This article establishes the capability of NMR, specifically using a mid-field 60 MHz NMR analyzer, as a viable alternative to amide analysis using the current HPLC method.

It describes the work in the following areas:

The identification of pure reference compounds, which constitute the mixture present in the amide stage including:

  • Methacrylamide (MAM)
  • Methacrylic acid (MAA)
  • α-Hydroxyisobutyramide (HIBAM)
  • α-Sulphatoxyisobutyramide (SIBAM)
  • Ammonium acetonedisulphonate (AADS)

The profiling of simple, binary mixtures of components to establish resolution and the suitability of the technique for quantification:

  • HIBAM+SIBAM
  • MAM+MAA

The analysis of “real” amide samples from the mixer and reactor of MM7.

Introduction

Presently, amide mixer and reactor samples from MM7, MM8 and MAA2 are analyzed offline via high performance liquid chromatography (HPLC). Values obtained from this analysis include conversion index (CI), water balance and decomposition index (DI).

Recent requirements for amide reactors have created a need for a continuous optimization program, which will ensure that reactors run at consistently high efficiencies and give the maximum possible output.

Despite the reliance on HPLC to assess the performance of the amide stage (what is the amide ‘stage’),  there  are several  drawbacks that  limit  the  convenience of  this  analysis  for optimization purposes:

1.  Frequency of data collection – Currently 1-2 samples are collected and analyzed every 24 hour period per reactor, this is dictated by the long HPLC system preparation and analysis times required for each sample. Each analysis only yields one data point for each parameter. This low rate of data collection severely limits the rate with which the reactors can be optimized and is insufficient.

2.  Offline sampling – There are difficulties in the handling of amide mixture samples. Many of the components within the amide samples themselves are not stable towards ambient conditions (for instance HIBAM/SIBAM), meaning that the sample composition is prone to change post-sampling. This introduces a further degree of uncertainty and impacts the reliability of the technique. In addition, samples solidify below 40°C and due to preferential crystallization of components; if this is allowed to happen to any degree then samples may not be representative of the amide mix. Both of these factors may give rise to high standard deviations in the observed data and hamper optimization efforts.

3.  Delay – It will often be a matter of hours before HPLC data are fed back for response of the technical team. This means that any measures taken are well after the event that gave rise to the results, by which time the situation may well have already changed. This long delay between sampling and feedback is clearly not ideal.

4.  Reliability – Although not proven, the reliability of the HPLC method itself has often been questioned. Being high in sulfuric acid and sensitive to moisture, the nature of the samples makes them difficult to analyze by HPLC. To overcome this, the current measurement uses specialized graphite columns in the HPLC instruments but resolution of certain peaks in the analysis is not good and the column deteriorates quickly (requiring replacement every 6 months).

Clearly, a need exists for an alternative analysis method which addresses some or preferably all of these challenges in order to provide the technical team with frequent, quick, sample-free (?) and reliable feedback from the amide stage.

Proton Nuclear Magnetic Resonance (1H NMR) Spectroscopy has been identified as a potential technique for the achievement of this aim and offers numerous advantages:

·    Very little system preparation – 1H NMR only requires periodic “shimming”, in order to maintain  homogeneity  of  the  magnetic  field.  

·    Frequency of feedback – A major advantage of 1H NMR is that individual scans only require a matter of seconds to collect. Depending on sample concentration, several scans may be required to obtain a reliable, composite spectra but the high concentration of the amide stage mix will work in our favor in this respect. This has the potential to generate data points numbering in the order of thousands for any given 24 hour period.

·    Online analysis – 1H NMR does not need to come into physical contact with samples and therefore lends itself to being adapted to online analysis. Thus all the problems associated with physical sampling can be circumvented.

·    Reliability —  1H NMR is  inherently quantitative as it  measures the  relative  signals from hydrogen nuclei (protons) present on each of the molecules in the mix, given that complete relaxation  of  these  nuclei  is  allowed  to  occur.  The relative ratios of components can therefore be accurately determined from first principles, without the need for continuous calibration.

1H NMR has the potential to address all the issues surrounding the current HPLC analysis. Conventional NMR spectroscopy however, is both expensive and unwieldy due to the high cost of the instrumentation, the constant supply of cryogens required to cool the magnets and the large, magnetically shielded space required to house equipment.

Technological advances have enabled the use of rare-earth permanent magnets for building relatively small, albeit relatively low field-strength spectrometers. These instruments are typically small enough to fit on the bench, require no cryogens and have their own built-in   magnetic   shielding.   Such   an   instrument, 4IR AI-60 60 MHz   NMR analyzer, was used for this study.

This report describes the preliminary work to analyze pure reference compounds, binary mixtures of particular analysts of interest and finally the analysis of amide samples from MM7 in order to assess the suitability of the technique.

Methods of Investigation

Pure reference compounds were analyzed diluted in both deuterated DMSO (DMSO-d6) and pure sulfuric acid (H2SO4). DMSO-d6 is a commonly used NMR solvent and produces virtually no interference with the rest of the spectrum. However in order to realize the goal of operating a NMR spectrometer on-line with recycling of materials, only components already present in the amide mix should be added and recycled, therefore H2SO4 will also be utilized despite its strong signal at about

11.0 ppm in the resultant spectra.

The analysis of the following reference compounds in both DMSO-d6 and H2SO4 were recorded:

·    Methacrylamide (MAM)

·    Methacrylic acid (MAA)

·    α-Hydroxyisobutyramide (HIBAM)

·    α-Sulphatoxyisobutyramide (SIBAM)

·    Ammonium acetonedisulphonate (AADS)

Some simple binary mixture experiments were also carried out, investigating the resolution and quantitative relationship between the following analysts:

·    HIBAM/SIBAM

·    MAM/MAA

Two final experiments will be carried out to analyze and quantify one amide mixer and one amide reactor sample from MM7.

Preliminary spectra were carried out using a standard 1 dimensional, high-sensitivity 1H NMR experiment dubbed “power scan” in the AI-60 software unless otherwise stated. The parameters of a power scan are as follows:

·    40 scans

·    6.4 second acquisition time

·    15 second repeat time.

·    90° pulse angle

Results and Discussion

Reference Spectra

MAM

Figure 1. The structure of MAM.

Shift/ppmMultiplicityIntegralgroup
1.63Singlet3.00CH3
5.11Triplet1.03C=CHH
5.46Singlet1.03C=CHH
6.81Doublet (broad)1.83CONH2
Figure 2. The 1H NMR Spectrum of MAM in DMSO-d6.

The multiplicity only identifies the visual appearance of the signals, as  some of the true NMR fine structure of the peaks might  not be resolved at this field strength

When reviewing the 1H NMR spectrum of MAM in DMSO-d6, a large singlet is observed at 1.60 ppm with an integral of 3, this signal originates from the 3 methyl protons. Two signals occurring at

5.11 and 5.46 ppm respectively and each integrating for a single proton are assigned to the 2 methylene protons. A broad doublet is observed at 6.81 ppm and is assigned to the 2 amide protons.

Tal, I don’t think that we should enter into any speculations about the fine structure. It is almost certainly more complicated than you think. We can be sure about the assignments only if I look in the literature for high-field spectra. I can do it but it will take much more time

Shift/ppmMultiplicityIntegralgroup
2.20Singlet3.00CH3
6.29Singlet0.98C=CHH
6.62Singlet0.97C=CHH
8.20Singlet (broad)1.53CONH2
Figure 3. The 1H NMR Spectrum of MAM in H2SO4.

When reviewing the 1H NMR spectrum of MAM in H2SO4, the HsSO4 contributes a large singlet peak at 11.22 ppm but otherwise many of the same features are present. All signals have shifted downfield, the methyl singlet to 2.20 ppm, the two  methylene  singlets  to 6.29  and  6.62  ppm respectively and the broad amide singlet to 8.20 ppm.

  • MAA
Figure SStructure of MAA.

Shift/ppmMultiplicityIntegralgroup
1.26Singlet3.00CH3
4.98Triplet1.00C=CHH
5.56Singlet0.99C=CHH
12.10Singlet0.99COOH
Figure 5. The 1H NMR Spectrum of MAA in DMSO-d6.

When reviewing the 1H NMR spectrum of MAA in DMSO-d6, a large singlet is observed at 1.26 ppm with an integral of 3, this signal originates from the 3 methyl protons. Two signals occurring at 4.98 and  5.56ppm  respectively and  each  integrating  for  a  single  proton  are  assigned  to the  2 methylene protons. The singlet at 12.10 ppm is assigned to the carboxylic acid proton.

Shift/ppmMultiplicityIntegralgroup
2.13Singlet3.00CH3
6.45Singlet0.99C=CHH
6.89Singlet0.98C=CHH
Figure 6. The 1H NMR Spectrum of MAA in H2SO4.

When reviewing the 1H NMR spectrum of MAA in H2SO4, many of the same features are present. Similar to  MAM,  all  signals  have  shifted  downfield,  the  methyl  singlet  to  2.13  ppm,  the  two methylene  singlets  to  6.45  and  6.89  ppm  respectively  and  the  carboxylic  acid  singlet  has presumably disappeared beneath the H2SO4 peak.

  • SIBAM
Figure 7. The structure of SIBAM.

Shift/ppmMultiplicityIntegralgroup
0.94Singlet6.00CH3
9.56Singlet (broad)0.81CONH2
Figure 8. The 1H NMR Spectrum of SIBAM in DMSO-d6.

When reviewing the 1H NMR spectrum of SIBAM in DMSO-d6, a large singlet is observed at 0.94 ppm; this signal is assigned to the 6 chemically equivalent methyl protons with an integral of 6. A broad singlet is present at 9.56 ppm, which is tentatively assigned to the 2 amide protons but the integral is difficult to determine due to the broad peak shape.

Shift/ppmMultiplicityIntegralgroup
1.94Singlet6.00CH3
2.17Singlet1.34?
9.56Doublet (broad)1.30CONH2
Figure 9. The 1H NMR Spectrum of SIBAM in H2SO4.

When reviewing the 1H NMR spectrum of SIBAM in H2SO4, the methyl singlet appears to have been shifted downfield and now occurs at 1.94 ppm. Finally, the broad amide doublet can be seen at 8.86 ppm but is poorly resolved from the large, sulfuric acid singlet.

Tal, both up field ‘singlets’ could be from the methyl groups, the solvent could have induced some in equivalence.

  • HIBAM
Figure 10. The structure of HIBAM.

Shift/ppmMultiplicityIntegralgroup
0.93Singlet6.05CH3
4.86Singlet1.00COH
6.67Singlet (broad)1.56CONH2
Figure 11. The 1H NMR Spectrum of HIBAM in DMSO-d6.

When reviewing the 1H NMR spectrum of HIBAM in DMSO-d6, a large singlet with an integral of 6 is observed at 0.93 ppm; this signal is assigned to the 6 chemically equivalent methyl protons. A singlet with an integral of 1 is observed at 4.86 ppm, which is assigned to the OH group proton. A broad singlet is present at 6.67 ppm, which is assigned to the 2 amide protons but, as with SIBAM, the integral is difficult to determine due to the poor peak shape.

Shift/ppmMultiplicityIntegralgroup
1.95Singlet6.00CH3
2.17Singlet1.56?
8.88Doublet (broad)2.04CONH2
Figure 12. The 1H NMR Spectrum of HIBAM in H2SO4.

When reviewing the 1H NMR spectrum of HIBAM in H2SO4, the methyl singlet appears to have been shifted downfield and now occurs at 1.95 ppm and as with SIBAM, there is now a second singlet at

2.17ppm. Finally, the broad amide doublet can be seen at 8.88 ppm but is poorly resolved from the large, sulfuric acid  singlet.  It  is  interesting  that  the  alcohol  singlet  formerly  at  4.86ppm  has

disappeared entirely, which could be caused by rapid exchange with the solvent. The next sentence is too speculative. Interconversion does not necessarily give rise to identical spectra.

  • AADS
Figure 13. The structure of AADS.

Shift/ppmMultiplicityIntegralgroup
1.95Singlet4.00CH2COCH2
2.17Singlet7.69NH + 4
Figure 14. The 1H NMR Spectrum of AADS in DMSO-d6.

When reviewing the 1H NMR of AADS, a large singlet with an integral of 4 can be observed, which is assigned due to the 4 chemically equivalent methylene protons. A second, broader singlet is observed at 6.78ppm with integral value 8, and is likely due to the ammonium protons from the two ammonium groups.

Figure 15. The 1H NMR Spectrum of AADS in H2SO4.

Perhaps the most interesting reference spectrum reviewed here is the spectrum of AADS in H2SO4, which is vastly different from the spectrum acquired in DMSO-d6. At first, there appear to be three signals present, a large doublet at 4.72 ppm and two smaller singlets at 5.93 and 7.17 ppm. However this does not correlate to the known structure present in solution, in particular the doublet is anomalous as there can be no proton-proton coupling present.

A very neat way to explain this unexpected phenomenon is to suggest that 14N-1H coupling of the ammonium signal is occurring here. 14N has a nuclear spin of 1 and is >99% abundant, the resultant splitting pattern would be a triplet with the relative intensities being 1:1:1 as opposed to a triplet arising from proton-proton coupling, the relative intensities for which would be 1:2:1. This is further supported when considering that the peaks at 4.69, 5.93 and 7.17 ppm are separated by a constant

1.24 ppm, i.e. JNH = 52.7 Hz.

Shift/ppmMultiplicityIntegralgroup
1.95Singlet4.00CH2COCH2
5.93Triplet (JNH=52.7Hz)8.71NH + 4

The reason for the stark increase in resolution of this coupling (with respect to AADS in DMSO-d6) is explained in literature1  as being due to an increase in symmetry of the environment of the nitrogen center. This causes a vast increase in the spin-lattice relaxation time (T1) of the nitrogen nucleus and  thus  greatly  increases  the  sharpness  of  this  signal.  This  can  be  explained  by

+

considering  that  the  NH4

protons  are  essentially no  longer  exchanging  in  H2SO4   due  to  the

extremely low pH and as a result, all ammonium ions have now become symmetrical (tetrahedral).

4

The practical implications of this are encouraging as NH +

is an indicator of amide decomposition

and therefore, a decrease in efficiency of the amide stage. The ability to monitor any one of these

+

three triplet peaks would enable the accurate monitoring of NH4 oncentration.

All of this is speculations. The ‘doublet ‘ may not be a doublet at all but two singlets due to some inequivalence. The only way to prove something conclusively is to see a spectrum at higher field strength! The N14 coupling is rarely observed because  its T1 is very short AND the NH4 protons are exchanging with the solvent

The remaining signal in this spectrum must again be due to the methylene protons. This signal poorly resolved from the ammonium triplet peak at 4.69 ppm but can still be observed as a singlet at

4.74 ppm. Integrals are difficult to assign due to this overlap but if an approximate integral of 4 is given, the resultant sum integral for the ammonium triplet equates to 8.7, which is not drastically different from the 8 that is expected.

Binary Mixture Experiments

  • SIBAM/HIBAM

Equal weights of HIBAM and SIBAM were dissolved in DMSO-d6 and a spectrum was obtained from this solution using a “power scan”.

Figure 16. The 1H NMR spectrum of 1:1 SIBAM + HIBAM in DMSO-d6.

The resultant spectrum is not indicative of a simple mixture of the two components. A large singlet is observed at 0.89 ppm, which is slightly up-field of the methyl protons of either SIBAM or HIBAM. Another, much smaller singlet is observed at 1.10 ppm, which is slightly down-field of the methyl protons of  either  SIBAM or  HIBAM.  Interestingly,  the  alcoholic  proton  of  HIBAM is  no  longer observed and only one amide peak is present at 6.99 ppm, which does not correlate well to either compound. The spectrum here does not look particularly like either pure compound and perhaps hints at some kind of intermediate or complex between the two.

A further solution of equal weights of SIBAM and HIBAM was prepared in H2SO4  and a spectrum

was obtained from this solution using a “power scan”.

Figure 17. The 1H NMR spectrum of 1:1 SIBAM + HIBAM in H2SO4

Again the resultant spectrum does not appear to be a simple mixture of the two components. A large singlet is observed at 1.81 ppm, which is slightly up-field of the methyl protons of either SIBAM or HIBAM. Another, much smaller singlet is observed at 2.03 ppm, which is slightly down-field of the methyl protons of either SIBAM or HIBAM. Again, the alcoholic proton of HIBAM is no longer observed and only one amide peak is present at 8.76 ppm, which does not correlate well to either compound. In this instance it is noteworthy that the spectrum bears a striking resemblance to those obtained for either pure SIBAM or HIBAM in H2SO4, adding to the evidence that in this medium, the two are not discreet entities.

In order to better understand these spectra, another mixture of SIBAM + HIBAM was made in H2SO4, this time in a 9:1 ratio. Any signal exhibiting an increased integral with respect to the 1:1 mixtures could be presumed to be due to SIBAM

Figure 18. The 1H NMR spectrum of 9:1 SIBAM + HIBAM in H2SO4

Interestingly, this spectrum hardly differs from the one obtained of the 1:1 mixture; the two singlets at 1.81 and 2.04 ppm are essentially the same relative intensity. The most likely explanation for this would be that, even at room temperature and in either DMSO-d6 or H2SO4, a dynamic equilibrium exists between SIBAM and HIBAM. In this case the proportion of each can be affected by the addition of water so a simple experiment was carried out whereby a drop of water was added to the

1:1 SIBAM + HIBAM solution.

Figure 19. The 1H NMR spectrum of 1:1 SIBAM + HIBAM in H2SO4 with 1 drop of water added.

The relative intensities of the two singlets does indeed appear to shift; the singlet at 2.10 ppm now has an integral of 0.95 vs. its original 1.35. A further 9 drops of water were added to the solution to accentuate this effect.

Figure 20. The 1H NMR spectrum of 1:1 SIBAM + HIBAM in H2SO4 with 10 drops of water added.

With the addition of so much water, the singlet at 2.14 ppm diminishes in intensity quite significantly, now with an integral of 0.31. It can therefore be concluded that the singlet at 2.14 ppm is SIBAM and that the singlet at 1.94 ppm is HIBAM and that SIBAM is hydrolyzing upon addition of water to the system, driving the equilibrium further towards HIBAM.

Tal, is the conclusion of the last 3 pages that quantification is not possible? In this case, no speculations are necessary. It is sufficient to say that we have shown that this solvent cannot be used for quantidication

  • MAM/MAA

As no advantage between DMSO-d6 and H2SO4  had been established up to this point in the investigation, H2SO4  was chosen as the sole diluent for further work. H2SO4  has the advantage of being the process solvent, meaning that a diluted in-line sample could be recycled back into the bulk stream.

Equal weights of MAM and MAA were dissolved in H2SO4  and a spectrum was obtained from this

solution using a “power scan”.

Shift/ppmMultiplicityIntegralgroupComponent
2.01Singlet3.00CH3MAM + MAA
6.09Singlet0.47C=CHHMAM
6.24Singlet0.50C=CHHMAA
6.43Singlet0.49C=CHHMAM
6.64Singlet0.52C=CHHMAA
8.30Singlet (broad)0.40CONH2MAM + MAA

Figure 21. The 1H NMR spectrum of 1:1 MAM +MAA in H2SO4.

This spectrum very clearly shows the mixture of the two components. The methyl protons of each are compounded as a singlet at 2.01 ppm but the methylene protons of each are clearly resolvable from one another. The methylene protons of MAM occur at 6.09 and 6.43 ppm respectively and the methylene protons of MAA at 6.24 and 6.64 ppm respectively.

Figure 22. An expanded view of 1H NMR spectrum of 1:1 MAM + MAA in H2SO4, showing the methylene protons of the two compounds.

A  comparison  of  the  two  integrals  6.64  and  6.43  ppm  can  be  used  to  evaluate  the  relative concentrations of MAM and MAA:

The recorded weights used were:

Table 1. The actual weights of MAM and MAA used for the 1:1 binary mixture.

This crudely derived theoretical value only differs with the actual values used by 0.5 mol%.

MAM and MAA in a ratio of 99:1 were then dissolved in H2SO4 and a spectrum was obtained from this solution using another “power scan”.

Shift/ppmMultiplicityIntegralgroupComponent
2.01Singlet3.13CH3MAM + MAA
6.09Singlet1.00C=CHHMAM
6.24Singlet1.03C=CHH/ C=CHHMAM + MAA
6.43Singlet0.02C=CHHMAA
8.25Singlet (broad)1.79CONH2MAM + MAA
Figure 23. The 1H NMR spectrum of 99:1 MAM + MAA in H2SO4.

Again, the two components can be seen in the spectrum. The methylene protons of each compound are less resolved. The methylene protons of MAM occur at 6.09 and 6.43 ppm respectively and the methylene protons of MAA at 6.24 and 6.64 ppm respectively.

Figure 24. An expanded view of 1H NMR spectrum of 99:1 MAM + MAA in H2SO4, showing the methylene protons of the two compounds.

A  comparison  of  the  two  integrals  6.77  and  6.16  ppm  can  be  used  to  evaluate  the  relative concentrations of MAM and MAA:

The recorded weights used were:

Table 2. The actual weights of MAM and MAA used for the 99:1 binary mixture.

This time the difference between the actual and measured values is greater at 1.0 mol%, which is unsurprising, given the poor resolution between the peaks. However, the conditions presented here are completely optimised and can be considered to be reasonably accurate although lacking in precision. It is also noteworthy that this accuracy is consistent with binary mixtures ranging from 1:1 to 99:1, without any form of calibration.

A decrease in sample viscosity will almost certainly result in narrower line widths and therefore increase  precision.  This  is  an  obvious  area  for  optimisation  and  can  be  achieved  either  by increasing the amount of H2SO4 diluent or by resorting to a less viscous solvent if needs be.

Amide samples

  • MM7 mixer sample.

A sample was taken directly after the static mixer of MM7 and diluted to varying degrees using

H2SO4. The resultant solutions were analysed via 1H NMR using a “powerscan”.


Figure 25. The 1H NMR spectrum of a MM7 amide mixer sample diluted 1:1 with H2SO4

By using a 1:1 dilution factor, a very good signal-to-noise ratio is achieved. However, as anticipated the high viscosity of the same sample has caused to broad line-width and poor resolution.

A large singlet is observed at 1.62 ppm, which is due to the methyl protons of HIBAM. A second singlet at 1.84 ppm is assigned to SIBAM but is very poorly resolved.

A large singlet at 2.00 ppm is assigned to the methyl protons of MAM, although it is only partially resolved from the singlets at 1.62 and 1.84 ppm.

A weak signal is observed at 4.68 ppm and is thought to be one part of the ammonium (AADS) triplet, along with a signal at 7.13 ppm. The remaining part of the triplet is thought to be unresolved from the methylene protons of MAM, this idea is supported by the appearance of a shoulder on the peak at 6.07 ppm. Again, I am not sure about the triplet

The two singlets at 6.07 and 6.39 ppm are due to the methylene protons of MAM.

The large broad signal at 8.23 ppm is most likely a composite signal arising from amide groups in

MAM, SIBAM and HIBAM.

It is noteworthy that, at this dilution no signals accounting for MAA can be seen.

Shift/ppmMultiplicityIntegralgroupComponent
1.73Singlet3.42CH3HIBAM
1.96Singlet0.59CH3SIBAM
2.09Singlet3.00C=CHHMAM
5.88Triplet (JNH=51.0Hz)0.09*NH4AADS**
6.17Singlet1.00C=CHHMAM
6.50Singlet1.01C=CHH/ C=CHHMAM + MAA
6.85Singlet0.01C=CHHMAA
8.18Singlet (broad)2.05CONH2MAM
8.87Singlet (broad)0.53CONH2SIBAM + HIBAM

Figure 26. The 1H NMR spectrum of a MM7 amide mixer sample diluted 1:5 with H2SO4. *Obtained by taking the least interfered with signal from the triplet at 7.15 ppm and multiplying by 3.**All ammonium assumed to be arising from AADS at this time.

By increasing the dilution factor to 1:5, a good signal-to-noise ratio is still achieved and resolution is markedly improved.

In addition to the features noted in the previous spectrum, resolution between the three singlets at

1.73, 1.96 and 2.09 ppm is better but assigning an accurate integral for SIBAM remains challenging.

A new signal appearing at 5.88 ppm is thought to be the remaining part of the ammonium triplet and is no longer obscured by the singlet at 6.17 ppm.

Crucially, one of the MAA methylene singlets can now be resolved and can be seen as a very weak peak at 6.85 ppm.

Given the information already known about the components of the amide mix, the following rough approximation of their relative concentrations can be made.

Table 3. The predicted molar ratios of components in the MM7 mixer sample, based on 1H NMR integrals. *1.33 protons due to only one third of the NH + triplet being integrated.

It is emphasized that all values reported here are approximate and that the method used here must be subject to further optimization.

Figure 27. The 1H NMR spectrum of a MM7 amide mixer sample diluted 1:20 with H2SO4

When increasing the dilution factor further to 1:20, it can be remarked that the signal-to-noise ratio has now become unacceptable as the weaker signals are now obscured or lost altogether. This could in theory be remedied by increasing the number of scans and increasing run-times. However this spectrum shows no increased resolution with respect to the 1:5 dilution sample and therefore it appears to be unnecessary to operate at such high dilutions.

  • MM7 Reactor Sample.

A MM7 reactor sample was analyzed at the 1:5 “optimum” dilution using a “power scan”

Figure 28. The 1H NMR spectrum of a MM7 amide reactor sample diluted 1:5 with H2SO4. *Obtained by taking the least interfered with signal from the triplet at 7.15 ppm and multiplying by 3.**All ammonium assumed to be arising from AADS at this time.

A small singlet with an integral of 0.07 is observed a 1.75 ppm; this is most likely arising from the HIBAM methyl protons. The equivalent signal from SIBAM is barely visible at 1.95 ppm but is present with an integral of 0.04

A large singlet is once again produced at 2.11 ppm by the methyl protons of MAM and has an integral of 3.22.

All three peaks of the ammonium triplet can be observed at 4.67, 5.91 and 7.15 ppm. The peak at

7.15 ppm is thought to be least interfered with and has an integral of 0.07.

The  two methylene protons of  MAM once again produce two  singlets  at  6.20  and  6.52  ppm respectively. The signal at 6.20 ppm is assumed to be free of interference and has been assigned the integral of 1.00. The other singlet at 6.52 ppm has an integral of 1.07 ppm

One of the methylene peaks of MAA is clearly resolved at 6.86 ppm and has an integral of 0.03. The broad peak at 1.74 ppm is due to the amide protons of MAM.

As for the mixer sample, the following rough approximation of relative concentrations of components can be made.

Again, it is emphasized that the values reported here are approximate; further optimization will be required to develop a method for obtaining precise and repeatable quantitative results.

4.  Conclusions Summary

A range of purified components of the amide stage have been profiled via 1H NMR in DMSO-d6 and H2SO4, including:

  • Methacrylamide (MAM)
  • Methacrylic acid (MAA)
  • α-Sulphatoisobutyramide (SIBAM)
  • α-Hydroxyisobutyramide (HIBAM)
  • Ammonium acetonedisulphonic acid (AADS)

Binary mixtures of similar compounds have also been profiled:

  • SIBAM + HIBAM; 1:1 mixtures have been examined in both DMSO-d6 and H2SO4.
  • Interconversion between the two occurs within the NMR sample at room temperature, the degree of which is dependent on the water content. It is safe to assume the same is true of amide samples from the plants.
  • High degree of similarity between these compounds means the two can only be resolved from one another with difficulty. It remains to be seen if they can be separated mathematically to give a good indication of “water balance”.

MAM +MAA; varying concentrations have been investigated in  H2SO4

  • 1:1 mixture shows clear resolution between the two compounds
  • 99:1 mixture shows that resolution can still be achieved but would benefit from further optimization to give a reliable value for MAA.
  • A good quantitative approximation can be drawn from these mixtures.

Amide samples from MM7 have been studied from both mixer and reactor.

  • The ideal dilution of these samples has been identified as 1:5 in H2SO4.
  • At this dilution, all aforementioned components can be observed in these samples, however further development of the method will be needed to quantify some of the more minor components of the mixture, such as HIBAM, SIBAM and MAA.

5.  Recommendations.

·    Parameters of the 1H NMR experiment now need to be developed; until now the majority of

experiments have been conducted using the instrument’s default experiment settings. It is

possible that scan times can be significantly reduced without detrimentally affecting the quality of spectra. Both acquisition and repetition times will be investigated in order to achieve this.

·    A suitable internal reference may be required in order to determine absolute concentrations (as opposed to relative ones) in spectra. Many conventional 1H NMR reference compounds rely on the Si-Me group to provide a signal with a neutral chemical shift, i.e. 0 ppm. However this group is decomposed by H2SO4 and so an alternative should be found.

·    Further work needs to be done to investigate separating various signals mathematically to give more reliable integrals and aid with quantification.

·    If possible, values for CI, water balance and DI should be derived from 1H NMR, however more suitable metrics than these values may arise as a result of further work.

·    Pending the definition of an “optimum” experiment. A regime of sample analysis alongside the current HPLC method should be undertaken and a rolling comparison made to establish the suitability and reliability of 1H NMR for quantification.

Want to know more?
Let's talk.