1. Introduction
Each analytical method, including pharmaceutical analysis, requires validation. Validation of analytical procedures is carried out based on the guidelines of the International Conference on Harmonization (ICH) [
1]. One of the parameters when performing method validation is the detection limit. The limit of detection (LOD) of the analyzed compound is usually defined as the lowest quantity or concentration of a component that can be reliably detected with a given analytical method [
2]. There are several methods for determining the limit of detection (LOD), including the signal-to-noise method or by preparing the linear regression [
3].
In the case of the analyzed compounds, namely caffeine, propyphenazone and paracetamol, many methods for their determination have been developed. All three compounds are popular ingredients of pharmaceutical preparations available on the market. They appear both as the sole active ingredient and in complex preparations. For example, there are known caffeine preparations combined with paracetamol or salicylic acid, as well as preparations that contain all three tested compounds, i.e. caffeine, propyphenazone and paracetamol [
4]. Propyphenazone can be combined with ergotamine or allobarbital. However, paracetamol is the most often found in preparations as the only active ingredient, but it also can be combined with tramadol or diphenhydramine. Due to the multitude of pharmaceutical preparations containing the above-mentioned compounds (caffeine, propyphenazone and paracetamol), it seems necessary to propose new methods for their determination, for example, due to the need to control their quality or monitor the production process.
The most common methods used in pharmaceutical analysis are chromatographic methods. It is, for example, micellar liquid chromatography for the analysis of caffeine in anti-inflammatory drugs containing caffeine [
5]. The method was carried out on a C18-SN column and required daily tedious column washing to remove the mobile phase. In turn, caffeine in combination with ibuprofen was analyzed using gas chromatography at high temperatures [
6]. A popular method in chemical, medical or pharmaceutical analysis is high-performance liquid chromatography, and therefore, many analyses of caffeine, paracetamol and propyphenazone are carried out using this method. The analyses concern single compounds or multi-component mixtures, also with substances other than those mentioned [
7,
8,
9,
10,
11,
12,
13,
14,
15,
16]. Most often, these are preparations available on the pharmaceutical market. Some of them were conducted in combination with HPLC and mass spectrometry [
17,
18,
19,
20,
21]. Many analysts, however, tend to use a chromatographic method that is simpler, taking into account, for example, sample preparation. This method is thin-layer chromatography. One of the multi-component caffeine preparations (with codeine and paracetamol) was analyzed using two chromatographic methods - HPLC and TLC [
22]. Both methods gave good results and proved to be useful for both quality control and routine analysis. Egyptian scientists also compared two chromatographic methods [
23]. The assay concerned paracetamol, caffeine and propyphenazone in the presence of two paracetamol impurities - 4-aminophenol and 4-nitrophenol. RP-HPLC and TLC combined with densitometry were used. Both methods of simultaneous determination of the mentioned compounds have been validated. The comparison showed no significant differences. The literature base also includes studies concerning only thin-layer chromatography analysis, the most often combined with densitometric quantitative analysis of compounds such as caffeine, propyphenazone and paracetamol [
24,
25,
26,
27,
28]. The most commonly used stationary phase are plates pre-coated with silica gel with an agent that allows visualization of chromatographic spots under UV light. There are also HPTLC (high-performance thin-layer chromatography) plates that are pre-coated with a thinner layer of a very fine-grained sorbent than in TLC [
29]. It allows for greater sensitivity and resolution of the assay, which in practice should translate into a lower limit of detection value obtained during the determination. An interesting study is the analysis of a four-component preparation with anti-migraine activity (metoclopramide, ergotamine, caffeine, and paracetamol). The green high-performance thin-layer chromatography method was used for that purpose [
30]. Validation parameters were checked in accordance with the International Conference of Harmonization guidelines. The analysis gave promising results. Propyphenazone and caffeine were also determined by thin-layer chromatography combined with densitometry in the presence of three other components, namely ergotamine tartrate and two impurities: phenazone and theophylline [
31]. All compounds have been determined in human plasma and in one of the pharmaceutical preparations. The method was found economical and eco-friendly.
Methods other than chromatography are also important, although used less frequently. For example, all compounds analyzed in our work (caffeine, paracetamol and propyphenazone) were determined simultaneously by two electrochemical methods [
32]. Both methods were based on square-wave voltammetric detection and were statistically compared with each other. No significant differences have been found between them. Paracetamol, propyphenazone and caffeine were also determined using UV spectrophotometry [
33,
34]. An interesting approach was taken by researchers from Bulgaria then. The data obtained from the analysis were used to create chemometric models. The models were tested on an external data set at concentrations within the calibration range. Then, the obtained models were successfully used in the determination of active ingredients in pharmaceutical preparations containing paracetamol, caffeine and propyphenazone.
The aim of our work was to compare the LOD of caffeine, propyphenazone and paracetamol analyzed by thin-layer chromatography (TLC) and high-performance thin-layer chromatography (HPTLC) and to check whether it is possible to develop conditions using the TLC technique that would allow obtaining LOD lower or comparable to results obtained using the HPTLC technique.
3. Results and Discussion
The tables below present a summary of RF and LOD values for the tested compounds obtained by adsorption thin-layer chromatography (
Table 2) and partition thin-layer chromatography (
Table 3), respectively. The LODs presented in this paper are the average LOD values calculated based on the standard deviation of the intercept (s
a) and the residual standard deviation (s
xy).
Both adsorption (NP-TLC) and partition (RP-TLC) thin layer chromatography analyzes were performed, with plates 1.05554, 1.05548 and 1.13728 used for NP-TLC analysis, plates 1.05559 and 1.15498 used for RP-TLC analysis, and plates 1.13726 and 1.05747 were used for both types of analyses. In the case of mobile phase A (acetone - chloroform - ammonia) and chromatographic plates #1.13726, it was impossible to determine the detection limit for propyphenazone. The reason was the fact that under these chromatographic conditions, the analyzed compound migrated with the front of the mobile phase. It was not possible as well to determine detection limits for all compounds analyzed using mobile phase D (methanol-water, 25:25, v/v) and #1.15498 chromatographic plates because all compounds remained at the start. Based on
Table 2 and
Table 3, plots were prepared to show the results of the limit of detection for individual tested compounds, namely for caffeine (
Figure 1), propyphenazone (
Figure 2) and paracetamol (
Figure 3) for TLC and HPTLC plates, respectively. When analyzing the LOD values for the conducted analyses, the limit value of the detection limit was assumed to be 0.05 μg/spot, and the possibility of detecting caffeine, propyphenazone and paracetamol was determined based on this value.
For caffeine, in the case of TLC plates, the condition that the LOD is less than 0.05 is met by four chromatographic conditions: chromatographic plates 1.05747 and mobile phases A, B and D, giving the following LOD values, respectively: 0.023, 0.048 and 0.040 μg/spot and chromatographic plates 1.05559 and mobile phase F with a LOD value of 0.019 μg/spot. In the case of HPTLC plates, only two chromatographic conditions met this condition, additionally only for one stationary phase - 1.05548 plates. The detection limit values are then 0.023 (mobile phase A) and 0.010 μg/spot (mobile phase B), respectively.
In the case of propyphenazone, when using HPTLC plates, more (than in the case of TLC plates) chromatographic conditions give an LOD value of less than 0.05 μg/spot. These are 1.05548 plates and mobile phases B (LOD=0.046 μg/spot) and C (LOD=0.039 μg/spot), 1.13728 plates and mobile phases A (LOD=0.048 μg/spot) and C (LOD= 0.026 μg/spot), 1.13726 plates and mobile phases D (LOD=0.046 μg/spot) and F (LOD=0.041 μg/spot) and 1.15498 plates and mobile phases E (LOD=0.035 μg/spot) and F (LOD=0.030 μg/spot). However, in the case of HPTLC plates, there are also chromatographic conditions that give the detection limits values that are definitely unfavorable, being much higher than the others, namely 0.175 and 0.247 μg/spot for plates 1.13728 and 1.13726 and the mobile phase B, respectively. For propyphenazone analyzed using the method TLC, the best LOD values were obtained for 1.0554 plates and mobile phase C - then the LOD was 0.029 μg/spot; for 1.05747 plates - then the LOD was 0.045, 0.030 and 0.041 μg/spot for mobile phases C, E and F, respectively; for the mobile phase 1.05559 and mobile phase F (LOD=0.024 μg/spot).
As in the case of propyphenazone, analysis on HPTLC plates gives more LOD of paracetamol results, which meet the condition that the LOD is less than 0.05 μg/spot. These are 1.05548 plates, mobile phases B and C, LOD 0.030 and 0.037 μg/spot, respectively; plates 1.13728, mobile phases A, B and C, LOD 0.016, 0.025 and 0.035 μg/spot, respectively; plates 1.13726 and mobile phase A, LOD=0.021 μg/spot. For TLC plates, there are only three results: plates 1.05554, mobile phase C, for which the detection limit is 0.016 μg/spot and plates 1.05747 and mobile phases A and C, for which the detection limits are 0.032 and 0.025 μg/spot, respectively. There are also significantly more cases of high LOD values for TLC plates.
The lowest LOD values for individual compounds obtained for caffeine, propyphenazone and paracetamol are 0.01, 0.024 and 0.016 µg/spot, respectively. For caffeine and propyphenazone, the analysis was carried out on TLC plates - 1.05548 (caffeine) and 1.05559 (propyphenazone). For paracetamol, a value of 0.016 µg/spot was obtained for plates TLC (1.13728) and HPTLC (1.05554). The list of the lowest and highest LOD values for caffeine, propyphenazone and paracetamol is shown in
Table 4.
This comparison (
Table 4) shows that in the reversed-phase system (RP), lower LOD values for all tested compounds can be obtained by TLC than by HPTLC. However, using analyses in the normal phase system (NP), similar (of the same order) LOD values are obtained for caffeine, propyphenazone and paracetamol when using both TLC and HPTLC plates. Therefore, for economic reasons, TLC plates should be recommended for the analysis of caffeine, propyphenazone and paracetamol because they are several times cheaper than HPTLC plates.
Moreover, a cluster analysis was performed based on the obtained LOD values of caffeine, propyphenazone and paracetamol. It is illustrated in the next figure (
Figure 4). Due to missing data, the LOD values for mobile phase A and 1.13726 plates (propyphenazone in this case migrates with the front of the mobile phase) and the LOD values for mobile phase D and 1.15498 plates (all analyzed compounds remain at the start during chromatographic analysis) were omitted.
The analysis (
Figure 4) indicates several clearly marked five-element clusters and one three-element cluster. These clusters include specific chromatography conditions for caffeine, propyphenazone, and acetaminophen. Subsequent clusters indicate similarity in LOD values for specific stationary and mobile phases. The similarities are as follows:
- mobile phase A and 1.05554 plates (TLC) and mobile phase E and 1.05559 plates (TLC)
- mobile phase B and 1.05554 plates (TLC) and mobile phase E and 1.13726 plates (HPTLC)
- mobile phase C and 1.05554 plates (TLC) and mobile phase C and 1.13728 plates (HPTLC)
- mobile phase C and 1.05747 plates (TLC), mobile phase A and 1.13728 plates (HPTLC) and mobile phase C and 1.05548 plates (HPTLC)
- mobile phase B and 1.05548 plates (HPTLC) and mobile phase F and 1.05559 plates (TLC)
- mobile phase E and 1.05747 plates (TLC) and mobile phase F and 1.15498 plates (HPTLC).
The above cluster analysis for all compounds shows that many chromatographic conditions give similar detection limits for the compounds tested. Often happens too that similar detection limits are obtained simultaneously on TLC and HPTLC plates, which is confirmed by the summary in
Table 4. This is also confirmed by similarity analyses performed for each mobile phase separately. They are presented in
Figure 5.
Only in the case of the mobile phase D (methanol-water, 25:25, v/v) are similar results shown by analyses using the same type of plates, i.e. TLC (1.05559 and 1.05747). In the case of the remaining mobile phases, i.e. A, B, C, E, F, the highest LOD similarity was always obtained for TLC and HPTLC plates, which is another confirmation that cheaper TLC plates can be used and give values similar to those obtained on HPTLC plates.
Due to the fact that in the pharmaceutical market exist many multi-component preparations containing all the analyzed compounds, using the obtained LOD data, chromatographic conditions were searched for those which would give the best results in LOD values and would allow to determine all components during one analysis. For this purpose, only those values of the detection limit were considered that were less than 0.05 μg/spot, i.e. those that were discussed earlier in terms of the use of chromatographic plates for TLC and HPTLC. A table has been prepared, giving an overview of the chromatographic conditions and the compounds determined (
Table 5). It was found that there is only one possibility for the analysis of all three compounds with such a limitation of the LOD value, namely using HPTLC plates 1.05548 and mobile phase B (n-hexane - acetone - ammonia, 25:25, 0.5, v/v/v).
The most optimal conditions for analyzing all compounds at once would be NP-HPTLC analysis. The detection limit values for caffeine, propyphenazone and paracetamol are 0.010, 0.046 and 0.030 μg/spot, respectively. The RF values obtained under these chromatography conditions are 0.38, 0.75 and 0.57 for caffeine, propyphenazone and paracetamol, respectively. Differences in these values will allow for a good separation of compounds.
However, also good results that are comparable in quality to the above can be obtained on TLC plates:
a) precoated with silica gel 60 F254 (#1.05554) using mobile phase C (chloroform - toluene - ethyl acetate - methanol - 80% acetic acid, 18:18:7.5:6:0.3, v/v). The detection limit values for caffeine, propyphenazone and paracetamol are 0.054, 0.029 and 0.016 μg/spot, respectively. The RF values obtained under these chromatography conditions are 0.47, 0.60 and 0.38 for caffeine, propyphenazone and paracetamol, respectively.
b) precoated with silica gel 60 F254, modified with C18 groups (#1.05559) using a mobile phase F (methanol-water, 40:10, v/v). The detection limit values for caffeine, propyphenazone and paracetamol are 0.019, 0.024 and 0.053 μg/spot, respectively. The RF values obtained under these chromatography conditions are 0.51, 0.47 and 0.74 for caffeine, propyphenazone and paracetamol, respectively.
The overview presented in
Table 5 shows that TLC plates may be particularly useful for the analysis of caffeine for which LOD is less than 0.05 in as many as four chromatographic conditions (plates 1.05747 and mobile phases A, B, and D, and also plates 1.05559 and mobile phase F). Only one chromatographic condition, using HPTLC plates, gave a LOD for caffeine of less than 0.05 (plates 1.05548 and mobile phase B, i.e. n-hexane – acetone – ammonia, 25:25:0.5, v/v/v). The data in
Table 5 also indicate the conditions under which low LOD values can be obtained when testing caffeine, propyphenazone, paracetamol in single-component and two-component caffeine-paracetamol, caffeine-propyphenazone, propyphenazone-paracetamol samples.