Optimization of microplate-based phenol-sulfuric acid method and application to the multi-sample measurements of cellulose nano�bers

Abstract


Introduction
Cellulose nanofibers (CNFs) are renewable materials with unique properties, including thixotropy, high mechanical strength, elastic modulus, gas barrier properties, viscosity, and low thermal expansion.
They have diverse applications, such as in structural and packaging materials, filters, foods, and cosmetics [1,2].However, because of their nanoscale features, ensuring their safety is important for promoting commercialization [3].Quantifying small amounts of CNFs is necessary for assessing their safety.For example, in an ecotoxicity test, it is necessary to measure the concentrations of CNFs in the media and confirm their dispersibility and stability during the test period.
The PSA method [4,5] is a simple and rapid colorimetric method that is widely used to determine the total saccharides (monosaccharides, disaccharides, oligosaccharides, polysaccharides, and their derivatives).Recently, this method has been adapted to quantify CNFs in aqueous solutions [6].The original method, proposed by DuBois et al. [7], used an 80% phenol solution.Hodge and Hofreiter [4] later improved on this by using a 5% phenol solution.The conventional method is carried out in test tubes by combining 1 mL of the sample with 1 mL of 5% phenol solution and 5 mL of concentrated sulfuric acid.To handle a large number of samples and reduce hazardous chemicals, microplate-based PSA methods have been developed [8][9][10][11].However, their analytical procedures vary slightly (Table 1), particularly in terms of the timing of the sulfuric acid reaction and the addition of phenol.In the test-tube-based method, the heat generated by the sulfuric acid reaction promotes the reaction, but in microplate-based methods, which use a smaller amount of solution, the temperature is raised with a heating device.The main difference between Li et al. [10] and the other studies is whether the thermal sulfuric acid reaction was performed after or before the addition of the phenol solution.In the conventional test-tube-based method, phenol was added before concentrated sulfuric acid.However, Rao and Pattabiraman [12] reported that phenol underwent sulfonation in situ and that the phenol-sulfonic acid formed decreased the color intensity for many hexoses and pentoses.They showed that it was preferable to add phenol after the dehydration of carbohydrates by sulfuric acid and after cooling the system.
In this study, we investigated the optimal procedure for the microplate-based PSA method that can be used for the multi-sample measurements of CNFs.Using two types of CNFs and glucose (Glc), we evaluated whether it was better to perform the thermal sulfuric acid reaction after or before the addition of the phenol solution.In addition, when the thermal sulfuric acid reaction was performed before the addition of the phenol solution, we investigated the optimal reaction conditions for the phenol solution, such as the shaking/resting time and reaction temperature.mmol/g and a typical width and length of 3-4 nm and 0.1-1 μm, respectively.Mechanically fibrillated CNFs (MFC) were obtained from Daio Paper Corporation (Tokyo, Japan).The MFC were produced by the mechanical fibrillation of leaf-bleached kraft pulp and had a typical width and length of 20-100 nm and >1 μm, respectively.D(+)-Glc (guaranteed reagent), concentrated sulfuric acid (96.0-98.0%,super special grade), and phenol (guaranteed reagent) were obtained from Fujifilm Wako Pure Chemical Corp.(Osaka, Japan).To obtain a 5% phenol solution, phenol (1000±0.1 mg) was added to ultrapure water, the total weight of the solution was adjusted to 20 g, and the mixture was shaken on a vortex shaker for 30 s.

Basic procedure of PSA method in this study
The volume ratio of the sample, concentrated sulfuric acid, and 5% phenol solution was set to 1:5:1, referring to the conventional test-tube-based PSA method [4,5].Thirty microliters of the sample was added to each well of a 96-well flat-bottom microplate (1-1601-06, Violamo, AS ONE, Osaka, Japan), followed by 150 µL of concentrated sulfuric acid.The microplate was heated at 90 °C and shaken at 600 rpm for 15 min in a thermomixer (ThermoMixer® C, Eppendorf, Hamburg, Germany) and then cooled to room temperature using a block bath (CDB-105, AS ONE, Osaka, Japan) at 4 °C for 2.5 min (the use of a block bath is not essential; see Results and Discussion).The microplate was ultrasonicated for 10 s to remove air bubbles and read on a plate reader (SpectraMax ABS Plus®, Molecular Devices, San Jose, CA, USA) at an absorbance of 490 nm to obtain the background absorbance before the color reaction with phenol.Thirty microliters of 5% phenol solution was added to the microplate, and the mixture was shaken at 600 rpm for 5 min at 25 °C in a thermomixer.After resting the plate for 60 min (or more), the plate was read on the plate reader again to obtain the absorbance after the reaction with phenol.The absorbance due to the color reaction with phenol was obtained from the difference in absorbance before and after the phenol reaction.
Furthermore, it was normalized by subtracting the absorbance of a reference sample (ultrapure water reacted with sulfuric acid and phenol in the same manner).An 8-channel electronic pipette (Eppendorf Xplorer® plus, ref 4861000805, Eppendorf, Hamburg, Germany) was used when concentrated sulfuric acid or 5% phenol was added to the microplate to minimize reaction time differences between wells.Outliers were removed based on the modified Thompson Tau test (95% confidence level).The procedure is presented in Table 1.

Examination of reaction conditions with concentrated sulfuric acid
We evaluated the difference between the case where the thermal sulfuric acid reaction was performed after the addition of phenol solution (i.e., with phenol) and the case where it was performed before the addition of phenol solution (i.e., without phenol).In both cases, the sulfuric acid reaction was performed at 90 °C for 15 min.In addition, the thermal sulfuric acid reaction without phenol at 100 °C for 15 min was examined (Fig. 1).When the thermal sulfuric acid reaction was performed without phenol, the reaction with phenol was performed based on the basic procedure.TOC, MFC, and Glc at concentrations of 0, 12.5, 25, and 50 mg/L (n=16 for 0 mg/L, n=5 for the others) were reacted.

Examination of reaction conditions with phenol
In the case where the thermal sulfuric acid reaction was performed without phenol (following the basic procedure), the optimum reaction conditions of the samples with phenol after the thermal sulfuric acid reaction were investigated using TOC.The cooling conditions after the sulfuric acid reaction, shaking conditions after the addition of phenol, and resting conditions after shaking were evaluated (Fig. 1).For Test 1, the microplate was cooled without a block bath during cooling.Tests 2-4 were performed based on the basic procedure; however, for Test 4, measurements were made not only at resting times of 60 min but also at 90 and 120 min.For Test 5, shaking was continued for a longer period (60, 90, and 120 min) than resting.For Tests 6 and 7, shaking and resting were performed at high temperatures (50 °C or 90 °C).In all tests, 100 mg/L of TOC was added to all 96 wells of the microplate and reacted.The mean and relative standard deviation (i.e., the standard deviation divided by the mean) of the measurements for the 96 wells were obtained.
slope and the coefficient of determination of the relationship between concentration and absorbance due to resting time (0, 30, 60, 90, 120, and 180 min) were evaluated (n=3).

Examination of reaction conditions with concentrated sulfuric acid
Fig. 2 shows the differences in the relationship between the concentration and absorbance depending on the reaction conditions with concentrated sulfuric acid.The thermal sulfuric acid reaction without phenol showed a higher absorbance (slope) than that with phenol at 90 °C.Therefore, it is preferable to perform a thermal sulfuric acid reaction without phenol to obtain higher sensitivity.Furthermore, the reaction without phenol at 90 °C exhibited a slightly higher absorbance (slope) than that at 100 °C, but the differences were not large.

Examination of reaction conditions with phenol
In the case where the thermal sulfuric acid reaction was performed without phenol, the optimum reaction conditions of the samples with phenol after the thermal sulfuric acid reaction were investigated.
For each reaction condition, the mean and relative standard deviation of the absorbances of the measurements for TOC added to all 96 wells are summarized in Table 2.
Comparing Test 1 with Tests 2-4, there was little difference between the microplate cooling methods (with or without the block bath).Although it is supposed that the use of a block bath can reduce the temperature more quickly and more uniformly, the use of a block bath is not considered essential.
According to the results of Test 4, longer resting times resulted in a higher absorbance and lower relative standard deviation.On the other hand, longer shaking times resulted in higher absorbance, but with a higher relative standard deviation (Test 5).Shaking and resting at higher temperatures (50 or 90 °C) resulted in lower absorbance (Tests 6 and 7).Overall, the conditions of Tests 1-4, which were based on the basic procedure, were considered better and simpler.
In the measurements of the TOC, MFC, and Glc under the basic conditions, changes in the relationship between the concentration and absorbance as a function of resting time (0, 30, 60, 90, 120, and 180 min) after shaking with phenol were shown in Fig. 3.It can be seen that the slope and coefficient of determination increased with increasing resting time.A longer resting time will yield better results, but since the coefficient of determination is sufficiently high after approximately 60 min, a resting time of approximately 60 min seems to be a good choice to save time.070 *TOC at 100 mg/L was added to all 96 wells of a microplate, and the thermal sulfuric acid reaction was performed before the addition of the phenol solution following the basic procedure.

Considerations and limitations of the method
The color reaction depends significantly on the measurement conditions.In the basic procedure of this study, the absorbance gradually increased with the resting time after shaking with phenol.Therefore, it is desirable to measure absorbance at a fixed resting time.
An example of the difference in absorbance depending on the location of the 96 wells in the microplate is shown in Fig. 4.There was little difference in absorbance by row in the microplate, but there were differences by column, with a trend toward higher absorbance in column 1 and lower absorbance in column 12.These results can be explained by that concentrated sulfuric acid and 5% phenol were added in sequence from column 1 to column 12 of the microplate using an 8-channel electronic pipette.Therefore, it is necessary to consider the possibility of differences in absorbance among the columns.One solution is to use fewer columns (e.g., only columns 4-9).Another solution is to place the samples to be compared in the same column (e.g., place the samples for calibration and the samples of interest in the same column).Fig. 4. Difference in absorbances depending on the location of the 96 wells in the microplate.The TOC at 100 mg/L was added to all 96 wells of a microplate and they were reacted following the basic procedure (Test 4).Concentrated sulfuric acid and 5% phenol were added in sequence from column 1 to column 12 of the microplate using an 8-channel electronic pipette.
For similar reasons, a direct comparison of the absorbance between different microplates may not necessarily be valid.The results for the reaction at 90 °C without phenol shown in Fig. 2 and the results for a resting time of 60 min shown in Fig. 3 are slightly different, even though they were obtained under the same conditions.Furthermore, in our experience, the absorbance of blank samples tends to increase with time after the opening of the sulfuric acid bottle or after preparing the 5% phenol solution.In such cases, it is advisable to use a new bottle of sulfuric acid or prepare a new 5% phenol solution.
The absorbance per mass concentration (i.e., the slope of the concentration vs. absorbance plot) was slightly different among TOC, MFC, and Glc.When comparing the results obtained under the same conditions using the same microplate, the absorbance per mass concentration of MFC was 1.1 times that of Glc.This difference is reasonable, given that the ratio of the molar mass of Glc (180 g/mol) to the molar mass of a glucose residue in cellulose (162 g/mol) is 1.1.Thus, the absorbance per molar concentration of glucose residues was comparable between the MFC and Glc.In contrast, the absorbance per mass concentration of TOC was 0.75 times that of Glc under the same conditions.Since the TOC had a sodium carboxylate content of 1.3-1.6 mmol/g, approximately 1/4 of the TOC was composed of glucuronic acid residues.In the conventional PSA method, the absorbance per mass concentration of glucuronic acid is less than half that of Glc [13,14].Therefore, the lower sensitivity of TOC was considered to be due to the addition of carboxyl groups.
The standard deviation (σ) of the absorbance of the blank sample (ultrapure water) was approximately 0.01.From this value (σ = 0.01) and the slopes (a) of the concentration vs. absorbance plots, the detection limits (DL) for TOC, MFC, and Glc in the basic procedure were calculated as 6-8 mg/L for a resting time of 60 min and 4-5 mg/L for a resting time of 180 min using the formula DL=3.3σ/a.
The relative standard deviation of the 96-well measurements for 100 mg/L TOC was 0.06-0.07for the resting time of 60 min and ~0.04 for the resting time of 120 min (Table 2).If repeated measurements yield values outside the range of twice the relative standard deviation, it is important to consider the cause (e.g., visually checking the liquid volume and the presence of bubbles).Outliers may occur owing to contamination, improper pipetting, smudges, or scratches on the microplate.Outliers should be omitted using statistical methods.Multiple sample measurements (e.g., n=5) are recommended to identify and exclude outliers.

Conclusions
In this study, we investigated the optimal procedures and conditions for a microplate-based PSA method using CNFs and Glc.It was found that the thermal reaction with sulfuric acid before the addition of phenol resulted in a higher coloration than after the addition of phenol.In the reaction with phenol, which was carried out after the sulfuric acid reaction, a longer resting time was better than raising the temperature or shaking for a longer time.Furthermore, the longer the resting time, the more coloration progressed, and the smaller the variation.A resting time of 60 min or longer was considered optimal.This study showed that the optimized microplate-based PSA method can be used for the multi-sample measurements of CNFs (the detection limit was <8 mg/L).This method could also be applied to other sugars.

Funding statement
This study was based on results obtained from a project, JPNP20009, commissioned by the New Energy and Industrial Technology Development Organization (NEDO).

Fig. 2 .
Fig. 2. Difference in the relationship between concentration and absorbance depending on the reaction conditions with concentrated sulfuric acid.

Fig. 3 .
Fig. 3. Changes in the relationship between concentration and absorbance as a function of resting time after shaking with phenol.a-i.Relationship between concentration and absorbance for TOC, MFC, and Glc at resting times of 0, 60, and 180 minutes.j, k.Slope and coefficient of determination, respectively, of the relationship between concentration and absorbance as a function of resting time.

Table 1
Procedures of PSA methods in microplates

Table 2
Examination of reaction conditions with phenol.