Two Methods for the Determination of Polymer Hydroxyl Values Using 19F Benchtop NMR

The hydroxyl value (HOV) of a polymer, typically reported in mg KOH/g of polymer, is one of the key parameters typically reported in the technical data sheet (TDS) or certificate of analysis (CoA) of polymers containing hydroxyl functionalities. Several methods exist to determine the HOV of polymers, the most common of which is an acid-base titration. While conceptually simple, this approach is in fact elaborate, expensive, and wasteful from both a reagent and solvent perspective. In this application note, we present two quantitative benchtop nuclear magnetic resonance (qNMR) methods, which require very little solvent (deuterated solvents are not required) and small amounts of reagents. One of these methods is appropriate for polymers soluble in organic solvents, such as chloroform, while the other method is recommended for water-soluble polymers.

Introduction

The hydroxyl value (HOV) of a polymer, typically reported in mg KOH/g of polymer, is one of the key parameters typically reported in the technical data sheet (TDS) or certificate of analysis (CoA) of polymers containing hydroxyl functionalities. This value provides insight into the stability and reactivity of these materials and is commonly reported along with the number-average molecular weight (M_n), the melting point, and the reactive groups (hydroxyl functionality) of a polymer. As such, the determination of the HOV of a polymer is an important component of quality assurance/quality control (QA/QC) assays.

Several methods exist to determine the HOV of polymers, the most common of which is an acid-base titration. Typically, the hydroxyl functionalities in the polymer are acetylated using acetic anhydride, then water is added to neutralize the remaining anhydride to acetic acid. The amount of converted acetic acid is then determined by titration with potassium hydroxide. As such, the HOV is a measure of how many milligrams of potassium hydroxide are required to neutralize the residual acetic acid per gram of polymer being analyzed (i.e., mg KOH/g polymer).

From an industrial perspective, this method is typically performed according to ASTM E222.¹ While conceptually simple, this approach is in fact elaborate, expensive, and wasteful from both a reagent and solvent perspective. Briefly, this method requires the use of a pressure bottle, an acetylation mixture prepared fresh daily (127 mL of acetic anhydride in 1 L of pyridine), hydrochloric acid, potassium hydroxide, phenolphthalein, and heating to 98 ± 2 °C for 2-4 hours. By contrast, the two quantitative benchtop nuclear magnetic resonance (qNMR) methods we propose herein require very little solvent (deuterated solvents are not required) and small amounts of reagents. Additionally, the reactions occur quickly, and the analyses do not require long run-times or extensive post-processing to obtain accurate and reproducible results.

Trifluoroacetic Anhydride Method

This method is closely based on work previously published by Foli et al. and involves the use of trifluoroacetic anhydride in chloroform to transform the hydroxyl groups in the polymer into trifluoroacetyl functionalities (Figure 1).² Then, all volatiles are removed in vacuo,³ the residue is re-dissolved in chloroform, and a known amount of trifluorotoluene is added as a stock solution in chloroform. The trifluorotoluene acts as an internal calibrant (IC) for qNMR,⁴ allowing for the accurate quantification of hydroxyl groups via ¹⁹F NMR and using the signal from the new trifluoroacetyl groups.

Figure 1. General reaction scheme for the transformation of hydroxyl groups into trifluoroacetyl fragments using trifluoroacetic anhydride in chloroform.

This approach works well for determining the HOV of polymers soluble in chloroform and each analysis requires less than three minutes. Here, we analyzed different types of polymers using this method, including: poly(ethylene glycol) (PEG), polyoxyl 10 oleyl ether, polyoxyl 20 cetostearyl ether, and poly(ethylene glycol) monomethyl ether (Ymer^TM, mPEG). Within these types of polymers, several different products were analyzed, to demonstrate the universality of this method across products with varying HOV.⁵

For each product, the sample was analyzed in triplicate and the results were compared to the values included in the CoA provided by the supplier. A summary of the results is presented in Table 1. A representative ¹⁹F NMR spectrum collected on a 60 MHz benchtop NMR spectrometer (¹⁹F operating frequency = 56.5 MHz) is shown in Figure 2. Note that the residual trifluoroacetic acid present in the sample can be removed by successive washings with toluene and removal of volatiles, via the formation of an azeotrope.⁶ Since the presence of this signal does not impact the results, we opted to minimize the length of the process and proceed with the analysis.

Figure 2. ¹⁹F (57.1 MHz) NMR spectrum of Ymer^TM N90 (mPEG) in chloroform after transformation with trifluoroacetic anhydride. The trifluorotoluene is added as an IC for qNMR.

Pentafluoropyridine Method

This method is adapted from work previously published by Beckham et al. and involves the use of pentafluoropyridine in dimethyl sulfoxide (DMSO).⁷ In the original report, the authors describe the use of pentafluoropyridine and potassium carbonate in a 40% H₂O/DMSO mixture to transform the phenolic hydroxyl groups in lignins into a tetrafluoropyridyl-ether, which can be analyzed using ¹⁹F NMR (Figure 3). Here, we sought to explore the potential of this technique for the determination of HOV in poloxamers.⁸ Due to solubility concerns during method optimization, we have found that performing the transformation in DMSO and omitting water, followed by the addition of acetone to ensure solubility of all components prior to analysis, led to the most accurate and reproducible results. The addition of acetone was also performed in the original report for certain lignins exhibiting similar solubility concerns. Prior to analysis, a known amount of trifluorotoluene for use as an IC in qNMR was added as a stock solution in DMSO.

Figure 3. General reaction scheme for the transformation of hydroxyl groups into tetrafluoropyridyl-ether fragments using pentafluoropyridine and potassium carbonate in DMSO.

For each product, the sample was analyzed in triplicate and the results were compared to the values included in the CoA provided by the supplier. A summary of the results is presented in Table 1. A representative ¹⁹F NMR spectrum collected on a 60 MHz benchtop NMR spectrometer (¹⁹F operating frequency = 56.5 MHz) is shown in Figure 4. Note that the presence of excess pentafluoropyridine signals is important to confirm that sufficient reagent was used during the reaction. Additionally, the fluorine peak at δ = 104 ppm, relating to the fluorines ortho to the ether fragment on the pyridyl ring, was arbitrarily chosen as the integrated signal for all qNMR calculations.

For these experiments, each analysis took approximately one hour, because of the more dilute sample and a need for a longer interscan delay. For both methods, inversion-recovery experiments were performed to determine the longest T₁ time for the signals of interest. Then, the acquisition parameters were set such that an interscan delay of at least five times this value was used. This allows the spins of interest to fully relax between pulses, ensuring quantitative data collection.⁹

Figure 4. ¹⁹F (57.1 MHz) NMR spectrum of Synperonic^TM PE/L 101 (poloxamer) in DMSO/ acetone after transformation with pentafluoropyridine. The trifluorotoluene is added as an IC for qNMR. The polymer signal at δ = 104 ppm was used for quantification.

Table 1. Summary of the HOV obtained for all polymers analyzed in this work. The averages of triplicate analyses are shown, and the relative standard deviation (RSD) values are included in parentheses. The CoA values are included for comparison and were those included by the supplier.

Conclusion

The work presented herein demonstrates that benchtop NMR spectroscopy can be used for the accurate and reproducible determination of HOV in various types of polymers. In this study, we analyzed PEG, polyoxyl 10 oleyl ether, polyoxyl 20 cetostearyl ether, and Ymer^TM polymers using the trifluoroacetic anhydride method. Additionally, we analyzed poloxamers using the pentafluoropyridine method. Both approaches work well and can be used on a case-bycase basis depending on the solubility of the polymer being analyzed. The simplicity of the two methods shown here, combined with the accessibility, automatability, and affordability of benchtop NMR instruments makes this an appealing alternative to titration for these important QA/QC assays. Furthermore, the use of ¹⁹F NMR allows for the use of non-deuterated solvents and makes use of the much larger chemical shift range of fluorine signals, especially as compared to the relatively narrow range of signals observed via ¹H NMR. Finally, NMR as a technique allows for many more types of analyses than titration, such as structural elucidation, purity determination, numberaverage molecular weight (M_n) determination, confirmation of number of repeating monomeric units in a polymer, and many more.

References

(1) ASTM Standard E222, 2017, Standard Test Methods for Hydroxyl Groups Using Acetic Anhydride

Acetylation, ASTM International, West Conshohocken, PA.

(2) Foli, G.; Degli Esposti, M.; Toselli, M.; Morselli, D.; Fabbri, P. Analyst 2019, 144, 2087–2096.

(3)(a) Caution: trifluoroacetic acid is corrosive, harmful when inhaled, and causes severe skin burns.

When removing volatiles in vacuo for this method, we opted to introduce a trap after the rotary

evaporator to both neutralize the acid and eliminate the strong odors. This was done by preparing a

basic solution of potassium hydroxide in water and venting the volatiles into this solution by using a

piece of Tygon® tubing.

(b) Trifluoroacetic acid safety data sheet: https://www.sigmaaldrich.com/CA/en/sds/SIGALD/

T6508 (accessed May 2023).

(4) Araneda, J. F.; Bernstein, M. A.; Chu, T.; Leclerc, M. C.; Riegel, S. D. Chapter 8 – Quantitative NMR

in Undergraduate Labs in NMR Spectroscopy in the Undergraduate Curriculum, Volume 4: In-Person

and Distance Learning Approaches; American Chemical Society: Washington, DC, 2021. pp 111-120.

(5) Polymers analyzed using the trifluoroacetic anhydride method were received from various suppliers

and used without further purification. MilliporeSigma: Poly(ethylene glycol) p/n: 81260, lot: BCBR7197V;

Brij® O10 p/n: P6136, lot: MKCG0705. Ethox Chemicals: ETHOXCARE® OA-10 lot: 31H054;

ETHOXCARE® CSA-20 lot: 32A062. Protameen Chemicals: ProcolTM OA-10 lot: 17-1280-21B03. BASF:

Kolliphor® CS 20 p/n 30554459, lot: 0023742681. Croda: Brij CS20 lot: 0001778834. Perstorp: YmerTM

N90 lot: 201008930; YmerTM N120 lot: 230304274; YmerTM N180 lot: 230100772.

(6) Delsuc, N.; Léger, J.-M.; Massip, S.; Huc, I. Angew. Chem. Int. Ed 2007, 46, 214–217.

(7) Kenny, J. K.; Medlin, J. W.; Beckham, G. T. ACS Sustain. Chem. Eng. 2023, 11, 5644–5655.

(8) Polymers analyzed using the pentafluoropyridine method were received from various suppliers

and used without further purification. BASF: Kollisolv® P124 p/n: 30631058, lot: GNE08023B. Croda:

SynperonicTM PE/L 101 lot: 0401AD6451; SynperonicTM PE/L 64 lot: 1902AD6532; SynperonicTM PE/L62

lot: 1603BC7183; SynperonicTM PE/L61 lot:0001853719. Millipore Sigma: Synperonic® F108 p/n: 07579,

lot: BCBX1044.

(9) Pauli, G. F. Phytochemical Analysis 2001, 12, 28–42.