色谱  2016, Vol. 34 Issue (2): 130-139   PDF (2089 KB)    
扩展功能
加入收藏夹
复制引文信息
加入引用管理器
Email Alert
RSS
本文作者相关文章
杨三东
唐涛
李彤
王风云
郝青丽
光学吸收检测器检测池设计的研究进展(英文)
杨三东1,2, 唐涛1,2, 李彤2 , 王风云1, 郝青丽1    
1. 南京理工大学化工学院, 江苏 南京 210094;
2. 大连依利特分析仪器有限公司, 辽宁 大连 116023
摘要: 光学吸收检测器是液相色谱仪最常用的检测器之一,检测池是此种检测器的核心部分,是获得光穿过样品信息的重要部件,其设计影响检测器的性能。为了提高检测器的信噪比,减小检测池引起的色谱峰展宽,需要设计更长光程、更小体积的检测池,同时还要保证一定的光通量。受加工技术的限制,增大光程、减小池体积和保证光通量通常很难兼顾,这对优化检测池的设计和加工、提高检测器性能提出了挑战。本文以光学吸收检测器的检测原理为基础,从增大光程、减小体积、利用全反射等角度综述了检测池设计的研究进展,同时对部分检测池的设计进行了详细说明。这些不同的设计思路和结构对于检测池的设计以及光学吸收检测器的研发具有指导意义。
关键词: 高效液相色谱仪     光学吸收检测器     检测池     综述    
Research development of designing flow cells for optical absorption detectors
YANG Sandong1,2, TANG Tao1,2, LI Tong2 , WANG Fengyun1, HAO Qingli1    
1. School of Chemical Engineering, Nanjing University of Science and Technology, Nanjing 210094, China;
2. Dalian Elite Analytical Instruments Co., Ltd., Dalian 116023, China
Foundation Item: National Key Scientific Instrument and Equipment Development Projects (Grant No.2012YQ12004401).
Abstract: The optical absorption detector is one of the most commonly used detectors for high performance liquid chromatography (HPLC). As a core part of this kind of detector, the designs of flow cells, where light passes through samples for acquiring samples information, will affect the performance of a detector. In order to enhance the signal to noise ratio of detectors and reduce the bands broadening that come from flow cells, it is necessary to design a flow cell with a longer optical path length and a less cell volume while maintaining the luminous flux. However the limitations of the machining capacity make it difficult to increase the optical path length, reduce the cell volume and keep or increase the luminous flux simultaneously. It is a challenge to optimize the designing and machining of flow cells so as to improve the performance of detectors. This review discusses the development of designing flow cells based on the detection principle in some aspects of increasing the optical path length, reducing the cell volume, taking the advantages of total reflection and so on. At the same time, some of the designs are illustrated in detail. These various ideas and structures are significant references for designing flow cells and developing optical absorption detectors.
Key words: high performance liquid chromatography (HPLC)     optical absorption detector     flow cell     review    

High performance liquid chromatography (HPLC) is one of the most widely used analytical instruments in analytical chemistry. As the “eyes” of HPLC,detectors can turn the information of separated samples into electronic level,which regarded as y axis with time as x axis to form a chromatogram. Operators will obtain the qualitative and quantitative information of samples according to the positions,sizes and shapes of the peaks.

At present,among the common HPLC detectors,optical absorption detectors are the most popular selective detectors due to their wide application,high sensitivity and fair price [1]. Optical absorption detectors such as ultraviolet visible (UV-Vis) detectors and diode array (DAD) detectors possess relatively complex and precision optical systems in which flow cells are located. As the samples absorb light energy in flow cells to express their absorbance information,the designs of flow cells are of importance for the detectors even for the whole instruments.

According to the optical absorption detection principle,the optical path length should be longer to achieve a better signal to noise ratio (S/N ratio) . However,the cell volume should be restricted for minimizing the band broadening of chromatographic peaks. At the same time,it is necessary to maintain an appropriate luminous flux. Limited by the ability of machining,concerning the above three aspects simultaneously the designing of flow cells may be difficult to detector designers.

In this article,the design ideas of flow cells in optical absorption detectors are analyzed on the basis of the detection principle and the development of designs. Meanwhile,the flow cells with different types and materials are introduced in detail.

1 Detection principle and flow cell designing

Spectral resolution is the research focus for spectrophotometers,while sensitivity,stability and repeatability should be taken into account for optical absorption detectors. Samples need to be decanted into the cuvettes with an optical path length of 1 cm in general for spectrophotometers before analyses. As for HPLC detectors,the information of samples flowing into the flow cells is acquired in real time,so it is necessary to design special flow cells for analyzing fluid samples. Now,most of the flow cells in commercialized optical absorption detectors have an optical path length of 1 cm and a volume of about 10 μ L [2]. Although these flow cells are suitable for HPLC using 4.6 mm i. d. columns,band broadening always occurs when they are used in capillary electrophoresis (CE) or μ -LC systems.

The principle of optical absorption detectors is based on the Lambert-Beer’s law as same as that of spectrophotometers [3]:

where A is the absorbance; T is the transmittance; I0 is the incident intensity (cd); I is the transmitted intensity (cd); ε is the molar absorptivity of the sample (L/(mol\5cm)); b is the optical path length in the flow cell (cm); c is the sample concentration (mol/L).

For optical absorption detectors,when the sample is determined,A is proportional to b. The longer the optical path length is,the higher the absorbance is,the higher the sensitivity of the detector obtains and the lower the minimum detectable concentration reaches. However,if the optical path length is too long,the chances of incident light reflection on the cell wall will increase and so will the noise. At the same time,the volume of the cell will be larger,along with extra-column band broadening and column efficiency decreasing [4]. So the main ideas for designing the flow cells are increasing the optical path length,reducing the cell volume,and enlarging the luminous flux.

Fig. 1 shows the basic development trend of flow cells. Earlier designs for flow cells in CE or μ -LC systems were aimed at increasing the optical path length. Then it was found that controlling the cell volume was also essential to detectors,especially to μ -LC systems employing micro columns or capillary columns. The early concept of liquid-core waveguide (LCW) did not get enough attention until the appearance of a new material. In addition,eliminating the refraction effect of mobile phase in flow cells,applying other materials to cell walls and controlling the temperature of flow cells could also be taken into account.

Fig.1 Development trend of flow cells
2 Research progress
2.1 Increasing the optical path length for flow cell designs

Micro separation systems like CE systems and μ -LC systems usually have less sample consumption. In order to minimize the band broadening,on-column detection was a good solution obtained by stripping the polyimide external coating from a segment of a capillary column. This modified part was the detection window [5] as shown in Fig. 2. The optical path length of this flow cell was equal to the inner diameter of the capillary that greatly limited the sensitivity of the detector. After studying this kind of flow cell,Vindevogel et al. [6] considered that not only the inner diameter of the capillary but also the light beam width or the distance between a flow cell and a photocell could also affect the performance of the detectors.

Fig.2 Schematic diagram of the on-column detection cell

In order to improve the sensitivity of the capillary on-column detection,bubbling in the capillary was employed to increase the optical path length [7]. The diameter of bubble was usually 3-5 times to the diameter of the capillary and so was the S/N ratio. Detection limits for simple dipeptides and aromatics were shown to be 1 μ g/mL or 0.01 μ g/mL. Although the sensitivity was enhanced,the resolution and the column efficiency decreased [8].

No matter on-column detection cells or bubble cells,the optical scattering and refraction caused by the curvature of the capillary walls could bring down the light energy and lift the noise. The rectangular cell could solve this problem. Tsuda et al. [9] studied several rectangular cells of different sizes among which the longest optical path length was up to 1 mm. Compared with a 50 μ m on-column detection cell,a 20-fold increase in optical path length was achieved but only a 15-fold enhancement in sensitivity was gained by the rectangular cell. Meanwhile,the flow profile in the rectangular cells,as opposed to the laminar flow profile,greatly reduced the boundary effects.

Besides the rectangular cells,irradiation along the capillary could also solve the problems caused by the curvature of the capillary walls and enlarge the optical path length as well. Xi et al. [10] designed an axial-beam flow cell. As the beam must get through the whole capillary and the samples were injected in plug,the effective optical path length in the capillary would be changed with the eluted situation. When the effective optical path length was 3 mm,a 7-fold improvement of the sensitivity would be obtained comparing to the 50 μ m on-column detection cell. The LODs of acridine and 3-aminoquinoline were 1.9×10-7mol/L and 1.1×10-6mol/L respectively. However,the choice of the mobile phase was limited as total reflection must happen on the inner wall of the capillary.

Chervet et al. [11] developed a Z-shaped flow cell with an optical path length of 3 mm by bending the detected section of the capillary as shown in Fig. 3. Compared with the on-column detection cell,there were enhancements in S/N ratio of up to 6-fold and in signal of up to 14-fold. But a loss in column efficiency was found to be 17% -32% . The LODs (S/N=2) were 0.31 μ g/mL for nucleosides and 0.51 μ g/mL for enkephalins. Although the optical path length had an improvement of 40-fold,the gain in S/N ratio was only 6-fold that was much less than the expectation just considering the increase in path length. This was probably due to the fact that a part of light was guided by the wall of the capillary that resulting in a noise increase. For better results,the light path needed to be optimized.

Fig.3 Schematic diagram of the Z-shaped flow cell with 3 mm optical path length

Wang et al. [12] designed a kind of multireflection cell for extending the optical path length as shown in Fig. 4. It was made from a 75 μ m i. d. capillary by burning off a part of the protected coating on the capillary. Then a silver layer was deposited on the part by silver mirror reaction to increase the reflectance. A black paint was applied on the outer wall to protect it. On the condition of an incident angle of 5° and a distance of 1 mm between the entrance and exit of light,the effective optical path length was 44 times longer and the sensitivity was over 40 times higher than the on-column detection cell. The LOD of brilliant green was 3.0×10-7 mol/L. But the linear dynamic range was only 2 orders of magnitude for the light passing through the capillary wall for many times. In addition,there was a higher demand for wavelength monotonicity otherwise a lot of stray light might be introduced resulting in noise increasing.

Fig.4 Schematic diagram of the multireflection cell

After comparing some flow cells of various commercial detectors,Moring et al. [13] proposed a new Z-shaped flow cell using a quartz ball lens to optimize the light path. The new flow cell provided more than 1 order of magnitude improvement in S/N ratio over that of a conventional cell with a smaller sapphire ball lens. The minimum detectable concentrations (MDC,S/N=3) of pyrrolidinone analogs were on the order of 10-8mol/L. The optical path length extended from 0.075 mm to 3 mm while the S/N was 45 times and 12 times higher than the conventional cell with sapphire ball lens and normal Z-shaped flow cell respectively. The MDC of dimethylsulfoxide reached 40 ng/mL. The new Z-shaped flow cell also provided an improvement of linear dynamic range from 3.5 to 4 orders of magnitude. The loss in efficiency was less than 14% owing to the extended cell volume. However,because of the energy loss caused by the bending end,the effective optical path length of this flow cell was only 1.67 mm.

In order to eliminate the influence that came from the bending of Z-shaped flow cell,Kim et al. [14] developed a new type of post-column T-shaped flow cell whose optical path length was also 3 mm as shown in Fig. 5. For the bottom part of the T-shaped flow cell,a groove was made on the top surface of a 3 mm o. d. glass rod and then a 100 μ m i. d. capillary was adhered in it. For the upper part,the end of the capillary column was inserted through a piece of heavy-wall capillary. At last the two parts were put together with epoxy resin. The eluted solution flowed from the middle to the two ends of the microchannel and then flushed by the buffer solution around the glass rod. An 8-fold S/N ratio improvement over on-column detection cell was achieved. The LOD (S/N=2) of 3.2×10-7 mol/L for fluorescein isothiocyanate was obtained. However,this T-shaped flow cell suffered 44% loss in resolution. In addition,this flow cell could easily realize the post-column derivatization by adding an inlet which extended its application. Indeed this flow cell avoided the bending part of the former Z-shaped flow cell,but the gain of the sensitivity was still less than that of the path length. Lim et al. [15] modified the T-shaped flow cell by changing the shape of the inner part from a cylindrical rod to a rectangular cube as shown in Fig. 6. The modification made the shape of the inner part in accordance with the outer part and in consequence decreased the consumption of the auxiliary and minimized the background absorption. A 3-fold enhancement in S/N ratio had been achieved compared to the previous T-shaped flow cell,which was 24 times better than that of the on-column detection cell.

Fig.5 Schematic diagram of the T-shaped flow cell a. whole construction; b. anatomy of the cylindrical rod-shaped assembly.

Fig.6 Schematic diagram of the modified T-shaped flow cell a. whole construction; b. anatomy of the cube-shaped assembly.

Another Z-shaped flow cell with a longer optical path length was designed by Chervet et al. [16, 17] as shown in Fig. 7. This flow cell was made by inserting a capillary of 75 μ m i. d. and 300 μ m o. d. to a stainless steel template directly. Its total volume did not exceed 90 nL and its optical path length was 20 mm. The sensitivity enhancement of 100-500 times could be realized in comparison with on-column detection cell. The LOD (S/N=2) of fluorene reached 3 pg. However this kind of flow cell also had the problems of light energy loss occurring at the bending and noise increasing caused by light getting through the wall of the capillary. Turning the stainless steel template for a certain angle could alleviate these problems but could not eliminate them totally [18].

Fig.7 Schematic diagram of the Z-shaped flow cell with a long optical path length

Tsunoda et al. [19] proposed a long capillary cell (LCC) based on total reflection between the cell wall and the atmosphere. The flow cell was made by drawing a standard borosilicate glass tubing to a capillary with 215 μ m i. d.,360 μ m o. d. and coiling to a coil about 13-14 cm in diameter. And then 160 cm of the coil was cut for use. A tee joint acted as the interface of sample introduction system,light and LCC,as shown in Fig. 8. L1/2,meaning the cell length when the energy of the incident light became half of the original value,was used to evaluate the light transmissivity. The L1/2 of this LCC was 38 cm,while the L1/2 of the mirror reflection cell was only 1.2 cm. When the absorbance of methylene blue solution in 1 cm optical path length equaled 2×10-4,the sensitivity enhancement factor of 220 times was obtained compared with a 1 cm cuvette. Owing to the fact that light transmitted through not only the solvent phase but also the cell wall for many times,the linear range was not satisfactory. The lower the absorbance was,the higher the sensitivity got,and vice versa.

Fig.8 Schematic diagram of the interface of sample introduction system, light and LCC (long capillary cell) a. inlet system of source light and sample; b. outlet system of source light and sample.

Table 1 summarizes some parameters of the above flow cells for increasing optical path length.

Table 1 Some parameters of the flow cells for increasing optical path length
2.2 Reducing the volume for flow cell designs

In order to reduce the extra-column band broadening and increase the sensitivity of LC systems,it was important to minimize the volume of flow cells especially for μ -LC systems [20]. There were some common methods to minimize the volume: 1. reducing the aperture diameter of flow cells while maintaining the optical path length; 2. reducing the optical path length while maintaining the aperture diameter; 3. reducing the aperture diameter and the optical path length simultaneously. For flow cells with the same optical path length,the smaller the cell volume was,the less the peak broadened and the higher the sensitivity acquired. Unfortunately,the above designs for reducing the cell volume were limited by the capacity of machining [21],the ability of focusing light [22] and the short-term noise [3]. At present,the optical path lengths of analytical flow cells for commercial HPLC UV-vis detectors are mostly 10 mm. But the cell volumes based on different ideas are obviously distinct from each other. This may affect the performance of entire detectors. Meanwhile,as the microminiaturization was a recent research hotspot [23, 24, 25, 26, 27],reducing the cell volume can also make a contribution to the instrument microminiaturization.

For reducing the extra-column band broadening,Enami et al. [28] proposed a packed flow cell made of 250 μ m i. d. capillary column instead of normal on-column detection cell. Even if the flow cell was packed with packing materials,UV light could still pass through it and reach the optical sensor. Compared with the normal on-column detection cell,the optical path length of the packed flow cell was less than 250 μ m,but the cell volume was smaller and the sensitivity was higher. The obtained peak height ratios of the packed flow cell to the normal on-column detection cell were 4.1 for Vitamin E-Ac and 6.2 for Vitamin K1. At the same time,thanks to the enrichment effect of the packed material,it was found that the sensitivity increased with the retention time. Nevertheless,the packed flow cell might yield a negative ghost peak that would hide the peak of an analyte with less retention.

Aiello et al. [29] made a scored flow cell from two blocks of stainless steel. A 12 mm long optical channel was machined into one of the mated sides by a high-speed steel tool. The two mated sides were both polished to a smooth surface and applied to assemble the scored flow cell. Both the incident and emergent light were guided by the optical fibers,as shown in Fig. 9. The optical path length was 12 mm,and the cell volume was 30 nL. It was found that increasing the spectral bandwidth could bring down the noise level caused by the limitation of the small aperture. Under the condition of the total absorbance detection (a bandwidth of 181 nm),the LODs of organic carbonyl in air corresponded to 5-8 ng/L.

Fig.9 Schematic diagram of the scored flow cell

According to the Lambert-Beer’s law,when monochromatic light passes through a sample,the absorbance is positively related to the concentration of the sample. Xu et al. [30, 31] found that the absorbed intensity was also positively related to the concentration of the sample when broadband light was used for a thin flow cell. As a consequence,wider broadband light could be applied to increasing incident light to achieve a lower LOD and a higher S/N ratio. Based on the technique of broadband absorption,a super-thin cell was designed as shown in Fig. 10. A modified union joint was served as the place where light crossed the samples with optical fibers guiding the incident and emergent light. The optical path length of the super-thin cell was 40 μ m but the cell volume was only 20 nL. For a 10 μ g/mL KMnO4 solution,the S/N ratio of absorbed intensity determination was about 10. The major advantage of the flow cell was that the S/N ratio could be improved by applying a broadband light source. At the same time,thanks to the optical fibers guiding light,a more flexible arrangement of components in detectors could be realized which was of great benefit to reducing the detector size. However,because the super-thin cell absorption theory was based on the infinitely short optical path length,there was a deviation between the reality and the theory. So the linearity at high concentration was unsatisfactory.

Fig.10 Schematic diagram of the super-thin cell

Wang et al. [32] designed a cross flow cell with a similar structure whose connection was the same as shown in Fig. 10. As a cross joint served as the body of the flow cell,the cell volume was only 3 nL. The cross flow cell could improve the linear dynamic range to 103,minimize the light leakage and acquire a lower level of stray light. The LOD of 5.1×10-8 mol/L for reserpine was observed.

Table 2 compares the former flow cells for reducing the cell volume.

Table 2 Some parameters of the flow cells for reducing the cell volume
2.3 Designs based on the LCW technique

Theoretically,the sensitivity of a detector will be improved along with the decrease of the column diameter. However,after studying two small-bore flow cells in commercial detectors,Doehl et al. [33] found that the improvement of sensitivity had not reached the theoretical value considering the decrease of the column diameter. This was mainly due to the increased noise levels. When the aperture diameter of a flow cell was reduced while the optical path length kept consistent,the signal was not affected but the noise was raised up resulting in the decrease of the S/N ratio. Therefore it was necessary to increase the luminous flux for minimizing the effect of noises. Most of the former designs for reducing the cell volume applied optical fibers to increase the incident intensity. However,the important part for the interaction between light and samples had not been improved essentially.

The concept of LCW was proposed in the 1970s [34]. A borosilicate glass capillary or a fused quartz capillary,whose refractive indexes were 1.51 and 1.45 respectively,was filled with liquids of higher refractive indexes. After incident light entering the capillary,total reflection would take place at the interface between the liquid and the inner wall to diminish the energy loss. As for HPLC,the commonly used mobile phases,such as water and methanol,had lower refractive indexes than the above liquids. So the total reflection could not occur at the interface and the LCW technique might not suitable for this kind of analytical instruments. In 1989,DuPont [35, 36] invented a new material with a lower refractive index than water—Teflon AF,which totally changed the situation to make the LCW technique become a research hotspot in analytical chemistry [37, 38, 39, 40, 41, 42, 43]. The refractive index of Teflon AF-2400 was 1.29,which was lower than most of the commonly used mobile phases. Total reflection could happen at the interface between the liquid and the inner wall if the wall was made from this material. It was significant to realize a high luminous flux with a long path length and a small cell volume. A number of instrument manufacturing corporations applied the technique to design flow cells in detectors for improving performance [44, 45] and applied for some patents [46, 47, 48, 49, 50, 51, 52, 53, 54, 55].

Gooijer et al. [56] designed a flow cell based on the LCW for HPLC. A Teflon AF-2400 tube of 530 μ m o. d. and 280 μ m i. d. was modified to be the flow cell. The optical path length was 90 cm and the cell volume was 55 μ L. A 1 m long quartz optical fiber was installed in the reference channel. An LC separation of three test analytes—pesticides atrazine,diuron and linuron,which were detected at 250 nm—provided LODs of 0.3-0.5 μ g/L. It was 30 to 50-fold better than that obtained with a standard flow cell. The flow cell was significantly suitable for analyzing water samples with pesticide residues at lower concentrations.

Munk [57] invented a flow cell inner coated Teflon AF. The patent disclosed two methods of coating: 1. dissolving Teflon AF in a fluorinert solvent and then filling the flow passage with the solution and drying the flow cell to deposit polymer on the interior wall while rotating the flow cell; 2. inserting a stainless steel mandrel coated by Teflon AF into the flow passage and then heating the assembly to leave Teflon AF adhering to the inner wall. At last the mandrel was removed from the flow passage. Due to the energy loss of light reflecting on the cell wall,conventional flow cells had incident angles of only about 1°. But this flow cell could accept an incident angle of at least about 30°,thereby increasing the luminous flux.

Although there was no energy loss in total reflection flow cells,noise might increase with some light entering the end cross-section of the cell wall when the diameter of incident light was larger than the aperture. It was very difficult to make the diameter of incident light smaller than a quite small aperture. Gilby [58] applied “carbon-doped Teflon AF” or “black Teflon AF” as the new material of the cell wall. Light entering the cell wall could be absorbed by the dopant in these materials resulting in less stray light and lower noise.

2.4 Other designs for flow cells

Considering a gradient elution by HPLC,the changes of solvent composition,flow rate,temperature and distinction between samples and solvents will induce the phenomenon of “liquid prism” which can bend the rays in flow cells and cut down the S/N ratios.

For the sake of eliminating “liquid prism”,pyramid and trapezium flow cells were proposed [59, 60]. Gilby et al. [60] invented a kind of trapezium flow cell,whose ends were of different shapes. One of the ends was oval or rectangular and the other was circular as shown in Fig. 11. In general,the flow cell is positioned after a slit in a UV-Vis detector or before a slit in a DAD detector. So as to increase the light intensity and S/N ratio,light could enter the flow cell from the oval or rectangular end and exit from the circular end for a UV-Vis detector while enter and exit from the opposite ends for a DAD detector. Meanwhile,since the diameter of the circle was between the long axis and the short axis of the oval or rectangular,the phenomenon of “liquid prism” could be partly avoided whatever the incident direction.

Fig.11 End surface schematic of the trapezium flow cell

Brooker [61] studied the effect of temperature to the stability and sensitivity of HPLC UV detectors. The stability of the detector could be markedly improved by introducing a heat exchange device to control the temperatures of the flow cell and the liquid in it. In the presence of 0.25 s time constant,the short-term noise and the drift were reduced from 0.000 2 A/min and 0.002 A/h to 0.000 04 A/min and 0.000 04 A/h,respectively.

Abbas et al. [62] designed a double bend capillary flow cell which could detect the refractive index,absorbance and fluorescence intensity of effluents simultaneously as shown in Fig. 12. For the absorbance function,the optical path length was 3 cm and the cell volume was 236 nL. A linearity of 3 orders of magnitude was achieved for Rhodamine 6G. The LOD of it was 2.3×10-8 mol/L. Under a performance criteria of ALOD/b (ALOD was absorbance limit of detection and b was the optical path length),a 12-fold improvement was obtained compared to Xi’s axial-beam flow cell [10].

Fig.12 Schematic diagram of the double bend capillary flow cell
3 Conclusions and prospects

With analytical samples getting increasingly complex,there is an urgent demand for the performance of detectors. As the core of detectors,flow cells have become a significant breakthrough for improving the performance. Enhancing the sensitivity of detectors and reducing the noise and band broadening are the key issues to be solved for designing flow cells. Increasing the optical path length to enhance the sensitivity and reducing the cell volume to control the band broadening are two important research directions. Along with the development of material technology,the application of new materials,especially Teflon AF will bring new ideas for the designing and enhance the performance of detectors distinctly. At the same time,micromachining technology will also provide more possibilities.

参考文献
[1]Ramni K, Navneet K, Ashutosh U, et al. Int Res J Pharm, 2011, 2(5): 1
[2]Riordon J R. Anal Chem, 2000, 72(18): 483
[3]Ling C J. Chinese Journal of Chromatography, 1985, 2(5): 263
[4]Prüβ A, Kempter C, Gysler J, et al. J Chromatogr A, 2003, 1016: 129  
[5]Yang F J. J High Resolut Chromatogr Chromatogr Comm, 1981, 4: 83  
[6]Vindevogel J, Schuddinck G, Dewaele C, et al. J High Resolut Chromatogr Chromatogr Comm, 1988, 11: 317  
[7]Heiger D N, Kaltenbach P, Sievert H J P. Electrophoresis, 1994, 15: 1234  
[8]LI S H. [MS Dissertation]. Nanjing: Nanjing University of Science and Technology, 2009
[9]Tsuda T, Sweedler J V, Zare R N. Anal Chem, 1990, 62(19): 2149  
[10]Xi X B, Yeung E S. Appl Spectrosc, 1991, 45(7): 1199  
[11]Chervet J P, Soest R E J V, Ursem M. J Chromatogr, 1991, 543: 439  
[12]Wang T S, Aiken J H, Huie C W, et al. Anal Chem, 1991, 63(14): 1372  
[13]Moring S E, Reel R T. Anal Chem, 1993, 65 (23): 3454  
[14]Kim S, Kim W, Hahm J H. J Chromatogr A, 1994, 680: 109  
[15]Lim K, Kim S, Hahn J H. Bull Korean Chem Soc, 2002, 23(2): 295  
[16]Chervet J P, Ursem M, Salzmann J P. J High Resolut Chromatogr, 1989, 12: 278  
[17]Chervet J P. US Patent, 5057216. 1991-10-15
[18]Chervet J P, Van Soest R E J. US Patent, 5423513. 1995-06-13
[19]Tsunoda K, Nomura A, Yamada J, et al. Anal Sci, 1988, 4: 321  
[20]Chervet J P, Ursem M. Anal Chem, 1996, 68(9): 1507  
[21]Dai C S, Liu T, Zhai F, et al. China Patent, 201220611696.8. 2013-4-24
[22]Zhang X H, Chen G L, Shang D F, et al. Journal of Instrumental Analysis, 2007, 26(2): 261
[23]Bui D A, Bomastyk B, Hauser P C. J Sep Sci, 2013, 36: 3152
[24]Sharma S, Plistil A, Simpson R S, et al. J Chromatogr A, 2014, 1327: 80  
[25]Sestak J, Moravcova D, Kahle V. J Chromatogr A, 2015, 1421: 2  
[26]Sharma S, Tolley L T, Tolley H D, et al. J Chromatogr A, 2015, 1421: 38  
[27]Nazario C E, Silva M R, Franco M S, et al. J Chromatogr A, 2015, 1421: 18  
[28]Enami T, Nagae N. Am Lab, 1998(24): 29
[29]Aiello M, Mclaren R. Anal Chem, 2001, 73(6): 1387  
[30]Xu C F, Zhang S Y, Fan H Z, et al. Acta Chimica Sinica, 2000, 58(8): 1022
[31]Shen H L, Xu C F, Huang Z Y, et al. Chinese Journal of Analytical Chemistry, 2003, 31(6): 755
[32]Wang H, Yi E C, Ibarra C A, et al. Analyst, 2000, 125(6): 1061  
[33]Doehl J, Greibrokk T. J Chromatogr Sci, 1987, 25: 99  
[34]Walrafen G E, Stone J. Appl Spectrosc, 1972, 26(6): 585  
[35]Yang M K. J Micro-Nanolith Mem, 2008, 7(3): 033010  
[36]Sun Z H, Zhou W, Xu Z T, et al. Spectrosc Spect Anal, 2011, 31(11): 2881
[37]Waterbury R D, Yao W S, Byrne R H. Anal Chim Acta, 1997, 357: 99  
[38]Yao W S, Byrne R H. Talanta, 1999, 48: 277  
[39]Duggan M P, Mccreedy T, Aylott J W. Analyst, 2003, 128(11): 1336  
[40]Du W B, Fang Q, Fang Z L. Chemical Journal of Chinese University, 2004, 25(4): 610
[41]Du W B, Fang Q, He Q H, et al. Anal Chem, 2005, 77(5): 1330  
[42]Villena G, Bejan I, Kurtenbach R, et al. Atmos Meas Tech, 2011, 4(8): 1663  
[43]Yang Q H, Gan W E, Sun H H. Journal of Instrumental Analysis, 2011, 30(2): 199
[44]Deland P. Am Lab, 1997(22): 19
[45]Cohen S, Murphy B J. Am Lab, 1999(6): 28
[46]Febre D L. Europe Patent, 0089157 A1. 1983-09-21
[47]Liu S Y. US Patent, 5444807. 1995-08-22
[48]Gilby A C, Carson W W. US Patent, 5184192. 1993-02-02
[49]Dourdeville T A, Gilby A C. WO Patent, 00/48023. 2000-08-17
[50]Jeannotte A, Gilby A C, Dourdeville T, et al. WO Patent, 2006/023525 A1. 2006-03-02
[51]Jeannotte A, Leveille M J, Denecke F. WO Patent, 2009/143217 A1. 2009-11-26
[52]Garrett M H. Canada Patent, 2355453C. 2009-08-18
[53]Mueller J, Mueller B. US Patent, 7005090B2. 2006-02-28
[54]Kheim T. Europe Patent, 2124036A1. 2009-11-25
[55]Kraiczek K, Mannion J, Killeen K P. US Patent, 20140373606A1. 2014-12-25
[56]Gooijer C, Hoornweg G P, Beer T D, et al. J Chromatogr A, 1998, 824: 1  
[57]Munk M N. US Patent, 5608517. 1997-03-04
[58]Gilby A C. WO Patent, 03/073069 A1. 2003-09-04
[59]Liu Z G, Zhang Y H, Han G F, et al. China Patent, 201010235405.5. 2012-02-01
[60]Gilby A C, Deluca J M. WO Patent, 99/02968. 1999-01-21
[61]Brooker G. Anal Chem, 1971, 43(8): 1095  
[62]Abbas A A, Shelly D C. Anal Chim Acta, 1999, 397: 191