- Email: [email protected]
Simple spectrophotometry method for the determination of sulfur dioxide in an alcohol-thionyl chloride reaction
Accepted Manuscript Simple Spectrophotometry Method for the Determination of Sulfur Dioxide in an Alcohol-Thionyl Chloride Reaction Jinjian Zheng, Feng Tan, Robert Hartman PII:
S0003-2670(15)00793-X
DOI:
10.1016/j.aca.2015.05.043
Reference:
ACA 233957
To appear in:
Analytica Chimica Acta
Received Date: 1 March 2015 Revised Date:
25 May 2015
Accepted Date: 27 May 2015
Please cite this article as: J. Zheng, F. Tan, R. Hartman, Simple Spectrophotometry Method for the Determination of Sulfur Dioxide in an Alcohol-Thionyl Chloride Reaction, Analytica Chimica Acta (2015), doi: 10.1016/j.aca.2015.05.043. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
3
TE EP
1.5 1
AC C
Absorbance
2
D
2.5
Sample solution Sample blank solution Sulfur dioxide standard solution Sulfur dioxide blank solution
0.5 0 200
250
300 Wavelength (nm)
350
400
ACCEPTED MANUSCRIPT
Simple Spectrophotometry Method for the Determination of Sulfur Dioxide in an Alcohol-
RI PT
Thionyl Chloride Reaction Jinjian Zheng, Feng Tan, and Robert Hartman
Analytical Development and Commercialization – API, Merck & Co, Inc., Rahway, New Jersey
SC
07065, USA
M AN U
*Corresponding author. Email: [email protected]. Tel: 732-594-4515. Fax: 732-594-
EP
TE
D
3887
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
1
ACCEPTED MANUSCRIPT
Abstract
RI PT
Thionyl chloride is often used to convert alcohols into more reactive alkyl chloride, which can be easily converted to many compounds that are not possible from alcohols directly. One important reaction of alkyl chloride is nucleophilic substitution, which is typically conducted under basic
SC
conditions. Sulfur dioxide, the by-product from alcohol-thionyl chloride reactions, often reacts with alkyl chloride to form a sulfonyl acid impurity, resulting in yield loss. Therefore, the alkyl
M AN U
chloride is typically isolated to remove the by-products including sulfur dioxide. However, in our laboratory, the alkyl chloride formed from alcohol and thionyl chloride was found to be a potential mutagenic impurity, and isolation of this compound would require extensive safety measures. As a result, a flow-through process was developed, and the sulfur dioxide was purged
D
using a combination of vacuum degassing and nitrogen gas sweeping. An analytical method that
TE
can quickly and accurately quantitate residual levels of sulfur dioxide in the reaction mixture is desired for in-process monitoring. We report here a simple ultraviolet (UV) spectrophotometry
EP
method for this measurement.
This method takes advantage of the dramatic change in the UV absorbance of sulfur dioxide with respect to pH, which allows for accurate quantitation of sulfur dioxide in the presence of the
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
strong UV-absorbing matrix. Each sample solution was prepared using 2 different diluents: 1) 50 mM ammonium acetate in methanol + 1% v/v hydrochloric acid, pH 1.3, and 2) 50 mM ammonium acetate in methanol + 1% glacial acetic acid, pH 4.0. The buffer solutions were carefully selected so that the UV absorbance of the sample matrix (excluding sulfur dioxide) at 276 nm remains constant. In the pH 1.3 buffer system, sulfur dioxide shows strong UV absorbance at 276 nm. Therefore, the UV absorbance of sample solution is the sum of sulfur dioxide and sample matrix. While in the pH 4.0 buffer system, sulfur dioxide has negligible UV 2
ACCEPTED MANUSCRIPT
absorbance at 276 nm, and the UV absorbance is attributed only to sample matrix. Quantitation
RI PT
of sulfur dioxide is achieved by subtracting the UV absorbance of sample solution at pH 4.0 from that at pH 1.3. The method is simple but sensitive, with a limit of quantitation of 80µg L-1. The method linearity was demonstrated from 2mg L-1 to 40mg L-1 with an R2 of 0.998, and the
SC
spiked recovery ranges from 94% to 105% within the same range. The results are comparable with those obtained using inductively coupled plasma-atomic emission spectrometry (ICP-AES)
EP
TE
D
M AN U
and gas chromatography-mass spectrometry (GC-MS), suggesting that this method is accurate.
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
3
ACCEPTED MANUSCRIPT
Key Words: Sulfur Dioxide; Sulfite; Spectrophotometry; Alkyl Chloride; Alcohol; Thionyl
EP
TE
D
M AN U
SC
RI PT
Chloride
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
4
ACCEPTED MANUSCRIPT
1. Introduction
RI PT
Alcohol is not reactive for nucleophilic substitution reaction because the hydroxyl leaving group is a strong base and thus not a good leaving group. One common approach to improve the reactivity of alcohol in nucleophilic substitution reactions is to convert the hydroxyl group into a
SC
better leaving group such as a sulfonate ester, phosphite ester, or alkyl halide. Thionyl chloride is commonly used for this conversion since the alkyl chloride product is very reactive, and the
M AN U
inorganic by-products such as sulfur dioxide and hydrogen chloride are gases, which can be purged by sparging with inert gas or degassing under vacuum [1-2]. In a process developed in Merck Research Laboratories, an alcohol is subjected to reaction with thionyl chloride to form an alkyl chloride intermediate (chlorination step), which subsequently
D
reacts with an amide (alkylation step) to form an active pharmaceutical ingredient (API). The
TE
schematic diagram of this synthesis is shown in Figure 1. In the original process, the alkyl chloride intermediate was isolated to remove the solvents and by-products. However, it was later
EP
determined that the alkyl chloride intermediate was a potential mutagenic impurity. Isolation and handling of this intermediate would require extensive safety measures. Therefore, a flow-through process was developed, in which the whole reaction mixture including solvents and by-products
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
from the chlorination step was carried onto the subsequent alkylation step. One issue that arises is that sulfur dioxide reacts with the alkyl chloride intermediate under the basic alkylation conditions, resulting in yield loss and elevated level of sulfonyl acid impurity as shown in Figure 1. As a result, sulfur dioxide has to be purged to below 5mg/mL to achieve satisfactory yield in the next step. The degassing process was very challenging due to the high solubility of sulfur dioxide in the reaction stream, which is mainly composed of the solvent, N-methyl-2-pyrrolidone (NMP) [3]. A combination of heating, sweeping of headspace with nitrogen gas, and application 5
ACCEPTED MANUSCRIPT
of vacuum was required to reduce the sulfur dioxide to the desired level. Due to stability
RI PT
concerns of the alkyl chloride intermediate at elevated temperature, it is preferable to stop the degassing process as soon as the sulfur dioxide is purged to below the desired level. For this purpose, an analytical method to quickly and accurately determine the sulfur dioxide
SC
concentration during degassing was desired.
There are many reports for the analyses of sulfur dioxide, including ion chromatography (IC) [4],
M AN U
fluorescence [5], gas chromatography-mass spectrometry (GC-MS) [6], colorimetric [7], ultraviolet (UV) spectrophotometry [8], and inductively coupled plasma-atomic emission spectrometry (ICP-AES). However, each method has some limitations. Very often, a large amount of organic compounds in the reaction mixture precipitate in the aqueous mobile phase
D
that is used for IC analysis, clogging the column and contaminating the ion suppressor. In GC-
TE
MS analysis, the inlet liner is easily contaminated. Additionally, the GC column deteriorates quickly due to a strongly acidic sample matrix, requiring frequent replacement. Colorimetric
EP
method [7] typically involves lengthy sample treatment such as color development, and is not preferred for in-process control, which requires a fast analysis. The UV spectrophotometry method described by A. Syty [8] cleverly utilizes strong acid to release sulfur dioxide from an
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
absorbing solution, and transfer it to a flow-through absorption cell using a carrier gas for the measurement of UV absorbance at 215 nm. Unfortunately, sulfur dioxide has very good solubility in NMP, the solvent that was used for the chlorination reaction even in the presence of about 1 M hydrochloric acid, the by-product of chlorination reaction. Therefore, this UV spectrophotometry method is not suitable for the samples studied in this article. A common method for fluorescence detection of sulfur dioxide is based on its reaction with ophthalaldehyde and amine in a basic solution to form an isoindole derivative [5]. This method
6
ACCEPTED MANUSCRIPT
has good sensitivity, but often requires extensive method development for this type of highly
RI PT
acidic and complex sample matrix. ICP-AES is very sensitive for measuring the total sulfur in the solution, but the instrumentation can be very expensive and therefore not available in most common analytical laboratories. Fourier transform infrared is commonly used for in-process
SC
monitoring, and works well for sulfur dioxide standard [9]. However, the detection limit for a sulfur dioxide in the process under investigation here is above 20mg mL-1 due to strong matrix
M AN U
interferences, and thus cannot meet the requirements of this analysis.
It is well known that free sulfur dioxide in a solution has strong UV absorbance at 276 nm under acidic conditions [10]. Therefore, it is possible to use UV spectrophotometry for the determination of sulfur dioxide. However, if the sample matrix also shows strong UV absorbance
D
at 276 nm, accurate quantitation of the sulfur dioxide will be extremely difficult.
TE
In this study, we report a simple but sensitive method for the quantitation of sulfur dioxide by taking advantage of the dramatic change in UV absorbance of sulfur dioxide with respect to pH.
EP
This method has been demonstrated to be sensitive and accurate for the quantitation of sulfur dioxide with minimal sample treatment even in the presence of strong background absorbance from reaction mixture. This method can be applied to quantitate sulfur dioxide in most thionyl
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
chloride reaction streams.
2. Experimental 2.1 Apparatus and Reagents All absorption spectral recordings and absorbance measurements were performed on a Cary-300 UV spectrophotometer (Varian, USA) with 1-cm light-path quartz cuvette. The cuvette was
7
ACCEPTED MANUSCRIPT
rinsed 3 times using the solution to be measured before filling up for UV absorbance
RI PT
measurement at the desired wavelength. All chemicals and reagents used were of ACS reagent grade or higher purity. Methanol, trimethylamine, and water were HPLC grade and were purchased from Fisher Scientific
SC
(Pittsburgh, PA, USA). Glacial acetic acid (Certified ACS grade), sodium sulfite (Sigma-Ultra, anhydrous), ammonium acetate (≥99.99% trace metal basis), and hydrochloric acid (ACS reagent,
M AN U
36-38%) were purchased from Sigma-Aldrich (St. Louis, MO, USA).
The alcohol-thionyl
chloride reaction mixture was synthesized by Merck Sharp & Dohme Corp., a subsidiary of
3. Results and Discussions
TE
3.1 Theory
D
Merck & Co. Inc. (Rahway, NJ, USA).
EP
In aqueous solution, sulfur dioxide forms SO2.nH2O complex, which dissociates into bisulfite:
Ka=1.3x10-2 or pKa=1.9
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
In the vicinity of the pKa, the composition of sulfur dioxide/bisulfite changes dramatically with the pH. For example, at pH 4.0, only about 1% of sulfur dioxide exists as free sulfur dioxide while at pH 1.3, about 80% of sulfur dioxide exists as free sulfur dioxide. It is generally agreed that the UV absorbance of a sulfur dioxide solution at 276 nm is due to the free sulfur dioxide dissolved in the solution [10]. Sodium sulfite solution in 50 mM ammonium acetate aqueous solution (pH~6.8) has strong UV absorbance below 220 nm but has almost no UV absorbance at 276 nm as shown in Figure 2. When the same solution was prepared in the 8
ACCEPTED MANUSCRIPT
presence of 1% hydrochloric acid (pH~1.3), it shows strong UV absorbance at 276 nm as shown
RI PT
in the overlaid UV spectra in Figure 2. In our process, the alcohol-thionyl chloride reaction is carried out in NMP. The chlorinated product will precipitate if the sample is diluted with water. Therefore, methanol was used as the
SC
diluent for the preparation of sample solutions as it provides good solubility for both the sample matrix and sulfur dioxide.
M AN U
We have demonstrated that the UV absorbance of sulfur dioxide in methanol changes with respect to pH in a similar way as that in an aqueous solution as shown in Figure 3. When sulfur dioxide is dissolved in methanol, it shows strong UV absorbance at 276 nm. However, the absorbance at 276 nm disappears when 0.1% v/v triethylamine was added to the solution.
D
In sulfur dioxide-methanol solution, majority of sulfur dioxide will react with methanol to form
TE
methyl ester of sulfurous acid in analog to its reaction with water to form sulfurous acid as suggested by Guss and Kolthoff[11]. Addition of TEA base will shift the equilibrium to the right,
EP
causing loss of UV absorbance at 276 nm.
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
In this article, we will focus on how to take advantage of this property of sulfur dioxide to selectively determine sulfur dioxide in the presence of strong UV absorbing organic compounds.
3.2 UV Absorbance of Sample Matrix vs. pH In the previous section, we reported that the UV absorbance of sulfur dioxide was reduced to non-detectable by the addition of 0.1% triethylamine. However, we noticed that the UV spectrum of the sample matrix also changed significantly. Therefore, it is difficult to determine the net absorbance of sulfur dioxide and to quantitate it. 9
ACCEPTED MANUSCRIPT
Due to the complexity of the sample matrix, it is difficult to predict how its absorbance is
RI PT
affected by pH changes. It is also not possible to predetermine the UV absorbance of a sample matrix and subtract it from subsequent sample measurements as it varies with the product concentration, reaction completion rate, time, etc. Therefore, experimental determination of the
SC
UV absorbance for each sample matrix is necessary.
To determine the variation of UV absorbance of the sample matrix with respect to pH changes, a
M AN U
reaction mixture that is free of sulfur dioxide was prepared by vacuum degassing for 40 hours followed by sparging with nitrogen gas for 72 hours. The solution was analyzed using ICP-AES, and was found to contain less than 0.05mg/mL sulfur dioxide, which is well below the target concentration of 5mg mL-1.
D
Ideally, the matrix absorbance should remain constant at 2 different pHs, while the absorbance of
TE
sulfur dioxide should change significantly. Given that sulfurous acid has a pKa of 1.9, the pH was studied in the range of 1.0 to 6.0. The pH values above 6.0 were not studied as sulfurous
EP
acid is expected to be fully ionized to bisulfite and/or sulfite at pH6.0. Since the sample matrix contains acidic by-products (HCl and SO2), ammonium acetate was added to provide buffering capacity between pH 3.0 and 6.0.
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
The sulfur dioxide free reaction mixture was diluted using diluents that contained 50 mM ammonium acetate in methanol with pH adjusted to 1.0, 1.3, 2.0, 3.0, 4.0, 5.0 and 6.0 using hydrochloric acid. The UV spectra of these solutions are shown in Figure 4. The UV spectra are superimposable for pH 1.0~4.0. At pH5.0, the UV spectrum starts to shift, although the UV absorbance between 270nm and 310nm does not change significantly. Therefore, the UV absorbance from the sample matrix is deemed constant as long as the diluent pH is controlled between 1.0 and 4.0. 10
ACCEPTED MANUSCRIPT
Based on the pKa of sulfurous acid (1.9), we can estimate that at pH 4.0, sulfite will not convert
RI PT
to free sulfur dioxide, and thus have no UV absorbance at 276 nm. Therefore, the sample solution prepared using pH 4.0 buffer can be used as a blank because the UV absorbance at 276 nm can be solely attributed to the matrix.
SC
On the other hand, from Figure 2, we demonstrated that sulfur dioxide shows strong absorbance at 276 nm in the presence of 1% hydrochloric acid (pH 1.3). The UV absorbance of the sample at
M AN U
this pH is the sum of UV absorbance from sample matrix and sulfur dioxide. Therefore, the difference in UV absorbance at 276 nm between pH 1.3 solution and pH 4.0 solution can be solely attributed to sulfur dioxide and can be used to quantitate sulfur dioxide.
D
3.3 Determination of UV Absorbance of Sulfur Dioxide in Chlorination Sample
Diluent 1 (50 mM ammonium acetate+1.0% hydrochloric acid in methanol). Prepared by
EP
TE
The optimized sample preparation procedures are as follows:
dissolving 385 mg ammonium acetate in 100mL methanol and adding 1.0mL hydrochloric acid (36~38 wt%). The pH is calculated to be 1.3 based on buffer
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
concentration
Diluent 2 (50 mM ammonium acetate + 1% glacial acetic acid in methanol). Prepared by dissolving 385mg ammonium acetate in 100mL methanol and adding 1.0mL glacial acetic acid. The pH is estimated to be 4.0 based on buffer concentration.
Sodium sulfite stock solution (20 mg mL-1): Accurately weigh about 500 mg sodium sulfite into a 25-mL volumetric flask. Dissolve and dilute to volume with water.
11
ACCEPTED MANUSCRIPT
Sulfur dioxide standard solution (0.04 mg mL-1 sodium sulfite Diluent 1, which
RI PT
corresponds to ~0.02 mg mL-1 sulfur dioxide), prepared by diluting sodium sulfite stock solution 1:500 with Diluent 1
Sulfur dioxide blank solution (0.04 mg mL-1 sodium sulfite Diluent 2, which corresponds
SC
to ~0.02 mg mL-1 sulfur dioxide), prepared by diluting sodium sulfite stock solution 1:500 with Diluent 2
Sample solution: Dilute reaction stream 1:500 with Diluent 1
Sample blank solution: Dilute reaction stream 1:500 with Diluent 2
M AN U
Measure the UV absorbance at 276 nm for sulfur dioxide standard solution (As1), sulfur dioxide blank solution (AS2), sample solution (Au1), and sample blank solution (Au2) using the same
D
cuvette. Rinse the cuvette 3 times using the solution to be measured before filling for UV
EP
Calculations:
TE
absorbance measurement.
SO2 (mg/mL)
Au 1 Au 2 Cs 0.507 DF As 1 As 2
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Where,
Cs=Sodium sulfite standard concentration (mg mL-1) 0.507=Conversion ratio of sodium sulfite to sulfur dioxide DF=Sample dilution factor, which is 500
An overlaid UV spectra of sulfur dioxide blank solution, sulfur dioxide standard solution, sample blank solution and sample solution is shown in Figure 5. For sulfur dioxide blank solution, 0.036 mg mL-1 sodium sulfite is prepared in Diluent 2 (pH4.0), and the absorbance at 276 nm was 12
ACCEPTED MANUSCRIPT
0.041 AU. When the same concentration of sodium sulfite was prepared in Diluent 1 (pH1.3), the
RI PT
absorbance at 276 nm increased to 0.651 AU. For sample blank solution prepared in Diluent 2 (pH4.0), the absorbance at 276 nm was 1.391 AU, which is mainly attributed to the absorbance from the sample matrix. For the same sample solution that was prepared in Diluent 1 (pH1.3),
SC
the absorbance at 276 nm was 1.873 AU, which is attributed to both sample matrix and sulfur dioxide. Using the equation above, we determined that the concentration of sulfur dioxide in the
M AN U
sample solution is 7.21 mg mL-1.
3.4 Limit of Quantitation/Limit of Detection
The quantitation limit and detection limit are estimated based on the standard deviation of a
D
blank solution [12], i.e. the sample solution prepared using pH 4.0 buffer. The detection limit is
TE
estimated based on 3 times of standard deviation of the UV absorbance of blank at 276 nm, which is equivalent to 24µg L-1, while the quantitation limit is estimated based on 10 times of
EP
standard deviation of the blank, which is equivalent to 80µg L-1. The quantitation limit is much lower than the desired sensitivity (10mg L-1 after 1:500 dilution), demonstrating the suitability of
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
this method.
3.5 Linearity and Spiked Recovery/Accuracy The method linearity has been studied for the range of 2 to 40 mg L-1. Both linear and quadratic regression analyses were performed using JMP software (Version 10), and the results were summarized in Table 1. Although the coefficient of determination (R2) from quadratic fit is slightly higher than that from linear fit, the R2 Adjusted from quadratic fit is actually slightly
13
ACCEPTED MANUSCRIPT
lower than that from linear fit. In addition, the 95% confidence interval around the quadratic
RI PT
coefficient ( -0.00058 to 0.00036) includes zero, indicating that the quadratic coefficient is insignificant. Therefore, we concluded that there was a linear relationship between absorbance and concentration. The high R2 value (0.998) demonstated good method linearity for the
SC
concentration range of 2 to 40 mg L-1. This linearity range covers 20% to 400% of our target concentration of 10 mg L-1. However, it is possible to extend the linearity to much lower
M AN U
concentration as the UV absorbance at lower concentration typically obeys Beer’s law better. The recovery was studied by spiking 2 to 40 mg L-1 sulfur dioxide into the fully degassed sample matrix. The recoveries are 105%, 105%, 103%, 102% and 94% for the spiking of 2, 4, 10, 20 and 40 mg L-1 sulfur dioxide, respectively. The results demonstrated good accuracy of this method.
D
Four sample solutions with different degassing times were analyzed using this UV
TE
spectrophotometry method as well as GC-MS and ICP-AES method. The results were
method.
EP
comparable from 3 different techniques as shown in Table 2, demonstrating the accuracy of this
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
3.6 Determination of Sulfur Dioxide during Degassing in Pilot Plant Manufacturing Shown in Figure 6 is the plot of sulfur dioxide concentration vs. degassing time during a recent pilot plant manufacturing of alkyl chloride intermediate. All pilot plant samples were diluted only once in both Diluent 1 and Diluent 2, and the absorbance at 276 nm was measured three times. The average absorbance for each solution was used to calculate the concentration of sulfur dioxide. The standard deviation from triplicate analyses is between 0.01 and 0.18 mg mL-1 for all samples. Accurate and fast quantitation of sulfur dioxide allowed the modeling of the degassing
14
ACCEPTED MANUSCRIPT
parameters, improvement of the degassing efficiency, and significant reduction of the degassing
RI PT
time.
SC
4. Conclusions
A UV spectrophotometry method has been developed to quantitate sulfur dioxide in an alcohol-
M AN U
thionyl chloride reaction mixture. This method has been demonstrated to be specific, sensitive, and accurate, and it can likely be applied to most alcohol-thionyl chloride reactions. It may also be of value in determining sulfur dioxide in foods and wines where matrix interference is expected. Depending on the actual sample matrix, additional development experiment might be required, but the principle described in this article should apply. The method is simple with no
D
special sample treatment, which facilitates the method transfer. In the situation studied here, the
TE
sample matrix absorbance remains constant from pH1.3 to pH4.0, whereas sulfur dioxide exhibits nearly maximum UV absorbance at 276 nm at pH1.3 but essentially no UV absorbance
EP
at 276 nm at pH4.0. However, this methodology can also be applied to other sample matrix that shows constant absorbance only in a narrow range of pHs. Accurate quantitation of sulfur
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
dioxide can still be achieved because the difference in UV absorbance at 276 nm between two pHs is proportional to the concentration of sulfur dioxide for both standard and sample solution. Of course, smaller pH range might result in smaller difference in the UV absorbance of sulfur dioxide at 276 nm, and thus lower sensitivity. In this case, accuracy should be assessed by spiked recovery or using an established method such as ICP-AES to confirm the lack of matrix interference. This method can be easily adapted for process analytical technology (PAT) for real-time monitoring by using some common methodology such as flow injection analysis (FIA) with UV 15
ACCEPTED MANUSCRIPT
absorbance detection. The procedure could be like this: take 100 µL sample, dilute it with
RI PT
Diluent 2 (pH4.0) to 50 mL, then pump the diluted sample solution through a UV absorbance flow cell in the FIA. Record the UV absorbance at 276 nm as Au2. Repeat the step except for diluting the sample in Diluent 1 (pH1.3), and record the UV absorbance at 276 nm as Au1. The
SC
difference between Au1 and Au2 is attributed to the UV absorbance of sulfur dioxide. The same procedure can be used to calibrate the flow cell using sodium sulfite standard. After calibration,
M AN U
the sample can be repeatedly analyzed per the procedures described above at desired interval (e.g. 30 min) until sulfur dioxide is below the target concentration. The implementation of PAT will enhance the degassing efficiency as the vacuum can be applied continuously during sampling, and the real time results will allow the degassing process to be optimized and process
5. Acknowledgement
TE
D
cycle time to be reduced.
EP
The authors would like to thank Don Gauthier and Yang Cao from Merck Research Laboratories for providing chlorination samples and valuable discussions on chlorination reaction, Zhihao Lin
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
from MMD PAT for advices on PAT applications, and thank Paul D. Oram, Kristine L. Cappuccio, and Charles W. Moeder from ADC-API for critical review of manuscript and valuable inputs. Thank Claire Lee and Tiebang Wang from Merck Research Laboratories for providing GC-MS results and ICP-AES results, respectively.
16
ACCEPTED MANUSCRIPT
References:
RI PT
[1]. E. S. Lewis, C. E. Boozer, The Kinetics and Stereochemistry of the Decomposition of Secondary Alkyl Chlorosulfites, J. Am. Chem. Soc. 74 (1952) 308-311.
[2]. C. E. Boozer, E. S. Lewis, The Decomposition of Secondary Alkyl Chlorosulfites. II.
SC
Solvent Effects and Mechanisms, J. Am. Chem. Soc., 75 (1953) 3182-3186.
M AN U
[3]. M. H. H. van Dam, A. S. Lamine, D. Roizard, P. Lochon, and C. Roizard, Selective Sulfur Dioxide Removal Using Organic Solvents, Ind. Eng. Chem. Res., 36 (1997) 4628–4637. [4]. H. J. Kim and Y. K. Kim, Analysis of Free and Total Sulfites in Food by Ion Chromatography with Electrochemical Detection. J. Food Sci. 51 (1986) 1360–1361.
D
[5]. E. S. Saltzman, S. A. Yvon and P. A. Matrai, Low-level atmospheric sulfur dioxide
TE
measurement using HPLC/Fluorescence Detection, J. Atmospheric Chem., 17 (1993) 73-90. [6]. A. R. Bandy, D. C. Thornton and A. R. Driedger III, Airborne measurements of sulfur
EP
dioxide, dimethyl sulfide, carbon disulfide, and carbonyl sulfide by isotope dilution gas chromatography/mass spectrometry, J. Geophys. Res.- Atmos., 98 (1993) 23423-23433.
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
[7]. K. Irgum, Pararosaniline colorimetric determination of sulfur dioxide stabilized with ethanedial, Anal. Chem., 57 (1985) 1335–1338. [8]. A. Syty, Determination of Sulfur Dioxide by Ultraviolet Absorption Spectrometry, Anal. Chem., 45(1973) 1744-1747 [9]. P. H. Berben, M. J. Kappers and J. W. Geus, An FT-IR study of adsorption of sulfur dioxide on alpha- and gamma-alumina, Microchimica Acta, 95 (1988) 15-18.
17
ACCEPTED MANUSCRIPT
[10]. Z. Yang, Y. Zhang, Q. Zhang, T. Pei, Z. Meng, Effect of HCl on spectral properties of
RI PT
sulfur dioxide and its derivatives dissolved in water, Procedia Environ. Sci. 18 (2013) 92 – 99. [11]. L. S. Guss , and I. M. Kolthoff, “The Behavior of Sulfur Dioxide as an Acid in Methanol”, J. Am. Chem. Soc., 66 (1944), 1484–1488.
EP
TE
D
M AN U
6.3.1., November 2005, accessed on April 9, 2015.
SC
[12]. ICH guideline Q2(R1) validation of analytical procedures: text and methodology, Section
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
18
ACCEPTED MANUSCRIPT
Figure 1. Reaction scheme of chlorination and alkylation
RI PT
Figure captions:
Figure 2. Overlaid UV spectra of 0.2mg mL-1 sodium sulfite aqueous solutions in 1) 50 mM ammonium acetate in water [pH6.8] – dashed line and 2) 50 mM ammonium acetate and 1%
SC
hydrochloric acid in water [pH1.3] – full line
M AN U
Figure 3. Overlaid UV spectra of 0.092mg mL-1 sulfur dioxide in 1) methanol – full line and 2) methanol with 0.1% triethylamine – dashed line
Figure 4. Overlaid UV spectra of chlorination sample matrix at different buffer pHs. Figure 5. Overlaid UV spectra of sulfur dioxide blank solution, sulfur dioxide standard solution,
D
sample blank solution and sample solution.
TE
Figure 6. Determination of sulfur dioxide in chlorination reaction samples during degassing for
EP
the removal of sulfur dioxide.
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
19
ACCEPTED MANUSCRIPT
Table title:
RI PT
Table 1. Assessment of method linearity
Table 2. Comparison of sulfur dioxide results for reaction samples using UV spectrophotometry,
EP
TE
D
M AN U
SC
GC-MS and ICP-AES
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
20
ACCEPTED MANUSCRIPT
Table 1.
R2 Adjusted 0.997021 0.997009
RI PT
R2 0.997766 0.998504
Equation Y=0.041726X-0.005387 Y=-0.000109X2 + 0.039683X- 0.030010
AC C
EP
TE
D
M AN U
Fit type Linear Quadratic
Absorbance 0.0765 0.1627 0.3884 0.8927 1.6694
SC
Sulfur Dioxide Conc. (mg L-1) 2.0287 4.0573 10.1433 20.2867 40.5734
ACCEPTED MANUSCRIPT
Table 2. GC-MS (mg mL-1) 35 21 18 <1 (NQ)
ICP-AES (mg mL-1) 34 24 20 2
RI PT
UV (mg mL-1) 35 26 22 2
AC C
EP
TE
D
M AN U
SC
Sample Name Sample at T0 Sample degassed for 4.5hrs Sample degassed for 6hrs Sample degassed for 22hrs
RI PT + SO2
M AN U
NMP
SOCl2
O
O R S 1 HO O Impurity
SC
R1 OH
O
R1 Cl
N R3 H
EP
TE
D
NMP, Base
AC C
R2
ACCEPTED MANUSCRIPT
R2
N R3 R1
RI PT
ACCEPTED MANUSCRIPT
SC
1
M AN U D
0.6
0 200
EP
0.2
TE
0.4
AC C
Absorbance (AU)
0.8
220
240
260
280
300
Wavelength (nm)
320
340
RI PT
ACCEPTED MANUSCRIPT
SC
2.1
M AN U
1.5
D
1.2
0.3 0 200
EP
0.6
TE
0.9
AC C
Absorbance (AU)
1.8
220
240
260
280
300
Wavelength (nm)
320
340
RI PT
ACCEPTED MANUSCRIPT
SC
1
M AN U
pH3 pH4 pH5
TE
0.4
0 230
pH6
EP
0.2
pH1.3 pH2
D
0.6
AC C
Absorbance (AU)
0.8
pH1
250
270 Wavelength (nm)
290
310
ACCEPTED MANUSCRIPT
RI PT
3
Sample solution Sample blank solution Sulfur dioxide standard solution Sulfur dioxide blank solution
M AN U
SC
2.5
0.5 0 200
EP
1
TE
D
1.5
AC C
Absorbance
2
250
300 Wavelength (nm)
350
400
RI PT
ACCEPTED MANUSCRIPT
70
SC M AN U
50 40
TE
D
30
EP
20 10 0
0
AC C
SO2 Conc. (mg/mL)
60
10
20
30 40 50 Degassing time (hr)
60
70
ACCEPTED MANUSCRIPT
EP
TE
D
M AN U
SC
RI PT
Simple UV spectrometry method to quantitate sulfur dioxide in complex matrix Excellent sensitivity, linearity and accuracy Potential for real time monitoring using process analytical technology Possible applications in chemical, pharmaceutical, beverage and wine industries
AC C
* * * *
S0003-2670(15)00793-X
DOI:
10.1016/j.aca.2015.05.043
Reference:
ACA 233957
To appear in:
Analytica Chimica Acta
Received Date: 1 March 2015 Revised Date:
25 May 2015
Accepted Date: 27 May 2015
Please cite this article as: J. Zheng, F. Tan, R. Hartman, Simple Spectrophotometry Method for the Determination of Sulfur Dioxide in an Alcohol-Thionyl Chloride Reaction, Analytica Chimica Acta (2015), doi: 10.1016/j.aca.2015.05.043. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
3
TE EP
1.5 1
AC C
Absorbance
2
D
2.5
Sample solution Sample blank solution Sulfur dioxide standard solution Sulfur dioxide blank solution
0.5 0 200
250
300 Wavelength (nm)
350
400
ACCEPTED MANUSCRIPT
Simple Spectrophotometry Method for the Determination of Sulfur Dioxide in an Alcohol-
RI PT
Thionyl Chloride Reaction Jinjian Zheng, Feng Tan, and Robert Hartman
Analytical Development and Commercialization – API, Merck & Co, Inc., Rahway, New Jersey
SC
07065, USA
M AN U
*Corresponding author. Email: [email protected]. Tel: 732-594-4515. Fax: 732-594-
EP
TE
D
3887
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
1
ACCEPTED MANUSCRIPT
Abstract
RI PT
Thionyl chloride is often used to convert alcohols into more reactive alkyl chloride, which can be easily converted to many compounds that are not possible from alcohols directly. One important reaction of alkyl chloride is nucleophilic substitution, which is typically conducted under basic
SC
conditions. Sulfur dioxide, the by-product from alcohol-thionyl chloride reactions, often reacts with alkyl chloride to form a sulfonyl acid impurity, resulting in yield loss. Therefore, the alkyl
M AN U
chloride is typically isolated to remove the by-products including sulfur dioxide. However, in our laboratory, the alkyl chloride formed from alcohol and thionyl chloride was found to be a potential mutagenic impurity, and isolation of this compound would require extensive safety measures. As a result, a flow-through process was developed, and the sulfur dioxide was purged
D
using a combination of vacuum degassing and nitrogen gas sweeping. An analytical method that
TE
can quickly and accurately quantitate residual levels of sulfur dioxide in the reaction mixture is desired for in-process monitoring. We report here a simple ultraviolet (UV) spectrophotometry
EP
method for this measurement.
This method takes advantage of the dramatic change in the UV absorbance of sulfur dioxide with respect to pH, which allows for accurate quantitation of sulfur dioxide in the presence of the
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
strong UV-absorbing matrix. Each sample solution was prepared using 2 different diluents: 1) 50 mM ammonium acetate in methanol + 1% v/v hydrochloric acid, pH 1.3, and 2) 50 mM ammonium acetate in methanol + 1% glacial acetic acid, pH 4.0. The buffer solutions were carefully selected so that the UV absorbance of the sample matrix (excluding sulfur dioxide) at 276 nm remains constant. In the pH 1.3 buffer system, sulfur dioxide shows strong UV absorbance at 276 nm. Therefore, the UV absorbance of sample solution is the sum of sulfur dioxide and sample matrix. While in the pH 4.0 buffer system, sulfur dioxide has negligible UV 2
ACCEPTED MANUSCRIPT
absorbance at 276 nm, and the UV absorbance is attributed only to sample matrix. Quantitation
RI PT
of sulfur dioxide is achieved by subtracting the UV absorbance of sample solution at pH 4.0 from that at pH 1.3. The method is simple but sensitive, with a limit of quantitation of 80µg L-1. The method linearity was demonstrated from 2mg L-1 to 40mg L-1 with an R2 of 0.998, and the
SC
spiked recovery ranges from 94% to 105% within the same range. The results are comparable with those obtained using inductively coupled plasma-atomic emission spectrometry (ICP-AES)
EP
TE
D
M AN U
and gas chromatography-mass spectrometry (GC-MS), suggesting that this method is accurate.
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
3
ACCEPTED MANUSCRIPT
Key Words: Sulfur Dioxide; Sulfite; Spectrophotometry; Alkyl Chloride; Alcohol; Thionyl
EP
TE
D
M AN U
SC
RI PT
Chloride
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
4
ACCEPTED MANUSCRIPT
1. Introduction
RI PT
Alcohol is not reactive for nucleophilic substitution reaction because the hydroxyl leaving group is a strong base and thus not a good leaving group. One common approach to improve the reactivity of alcohol in nucleophilic substitution reactions is to convert the hydroxyl group into a
SC
better leaving group such as a sulfonate ester, phosphite ester, or alkyl halide. Thionyl chloride is commonly used for this conversion since the alkyl chloride product is very reactive, and the
M AN U
inorganic by-products such as sulfur dioxide and hydrogen chloride are gases, which can be purged by sparging with inert gas or degassing under vacuum [1-2]. In a process developed in Merck Research Laboratories, an alcohol is subjected to reaction with thionyl chloride to form an alkyl chloride intermediate (chlorination step), which subsequently
D
reacts with an amide (alkylation step) to form an active pharmaceutical ingredient (API). The
TE
schematic diagram of this synthesis is shown in Figure 1. In the original process, the alkyl chloride intermediate was isolated to remove the solvents and by-products. However, it was later
EP
determined that the alkyl chloride intermediate was a potential mutagenic impurity. Isolation and handling of this intermediate would require extensive safety measures. Therefore, a flow-through process was developed, in which the whole reaction mixture including solvents and by-products
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
from the chlorination step was carried onto the subsequent alkylation step. One issue that arises is that sulfur dioxide reacts with the alkyl chloride intermediate under the basic alkylation conditions, resulting in yield loss and elevated level of sulfonyl acid impurity as shown in Figure 1. As a result, sulfur dioxide has to be purged to below 5mg/mL to achieve satisfactory yield in the next step. The degassing process was very challenging due to the high solubility of sulfur dioxide in the reaction stream, which is mainly composed of the solvent, N-methyl-2-pyrrolidone (NMP) [3]. A combination of heating, sweeping of headspace with nitrogen gas, and application 5
ACCEPTED MANUSCRIPT
of vacuum was required to reduce the sulfur dioxide to the desired level. Due to stability
RI PT
concerns of the alkyl chloride intermediate at elevated temperature, it is preferable to stop the degassing process as soon as the sulfur dioxide is purged to below the desired level. For this purpose, an analytical method to quickly and accurately determine the sulfur dioxide
SC
concentration during degassing was desired.
There are many reports for the analyses of sulfur dioxide, including ion chromatography (IC) [4],
M AN U
fluorescence [5], gas chromatography-mass spectrometry (GC-MS) [6], colorimetric [7], ultraviolet (UV) spectrophotometry [8], and inductively coupled plasma-atomic emission spectrometry (ICP-AES). However, each method has some limitations. Very often, a large amount of organic compounds in the reaction mixture precipitate in the aqueous mobile phase
D
that is used for IC analysis, clogging the column and contaminating the ion suppressor. In GC-
TE
MS analysis, the inlet liner is easily contaminated. Additionally, the GC column deteriorates quickly due to a strongly acidic sample matrix, requiring frequent replacement. Colorimetric
EP
method [7] typically involves lengthy sample treatment such as color development, and is not preferred for in-process control, which requires a fast analysis. The UV spectrophotometry method described by A. Syty [8] cleverly utilizes strong acid to release sulfur dioxide from an
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
absorbing solution, and transfer it to a flow-through absorption cell using a carrier gas for the measurement of UV absorbance at 215 nm. Unfortunately, sulfur dioxide has very good solubility in NMP, the solvent that was used for the chlorination reaction even in the presence of about 1 M hydrochloric acid, the by-product of chlorination reaction. Therefore, this UV spectrophotometry method is not suitable for the samples studied in this article. A common method for fluorescence detection of sulfur dioxide is based on its reaction with ophthalaldehyde and amine in a basic solution to form an isoindole derivative [5]. This method
6
ACCEPTED MANUSCRIPT
has good sensitivity, but often requires extensive method development for this type of highly
RI PT
acidic and complex sample matrix. ICP-AES is very sensitive for measuring the total sulfur in the solution, but the instrumentation can be very expensive and therefore not available in most common analytical laboratories. Fourier transform infrared is commonly used for in-process
SC
monitoring, and works well for sulfur dioxide standard [9]. However, the detection limit for a sulfur dioxide in the process under investigation here is above 20mg mL-1 due to strong matrix
M AN U
interferences, and thus cannot meet the requirements of this analysis.
It is well known that free sulfur dioxide in a solution has strong UV absorbance at 276 nm under acidic conditions [10]. Therefore, it is possible to use UV spectrophotometry for the determination of sulfur dioxide. However, if the sample matrix also shows strong UV absorbance
D
at 276 nm, accurate quantitation of the sulfur dioxide will be extremely difficult.
TE
In this study, we report a simple but sensitive method for the quantitation of sulfur dioxide by taking advantage of the dramatic change in UV absorbance of sulfur dioxide with respect to pH.
EP
This method has been demonstrated to be sensitive and accurate for the quantitation of sulfur dioxide with minimal sample treatment even in the presence of strong background absorbance from reaction mixture. This method can be applied to quantitate sulfur dioxide in most thionyl
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
chloride reaction streams.
2. Experimental 2.1 Apparatus and Reagents All absorption spectral recordings and absorbance measurements were performed on a Cary-300 UV spectrophotometer (Varian, USA) with 1-cm light-path quartz cuvette. The cuvette was
7
ACCEPTED MANUSCRIPT
rinsed 3 times using the solution to be measured before filling up for UV absorbance
RI PT
measurement at the desired wavelength. All chemicals and reagents used were of ACS reagent grade or higher purity. Methanol, trimethylamine, and water were HPLC grade and were purchased from Fisher Scientific
SC
(Pittsburgh, PA, USA). Glacial acetic acid (Certified ACS grade), sodium sulfite (Sigma-Ultra, anhydrous), ammonium acetate (≥99.99% trace metal basis), and hydrochloric acid (ACS reagent,
M AN U
36-38%) were purchased from Sigma-Aldrich (St. Louis, MO, USA).
The alcohol-thionyl
chloride reaction mixture was synthesized by Merck Sharp & Dohme Corp., a subsidiary of
3. Results and Discussions
TE
3.1 Theory
D
Merck & Co. Inc. (Rahway, NJ, USA).
EP
In aqueous solution, sulfur dioxide forms SO2.nH2O complex, which dissociates into bisulfite:
Ka=1.3x10-2 or pKa=1.9
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
In the vicinity of the pKa, the composition of sulfur dioxide/bisulfite changes dramatically with the pH. For example, at pH 4.0, only about 1% of sulfur dioxide exists as free sulfur dioxide while at pH 1.3, about 80% of sulfur dioxide exists as free sulfur dioxide. It is generally agreed that the UV absorbance of a sulfur dioxide solution at 276 nm is due to the free sulfur dioxide dissolved in the solution [10]. Sodium sulfite solution in 50 mM ammonium acetate aqueous solution (pH~6.8) has strong UV absorbance below 220 nm but has almost no UV absorbance at 276 nm as shown in Figure 2. When the same solution was prepared in the 8
ACCEPTED MANUSCRIPT
presence of 1% hydrochloric acid (pH~1.3), it shows strong UV absorbance at 276 nm as shown
RI PT
in the overlaid UV spectra in Figure 2. In our process, the alcohol-thionyl chloride reaction is carried out in NMP. The chlorinated product will precipitate if the sample is diluted with water. Therefore, methanol was used as the
SC
diluent for the preparation of sample solutions as it provides good solubility for both the sample matrix and sulfur dioxide.
M AN U
We have demonstrated that the UV absorbance of sulfur dioxide in methanol changes with respect to pH in a similar way as that in an aqueous solution as shown in Figure 3. When sulfur dioxide is dissolved in methanol, it shows strong UV absorbance at 276 nm. However, the absorbance at 276 nm disappears when 0.1% v/v triethylamine was added to the solution.
D
In sulfur dioxide-methanol solution, majority of sulfur dioxide will react with methanol to form
TE
methyl ester of sulfurous acid in analog to its reaction with water to form sulfurous acid as suggested by Guss and Kolthoff[11]. Addition of TEA base will shift the equilibrium to the right,
EP
causing loss of UV absorbance at 276 nm.
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
In this article, we will focus on how to take advantage of this property of sulfur dioxide to selectively determine sulfur dioxide in the presence of strong UV absorbing organic compounds.
3.2 UV Absorbance of Sample Matrix vs. pH In the previous section, we reported that the UV absorbance of sulfur dioxide was reduced to non-detectable by the addition of 0.1% triethylamine. However, we noticed that the UV spectrum of the sample matrix also changed significantly. Therefore, it is difficult to determine the net absorbance of sulfur dioxide and to quantitate it. 9
ACCEPTED MANUSCRIPT
Due to the complexity of the sample matrix, it is difficult to predict how its absorbance is
RI PT
affected by pH changes. It is also not possible to predetermine the UV absorbance of a sample matrix and subtract it from subsequent sample measurements as it varies with the product concentration, reaction completion rate, time, etc. Therefore, experimental determination of the
SC
UV absorbance for each sample matrix is necessary.
To determine the variation of UV absorbance of the sample matrix with respect to pH changes, a
M AN U
reaction mixture that is free of sulfur dioxide was prepared by vacuum degassing for 40 hours followed by sparging with nitrogen gas for 72 hours. The solution was analyzed using ICP-AES, and was found to contain less than 0.05mg/mL sulfur dioxide, which is well below the target concentration of 5mg mL-1.
D
Ideally, the matrix absorbance should remain constant at 2 different pHs, while the absorbance of
TE
sulfur dioxide should change significantly. Given that sulfurous acid has a pKa of 1.9, the pH was studied in the range of 1.0 to 6.0. The pH values above 6.0 were not studied as sulfurous
EP
acid is expected to be fully ionized to bisulfite and/or sulfite at pH6.0. Since the sample matrix contains acidic by-products (HCl and SO2), ammonium acetate was added to provide buffering capacity between pH 3.0 and 6.0.
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
The sulfur dioxide free reaction mixture was diluted using diluents that contained 50 mM ammonium acetate in methanol with pH adjusted to 1.0, 1.3, 2.0, 3.0, 4.0, 5.0 and 6.0 using hydrochloric acid. The UV spectra of these solutions are shown in Figure 4. The UV spectra are superimposable for pH 1.0~4.0. At pH5.0, the UV spectrum starts to shift, although the UV absorbance between 270nm and 310nm does not change significantly. Therefore, the UV absorbance from the sample matrix is deemed constant as long as the diluent pH is controlled between 1.0 and 4.0. 10
ACCEPTED MANUSCRIPT
Based on the pKa of sulfurous acid (1.9), we can estimate that at pH 4.0, sulfite will not convert
RI PT
to free sulfur dioxide, and thus have no UV absorbance at 276 nm. Therefore, the sample solution prepared using pH 4.0 buffer can be used as a blank because the UV absorbance at 276 nm can be solely attributed to the matrix.
SC
On the other hand, from Figure 2, we demonstrated that sulfur dioxide shows strong absorbance at 276 nm in the presence of 1% hydrochloric acid (pH 1.3). The UV absorbance of the sample at
M AN U
this pH is the sum of UV absorbance from sample matrix and sulfur dioxide. Therefore, the difference in UV absorbance at 276 nm between pH 1.3 solution and pH 4.0 solution can be solely attributed to sulfur dioxide and can be used to quantitate sulfur dioxide.
D
3.3 Determination of UV Absorbance of Sulfur Dioxide in Chlorination Sample
Diluent 1 (50 mM ammonium acetate+1.0% hydrochloric acid in methanol). Prepared by
EP
TE
The optimized sample preparation procedures are as follows:
dissolving 385 mg ammonium acetate in 100mL methanol and adding 1.0mL hydrochloric acid (36~38 wt%). The pH is calculated to be 1.3 based on buffer
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
concentration
Diluent 2 (50 mM ammonium acetate + 1% glacial acetic acid in methanol). Prepared by dissolving 385mg ammonium acetate in 100mL methanol and adding 1.0mL glacial acetic acid. The pH is estimated to be 4.0 based on buffer concentration.
Sodium sulfite stock solution (20 mg mL-1): Accurately weigh about 500 mg sodium sulfite into a 25-mL volumetric flask. Dissolve and dilute to volume with water.
11
ACCEPTED MANUSCRIPT
Sulfur dioxide standard solution (0.04 mg mL-1 sodium sulfite Diluent 1, which
RI PT
corresponds to ~0.02 mg mL-1 sulfur dioxide), prepared by diluting sodium sulfite stock solution 1:500 with Diluent 1
Sulfur dioxide blank solution (0.04 mg mL-1 sodium sulfite Diluent 2, which corresponds
SC
to ~0.02 mg mL-1 sulfur dioxide), prepared by diluting sodium sulfite stock solution 1:500 with Diluent 2
Sample solution: Dilute reaction stream 1:500 with Diluent 1
Sample blank solution: Dilute reaction stream 1:500 with Diluent 2
M AN U
Measure the UV absorbance at 276 nm for sulfur dioxide standard solution (As1), sulfur dioxide blank solution (AS2), sample solution (Au1), and sample blank solution (Au2) using the same
D
cuvette. Rinse the cuvette 3 times using the solution to be measured before filling for UV
EP
Calculations:
TE
absorbance measurement.
SO2 (mg/mL)
Au 1 Au 2 Cs 0.507 DF As 1 As 2
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Where,
Cs=Sodium sulfite standard concentration (mg mL-1) 0.507=Conversion ratio of sodium sulfite to sulfur dioxide DF=Sample dilution factor, which is 500
An overlaid UV spectra of sulfur dioxide blank solution, sulfur dioxide standard solution, sample blank solution and sample solution is shown in Figure 5. For sulfur dioxide blank solution, 0.036 mg mL-1 sodium sulfite is prepared in Diluent 2 (pH4.0), and the absorbance at 276 nm was 12
ACCEPTED MANUSCRIPT
0.041 AU. When the same concentration of sodium sulfite was prepared in Diluent 1 (pH1.3), the
RI PT
absorbance at 276 nm increased to 0.651 AU. For sample blank solution prepared in Diluent 2 (pH4.0), the absorbance at 276 nm was 1.391 AU, which is mainly attributed to the absorbance from the sample matrix. For the same sample solution that was prepared in Diluent 1 (pH1.3),
SC
the absorbance at 276 nm was 1.873 AU, which is attributed to both sample matrix and sulfur dioxide. Using the equation above, we determined that the concentration of sulfur dioxide in the
M AN U
sample solution is 7.21 mg mL-1.
3.4 Limit of Quantitation/Limit of Detection
The quantitation limit and detection limit are estimated based on the standard deviation of a
D
blank solution [12], i.e. the sample solution prepared using pH 4.0 buffer. The detection limit is
TE
estimated based on 3 times of standard deviation of the UV absorbance of blank at 276 nm, which is equivalent to 24µg L-1, while the quantitation limit is estimated based on 10 times of
EP
standard deviation of the blank, which is equivalent to 80µg L-1. The quantitation limit is much lower than the desired sensitivity (10mg L-1 after 1:500 dilution), demonstrating the suitability of
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
this method.
3.5 Linearity and Spiked Recovery/Accuracy The method linearity has been studied for the range of 2 to 40 mg L-1. Both linear and quadratic regression analyses were performed using JMP software (Version 10), and the results were summarized in Table 1. Although the coefficient of determination (R2) from quadratic fit is slightly higher than that from linear fit, the R2 Adjusted from quadratic fit is actually slightly
13
ACCEPTED MANUSCRIPT
lower than that from linear fit. In addition, the 95% confidence interval around the quadratic
RI PT
coefficient ( -0.00058 to 0.00036) includes zero, indicating that the quadratic coefficient is insignificant. Therefore, we concluded that there was a linear relationship between absorbance and concentration. The high R2 value (0.998) demonstated good method linearity for the
SC
concentration range of 2 to 40 mg L-1. This linearity range covers 20% to 400% of our target concentration of 10 mg L-1. However, it is possible to extend the linearity to much lower
M AN U
concentration as the UV absorbance at lower concentration typically obeys Beer’s law better. The recovery was studied by spiking 2 to 40 mg L-1 sulfur dioxide into the fully degassed sample matrix. The recoveries are 105%, 105%, 103%, 102% and 94% for the spiking of 2, 4, 10, 20 and 40 mg L-1 sulfur dioxide, respectively. The results demonstrated good accuracy of this method.
D
Four sample solutions with different degassing times were analyzed using this UV
TE
spectrophotometry method as well as GC-MS and ICP-AES method. The results were
method.
EP
comparable from 3 different techniques as shown in Table 2, demonstrating the accuracy of this
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
3.6 Determination of Sulfur Dioxide during Degassing in Pilot Plant Manufacturing Shown in Figure 6 is the plot of sulfur dioxide concentration vs. degassing time during a recent pilot plant manufacturing of alkyl chloride intermediate. All pilot plant samples were diluted only once in both Diluent 1 and Diluent 2, and the absorbance at 276 nm was measured three times. The average absorbance for each solution was used to calculate the concentration of sulfur dioxide. The standard deviation from triplicate analyses is between 0.01 and 0.18 mg mL-1 for all samples. Accurate and fast quantitation of sulfur dioxide allowed the modeling of the degassing
14
ACCEPTED MANUSCRIPT
parameters, improvement of the degassing efficiency, and significant reduction of the degassing
RI PT
time.
SC
4. Conclusions
A UV spectrophotometry method has been developed to quantitate sulfur dioxide in an alcohol-
M AN U
thionyl chloride reaction mixture. This method has been demonstrated to be specific, sensitive, and accurate, and it can likely be applied to most alcohol-thionyl chloride reactions. It may also be of value in determining sulfur dioxide in foods and wines where matrix interference is expected. Depending on the actual sample matrix, additional development experiment might be required, but the principle described in this article should apply. The method is simple with no
D
special sample treatment, which facilitates the method transfer. In the situation studied here, the
TE
sample matrix absorbance remains constant from pH1.3 to pH4.0, whereas sulfur dioxide exhibits nearly maximum UV absorbance at 276 nm at pH1.3 but essentially no UV absorbance
EP
at 276 nm at pH4.0. However, this methodology can also be applied to other sample matrix that shows constant absorbance only in a narrow range of pHs. Accurate quantitation of sulfur
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
dioxide can still be achieved because the difference in UV absorbance at 276 nm between two pHs is proportional to the concentration of sulfur dioxide for both standard and sample solution. Of course, smaller pH range might result in smaller difference in the UV absorbance of sulfur dioxide at 276 nm, and thus lower sensitivity. In this case, accuracy should be assessed by spiked recovery or using an established method such as ICP-AES to confirm the lack of matrix interference. This method can be easily adapted for process analytical technology (PAT) for real-time monitoring by using some common methodology such as flow injection analysis (FIA) with UV 15
ACCEPTED MANUSCRIPT
absorbance detection. The procedure could be like this: take 100 µL sample, dilute it with
RI PT
Diluent 2 (pH4.0) to 50 mL, then pump the diluted sample solution through a UV absorbance flow cell in the FIA. Record the UV absorbance at 276 nm as Au2. Repeat the step except for diluting the sample in Diluent 1 (pH1.3), and record the UV absorbance at 276 nm as Au1. The
SC
difference between Au1 and Au2 is attributed to the UV absorbance of sulfur dioxide. The same procedure can be used to calibrate the flow cell using sodium sulfite standard. After calibration,
M AN U
the sample can be repeatedly analyzed per the procedures described above at desired interval (e.g. 30 min) until sulfur dioxide is below the target concentration. The implementation of PAT will enhance the degassing efficiency as the vacuum can be applied continuously during sampling, and the real time results will allow the degassing process to be optimized and process
5. Acknowledgement
TE
D
cycle time to be reduced.
EP
The authors would like to thank Don Gauthier and Yang Cao from Merck Research Laboratories for providing chlorination samples and valuable discussions on chlorination reaction, Zhihao Lin
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
from MMD PAT for advices on PAT applications, and thank Paul D. Oram, Kristine L. Cappuccio, and Charles W. Moeder from ADC-API for critical review of manuscript and valuable inputs. Thank Claire Lee and Tiebang Wang from Merck Research Laboratories for providing GC-MS results and ICP-AES results, respectively.
16
ACCEPTED MANUSCRIPT
References:
RI PT
[1]. E. S. Lewis, C. E. Boozer, The Kinetics and Stereochemistry of the Decomposition of Secondary Alkyl Chlorosulfites, J. Am. Chem. Soc. 74 (1952) 308-311.
[2]. C. E. Boozer, E. S. Lewis, The Decomposition of Secondary Alkyl Chlorosulfites. II.
SC
Solvent Effects and Mechanisms, J. Am. Chem. Soc., 75 (1953) 3182-3186.
M AN U
[3]. M. H. H. van Dam, A. S. Lamine, D. Roizard, P. Lochon, and C. Roizard, Selective Sulfur Dioxide Removal Using Organic Solvents, Ind. Eng. Chem. Res., 36 (1997) 4628–4637. [4]. H. J. Kim and Y. K. Kim, Analysis of Free and Total Sulfites in Food by Ion Chromatography with Electrochemical Detection. J. Food Sci. 51 (1986) 1360–1361.
D
[5]. E. S. Saltzman, S. A. Yvon and P. A. Matrai, Low-level atmospheric sulfur dioxide
TE
measurement using HPLC/Fluorescence Detection, J. Atmospheric Chem., 17 (1993) 73-90. [6]. A. R. Bandy, D. C. Thornton and A. R. Driedger III, Airborne measurements of sulfur
EP
dioxide, dimethyl sulfide, carbon disulfide, and carbonyl sulfide by isotope dilution gas chromatography/mass spectrometry, J. Geophys. Res.- Atmos., 98 (1993) 23423-23433.
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
[7]. K. Irgum, Pararosaniline colorimetric determination of sulfur dioxide stabilized with ethanedial, Anal. Chem., 57 (1985) 1335–1338. [8]. A. Syty, Determination of Sulfur Dioxide by Ultraviolet Absorption Spectrometry, Anal. Chem., 45(1973) 1744-1747 [9]. P. H. Berben, M. J. Kappers and J. W. Geus, An FT-IR study of adsorption of sulfur dioxide on alpha- and gamma-alumina, Microchimica Acta, 95 (1988) 15-18.
17
ACCEPTED MANUSCRIPT
[10]. Z. Yang, Y. Zhang, Q. Zhang, T. Pei, Z. Meng, Effect of HCl on spectral properties of
RI PT
sulfur dioxide and its derivatives dissolved in water, Procedia Environ. Sci. 18 (2013) 92 – 99. [11]. L. S. Guss , and I. M. Kolthoff, “The Behavior of Sulfur Dioxide as an Acid in Methanol”, J. Am. Chem. Soc., 66 (1944), 1484–1488.
EP
TE
D
M AN U
6.3.1., November 2005, accessed on April 9, 2015.
SC
[12]. ICH guideline Q2(R1) validation of analytical procedures: text and methodology, Section
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
18
ACCEPTED MANUSCRIPT
Figure 1. Reaction scheme of chlorination and alkylation
RI PT
Figure captions:
Figure 2. Overlaid UV spectra of 0.2mg mL-1 sodium sulfite aqueous solutions in 1) 50 mM ammonium acetate in water [pH6.8] – dashed line and 2) 50 mM ammonium acetate and 1%
SC
hydrochloric acid in water [pH1.3] – full line
M AN U
Figure 3. Overlaid UV spectra of 0.092mg mL-1 sulfur dioxide in 1) methanol – full line and 2) methanol with 0.1% triethylamine – dashed line
Figure 4. Overlaid UV spectra of chlorination sample matrix at different buffer pHs. Figure 5. Overlaid UV spectra of sulfur dioxide blank solution, sulfur dioxide standard solution,
D
sample blank solution and sample solution.
TE
Figure 6. Determination of sulfur dioxide in chlorination reaction samples during degassing for
EP
the removal of sulfur dioxide.
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
19
ACCEPTED MANUSCRIPT
Table title:
RI PT
Table 1. Assessment of method linearity
Table 2. Comparison of sulfur dioxide results for reaction samples using UV spectrophotometry,
EP
TE
D
M AN U
SC
GC-MS and ICP-AES
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
20
ACCEPTED MANUSCRIPT
Table 1.
R2 Adjusted 0.997021 0.997009
RI PT
R2 0.997766 0.998504
Equation Y=0.041726X-0.005387 Y=-0.000109X2 + 0.039683X- 0.030010
AC C
EP
TE
D
M AN U
Fit type Linear Quadratic
Absorbance 0.0765 0.1627 0.3884 0.8927 1.6694
SC
Sulfur Dioxide Conc. (mg L-1) 2.0287 4.0573 10.1433 20.2867 40.5734
ACCEPTED MANUSCRIPT
Table 2. GC-MS (mg mL-1) 35 21 18 <1 (NQ)
ICP-AES (mg mL-1) 34 24 20 2
RI PT
UV (mg mL-1) 35 26 22 2
AC C
EP
TE
D
M AN U
SC
Sample Name Sample at T0 Sample degassed for 4.5hrs Sample degassed for 6hrs Sample degassed for 22hrs
RI PT + SO2
M AN U
NMP
SOCl2
O
O R S 1 HO O Impurity
SC
R1 OH
O
R1 Cl
N R3 H
EP
TE
D
NMP, Base
AC C
R2
ACCEPTED MANUSCRIPT
R2
N R3 R1
RI PT
ACCEPTED MANUSCRIPT
SC
1
M AN U D
0.6
0 200
EP
0.2
TE
0.4
AC C
Absorbance (AU)
0.8
220
240
260
280
300
Wavelength (nm)
320
340
RI PT
ACCEPTED MANUSCRIPT
SC
2.1
M AN U
1.5
D
1.2
0.3 0 200
EP
0.6
TE
0.9
AC C
Absorbance (AU)
1.8
220
240
260
280
300
Wavelength (nm)
320
340
RI PT
ACCEPTED MANUSCRIPT
SC
1
M AN U
pH3 pH4 pH5
TE
0.4
0 230
pH6
EP
0.2
pH1.3 pH2
D
0.6
AC C
Absorbance (AU)
0.8
pH1
250
270 Wavelength (nm)
290
310
ACCEPTED MANUSCRIPT
RI PT
3
Sample solution Sample blank solution Sulfur dioxide standard solution Sulfur dioxide blank solution
M AN U
SC
2.5
0.5 0 200
EP
1
TE
D
1.5
AC C
Absorbance
2
250
300 Wavelength (nm)
350
400
RI PT
ACCEPTED MANUSCRIPT
70
SC M AN U
50 40
TE
D
30
EP
20 10 0
0
AC C
SO2 Conc. (mg/mL)
60
10
20
30 40 50 Degassing time (hr)
60
70
ACCEPTED MANUSCRIPT
EP
TE
D
M AN U
SC
RI PT
Simple UV spectrometry method to quantitate sulfur dioxide in complex matrix Excellent sensitivity, linearity and accuracy Potential for real time monitoring using process analytical technology Possible applications in chemical, pharmaceutical, beverage and wine industries
AC C
* * * *