromatic C–O		within 155.0 to 156.3
aromatic CH and C	7.81 to 7.89 (24H), 7.50, 7.39, 7.29, 7.20 (8H each)	127.4 to 120.7, 111.9
NCH2	3.73 (16H)	64.7
CH2CH2CH2	1.96 (16H)	24.6
SiCH2	0.83 (16H)	9.9
29Si, T3	δSi −66.5

Comparing Ammonium, Amide, and Imine POSS

The ²⁹Si Shift Reports on the Cage, Not on the Terminal Group

Placed side by side, the ²⁹Si data of the three families make one point very clearly (Table 6). Although the terminal groups range from ammonium salts through amides to aromatic imines, the ²⁹Si resonances stay close to −66.5 ppm. The silicon atom senses mainly the group bonded directly to it, which in all of these compounds is the same propyl spacer, while the functional group sits at the far end of the three carbon chain. The ²⁹Si shift is therefore a reliable probe of cage integrity, but conversion at the terminal group must always be confirmed by ¹H and ¹³C NMR.

Table 6. ²⁹Si NMR chemical shifts (59.6 MHz) of the ammonium, amide, and imine POSS shown on this page, referenced to TMS or to DSS.

Compound	Terminal group	Solvent	δ(29Si) / ppm
OAS-POSS · HCl	ammonium, chloride salt	DMSO-d6	−66.5
OAS-POSS · CF3SO3H	ammonium, triflate salt	DMSO-d6	−66.53
4-fluorobenzamide POSS	aromatic amide	DMSO-d6	−66.12
hexanamide POSS	aliphatic amide	DMSO-d6	about −66.4
POSS-6	hydroxy(dimethyl)phenyl imine	CDCl3	−66.66
POSS-7	3-bromophenyl imine	CDCl3	−66.6
POSS-8	thiophene imine	CDCl3	−66.74
POSS-9	dibenzofuran imine	CDCl3	−66.5

The NCH₂ Group Reports on the Terminal Group

The NCH₂ group tells the opposite story. Its ¹H signal moves in a characteristic sequence, from 2.75 to 2.79 ppm in the ammonium salts to 3.15 to 3.19 ppm in the amides and 3.54 to 3.73 ppm in the imines, and each step can be followed directly during a reaction. Its ¹³C signal stays near 41 to 43 ppm in salts and amides but moves to about 62 to 65 ppm in the imines, which makes it a decisive marker of imine formation but not of amide formation. Together with the new signals of the terminal group, the amide NH and carbonyl carbon or the imine CH=N proton and carbon, these shifts allow each family to be identified from a single pair of ¹H and ¹³C spectra.

Common Pitfalls in Interpreting POSS Spectra

The most frequent error is to rely on a single technique. A clean ¹H NMR spectrum does not prove that the cage survived, because an opened or rearranged framework can carry the same organic arms, and a single ²⁹Si resonance does not prove that all arms reacted, because the silicon shift is almost insensitive to the terminal group. Cage integrity should be established by ²⁹Si NMR together with mass spectrometry, and the degree of substitution by ¹H and ¹³C NMR.

A second group of errors concerns signals that do not belong to the product. Residual aldehyde gives a formyl proton near 10 ppm and a carbonyl carbon near 190 ppm, unreacted aminopropyl arms leave an NCH₂ signal near 2.7 ppm, and residual CHCl₃ and water appear at 7.26 and about 1.56 ppm in CDCl₃. Very small signals of this kind should be integrated and reported rather than ignored, since their size relative to the main signals gives a direct estimate of the purity of the sample.

A third group concerns the solvent and the counterion. Exchangeable protons such as amide NH and ammonium NH₃⁺ disappear in methanol-d₄ or D₂O and shift with the counterion in DMSO-d₆, so their position or absence says little by itself. Spectra of starting material and product should be compared in the same solvent and with the same reference, particularly in D₂O, where no internal TMS can be used.

A fourth group concerns the stability of the product in the NMR tube. Imine bonds can hydrolyze in the presence of water, particularly in acidic solvents such as aged CDCl₃, which slowly forms traces of HCl. Hydrolysis is recognized by the simultaneous growth of an aldehyde signal and of a new NCH₂ signal of the regenerated amine. Spectra should therefore be recorded promptly in dry solvent, and a spectrum that changes on standing indicates decomposition rather than an impure starting material.

Finally, additional ²⁹Si resonances do not always indicate a defect. Mixtures of T₈, T₁₀, and T₁₂ cages give separate signals, and a T₁₂ cage of D_2d symmetry itself shows two resonances in a 2:1 ratio. Distinguishing these cases requires mass spectrometry, which reveals each cage size as a separate isotope envelope.

Quick Reference Tables

Table 7. Summary of the diagnostic spectroscopic data (chemical shifts in ppm). ²⁹Si spectra of the salts and amides were recorded in DMSO-d₆.

Compound	Solvent (NMR)	δ(29Si)	Key 1H signal	NCH2 δH	NCH2 δC	Key IR bands	ESI-MS envelope maximum, m/z
OAS-POSS · HCl	DMSO-d6	−66.5	NH3+ 8.28	2.77	41.0	broad N–H, Si–O–Si	881.29
OAS-POSS · CF3SO3H	DMSO-d6	−66.53	NH3+ 7.57	2.75	44.35 (D2O)	CF3, SO3, C–S, Si–O–Si	881.29
4-fluorobenzamide POSS	DMSO-d6	−66.12	NH 8.44	3.19	41.70	amide I near 1640 cm−1, Si–O–Si	1858.4
hexanamide POSS	methanol-d4	about −66.4	NH exchanged	3.15	42.92	amide I near 1640 cm−1, Si–O–Si	1666.87
POSS-6	CDCl3	−66.66	CH=N 8.20, OH 13.60	3.55	61.7	C=N 1633, Si–O–Si 1131 cm−1	1938.76
POSS-7	CDCl3	−66.6	CH=N 8.16	3.58	64.0	C=N 1647, Si–O–Si 1116 cm−1	2218.8
POSS-8	CDCl3	−66.74	CH=N 8.33	3.54	63.8	C=N 1635, Si–O–Si 1108 cm−1	1635.16
POSS-9	CDCl3	−66.5	CH=N 8.83	3.73	64.7	C=N 1647, Si–O–Si 1111 cm−1	2307.75

Table 8. Imine and SiCH₂ signals of octa-imine POSS in CDCl₃ (¹H at 500 MHz, ¹³C at 126 MHz). ^a See the note to Table 5.

Compound	CH=N, δH	SiCH2, δH	CH=N, δC	SiCH2, δC
POSS-6	8.20	0.71 to 0.75	165.1	9.6
POSS-7	8.16	0.67 to 0.71	159.5	9.8
POSS-8	8.33	0.65	154.4	9.8
POSS-9	8.83	0.83	155.0 to 156.3a	9.9

Frequently Asked Questions

What is the typical ²⁹Si NMR chemical shift of an alkyl substituted T8 POSS?

Octafunctional T₈ cages with alkyl substituents usually resonate between about −66 and −68 ppm relative to TMS. Cages with 3-aminopropyl, 3-amidopropyl, and 3-iminopropyl arms, such as those shown on this page, give single signals close to −66.5 ppm.

Why does my POSS show more than one ²⁹Si NMR signal?

Additional signals indicate either a lower symmetry or a mixture. Typical causes are incompletely condensed T² silicon atoms bearing OH groups, which appear several ppm downfield, mixtures of T₈, T₁₀, and T₁₂ cages, or inequivalent substituents on a single cage. A T₁₂ cage of D_2d symmetry also gives two signals in a 2:1 ratio.

How can I confirm by NMR that all eight arms of the POSS reacted?

Compare the integral of a signal unique to the new group, for example the amide NH or the imine CH=N proton, with those of the propyl methylene groups. A fully substituted cage gives a ratio of 1:2:2:2, and the NCH₂ signal of the starting ammonium salt, near 2.7 to 2.8 ppm, must be completely absent.

How can I tell which counterion my OAS-POSS salt contains?

Not from positive ion ESI-MS, which shows the same protonated octaamine at m/z 881.29 for different salts. Use ¹H NMR in DMSO-d₆, where the NH₃⁺ signal appears at 8.23 to 8.28 ppm for the hydrochloride and at 7.57 ppm for the triflate, together with anion specific evidence such as the CF₃ quartet near 122 ppm in the ¹³C spectrum or the CF₃ and SO₃ bands in the infrared spectrum of the triflate.

Which infrared band is characteristic of the POSS cage?

The asymmetric Si–O–Si stretching vibration, usually the strongest band of the spectrum, found near 1100 cm⁻¹. Because substituents, counterions, and solid state effects can overlap, split, or shift this band, cage integrity should be confirmed by ²⁹Si NMR and mass spectrometry.

Why is the most intense peak in the mass spectrum of POSS not always the monoisotopic ion?

A molecule with eight silicon atoms and many carbon atoms contains heavy isotopes such as ²⁹Si, ³⁰Si, and ¹³C in a large fraction of its molecules. For larger POSS derivatives the isotope envelope therefore peaks one or more mass units above the monoisotopic ion, and bromine or sulfur atoms broaden it further. For small cages such as the protonated octaamine the monoisotopic peak can still be the most intense.

Do I need a relaxation agent to record ²⁹Si NMR spectra of POSS?

It is not always required, but it is strongly recommended. Silicon nuclei in rigid cages relax slowly, and a small amount of chromium(III) acetylacetonate combined with inverse gated proton decoupling shortens the measurement and gives reliable relative intensities.

Sources and Related Articles

The NMR, infrared, and ESI-MS spectra of OAS-POSS · CF₃SO₃H and of the two amides, as well as the infrared and mass spectra of OAS-POSS · HCl, are reproduced from the Supporting Information of M. Janeta, Ł. John, J. Ejfler, and S. Szafert, Chemistry – A European Journal, 2014, 20, 15966, DOI 10.1002/chem.201404153. The NMR spectra of OAS-POSS · HCl and all spectra of POSS-6 to POSS-9 are reproduced from the Supporting Information of M. Janeta and S. Szafert, Journal of Molecular Structure, 2025, DOI 10.1016/j.molstruc.2025.143517. Crystallographic data for POSS-6 to POSS-9 are deposited with the Cambridge Crystallographic Data Centre under the numbers 1873806, 1522027, 2453809, and 2453808.

Further reading on this blog includes the article on synthesis routes to OAS-POSS, the post on amide POSS prepared from OAS salts, the overview of T10 decameric silsesquioxanes, the article on octamethylsilsesquioxane, and the complete POSS article index.