Improvement of the detection limit

In Sec. 3.3.4 the detection limit and related analytical quantities are described in detail. In this section we define the detection limit as the concentration c we get from a signal which is three times as high as the standard deviation of the background (Kaiser, 1965). The signal should be the absorbance which is proportional to the concentration. The mean value of the noise amplitude will be approximately 5 times the standard deviation (Doerffel, 1988). With this assumption we get 

It is common to use the peak absoiption coefficient instead of the integral absorption coefficient. In most cases only small errors are introduced by this simplification. It is evident that the detection limit decreases with the membrane method as I//At/.s - For gases extracted by the membrane the usual rotational vibrational bands vanish if the compound is dissolved in a polymer, since the molecule is no longer able to rotate freely. As a result, one relatively sharp absorption band is observed which has the same integral absorption coefficient as the rotational vibrational absorption band. So, for gases the detection limit is decreased by an additional factor h, (see Table 6.5-3). For the ATR-method the thickness of the sample is the effective thickness multiplied by the number of reflections N. So we get as detection limit for the membrane method 

In practice this relation is only an approximation because of the uncertainty of all of the parameters. Nevertheless it is still an useful estimation. For example CS2 m water should be measured. The ZnSe-IRE with a length of 50 mm, a thickness of 3 mm, and windows with an angle of 45 °, allow 12 reflections in the sample area. It is coated with a PDMS membrane (n w 1.4) of 20 Xm thickness. The band at 1521 cm (= 6.575 pm) with a peak absorption coefficient of 3100 L mol cm is evaluated. With Eq. 6.5-1 and 6.5-3 the pathlength is calculated to be (V 0.31 A = 24.5 pm. Within this spectral range the noise amplitude is measured as 0.001 absorbance units. The enrichment factor/yv//w is 66. 

Therefore the detection limit is calculated to be 1.2 10 mol/L, or to 0.090 mgA- (M = 76.13 g/mol). Wyzgol found experimentally 0.1 mg/L which is in sufficient agreement with our estimation. 

---

Conducting polymers have been studied as potentiometric ion sensors for almost two decades and new sensors are continuously developed. The analytical performance of solid-state ion sensors with conducting polymers as ion-to-electron transducer (solid-contact ISEs) has been significantly improved over the last few years. Of particular interest is the large improvement of the detection limit of such solid-contact ISEs down to the nanomolar level. Further optimization of the solid contacts as well as the ion-selective membranes will most certainly extend the range of practical applications. 

Although IPCR development is typically focused on improvement of the detection limit, the two examples discussed in 3.6.1 have already illustrated how the increase in sensitivity could be used as a means to an end instead of being the sole intended focus. 

In the example of a-human atrial natriuretic peptide (ANP), found at increased plasma levels in patients with heart failure, Numata et al. [70] demonstrated how IPCR sensitivity accelerated conventional assay procedures. For individual treatment of the cardiac patients, a prompt detection of atrial distension by the presence of the ANP marker would be desirable. Common ANP tests, however, take 2-3 days for the quantification of plasma by radiometric or ELISA techniques. With sandwich IPCR, the assay time could be shortened to 5 hours. A good correlation between IPCR and radiometric detection was maintained, combined with an additional improvement of the detection limit to 2 ng/L ANP. The average level of ANP in plasma for 25 patients with heart failure was found to be 117 100 ng/L, significantly higher than the typical level of 20 14 ng/L for healthy subjects. 

Castro et al. [57] studied the influence of various ion-pair agents on the response of diquat and paraquat. The ion-pair LC separation is based on a 0.5-40% acetonitrile gradient in 15 mmol/1 aqueous HFBA. The compounds were analysed in tap water after a Sep-Pak sample pretreatment. The detection limits were 0.9 and 4.7 pg/1 in ESI and 0.1 and 1.8 pg/1 in APCI for diquat and paraquat, respectively. In a subsequent study [58], on-line SPE is performed on EN h-8 disks, after addition of 15 mmol/1 HFBA to the filtered drinking water sample. Detection limits were 50 and 60 ng/1 for diquat and paraquat, respectively. Further improvement of the detection limits to 30 ng/1 for both compounds was achieved by the use of on-line SPE-LC-MS-MS. The intra-day and inter-day precision for diquat were 9.4% and 12.8%, respectively [60]. In another study from the same group [95], an oa-TOF-MS, operated in full-spectram acquisition mode, and a triple-quadrapole instrament, operated in SRM mode, were compared. The detection limits with the triple-quadmpole instmment were at least tenfold better than those obtained with the oa-TOF instmment, i.e., 60 and 3 ng/1 in tap water for paraquat and diquat, respectively. 

On the other hand, the capability of sample preconcentration for instruments such as AAS, ICP-AES, ICP-MS, and so forth was studied [3]. After metal ions were enriched, they were eluted almost simultaneously by inorganic acid at low pH, because of their diffusion in the column is at a disadvantage for improvement of the detection limits. It has been demonstrated that metal ions such as Ca, Cd, Mg, Mn, Pb, and Zn were enriched with a good recovery at a concentration of 10 ppb each in 500 mL of the sample solution. However, the final enriched sample volume eluted from the CCC column was as large as several milliliters, due to longitudinal diffusion of the sample band in the retained stationary phase [1,3]. Additional band spreading occurred in the flow tube when the concentrated solution was eluted with an acid solution for subsequent analysis. 

Measure the enzyme-catalyzed reaction over a 40-sec period (A is measured and compared with standard curve. Section 15.1.5) a longer period of measurement produces a 30-fold improvement of the detection limit. 

The on-line microcolumn technique allowing improvement of the detection limit of atomic absorption and atomic emission spectroscopy by... 

The adsorptive stripping voltammetric technique led to further improvements of the detection limit of germanium in biological material. By using pyrogallol, the detection limit is 0.1 ng/mL with a standard deviation of 13% [47]. 

Poleunis, C., Delcorte, A., Bertrand, P. (2006) Determination of orgaiuc contaminations on Si wafer surfaces by static ToF-SIMS improvement of the detection limit with Ceo primary ions. Applied Surface Science, 252,7258-7261. 

In a recent work, a bienzymatic sensor for the determination of polyphenols was presented. An ITO electrode was modified with multiwalled carbon nanotubes, and the enzymes laccase and tyrosinase were co-entrapped into a chitosan matrix. The resulting biosensor was calibrated at —50 mV (vs. the Ag/AgCl reference electrode) using rosmarinic acid, caffeic acid and galhc acid as the substrates. The new biosensor resulted in a 10.7-fold increase in response sensitivity and a considerable improvement of the detection limit (42 nM for rosmarinic acid). The sensor was used to evaluate the total phenolic content from plant extracts of Salvia officinalis and cultures of Basilicum callus [54]. 

OPH can be integrated with an amperometric transducer to monitor the oxidatimi or reduction current of the hydrolysis products such as 4-nitrophenol and 2-(diisopropylamino) ethanethiol. An improvement of the detection limit more than one order of magnitude for paraoxmi and methyl parathion was achieved... 

Compared to EDS, which uses 10-100 keV electrons, PEXE provides orders-of-magnitude improvement in the detection limits for trace elements. This is a consequence of the much reduced background associated with the deceleration of ions (called bremsstrahlun compared to that generated by the stopping of the electrons, and of the similarity of the cross sections for ioiuzing atoms by ions and electrons. Detailed comparison of PIXE with XRF showed that PDCE should be preferred for the analysis of thin samples, surfrce layers, and samples with limited amounts of materials. XRF is better (or bulk analysis and thick specimens because the somewhat shallow penetration of the ions (e.g., tens of pm for protons) limits the analytical volume in PIXE. 

VPD-TXRF is also a facile technique for interface analysis [4.78, 4.79]. Automated VPD equipment (Fig. 4.16) improves both the detection limit (upper range 10 atoms cm ) and the reliability (by > 50%) of the VPD-TXRF measurement [4.14]. Current research focuses on sample holders [4.80, 4.81] and light-element detection capability [4.82-4.84]. 

The selection of a technique to determine the concentration of a given element is often based on the availability of the instrumentation and the personal preferences of the analytical chemist. As a general rule, AAS is preferred when quantifications of only a few elements are required since it is easy to operate and is relatively inexpensive. A comparison of the detection limits that can be obtained by atomic spectroscopy with various atom reservoirs is contained in Table 8.1. These data show the advantages of individual techniques and also the improvements in detection limits that can be obtained with different atom reservoirs. 

An added benefit of using thin mica is that the mica is optically transparent down to 1200 cm-l and at least partially transparent down to the spectrometer frequency limit at 200 cm-l. This allows spectral subtraction routines to be employed below 1200 cm-l for extraction of spectral information of the adsorbed species from the mica background (11). Secondly, the thin films of mica can be placed in series and this increases the amount of sample probed by the beam which has led to a direct improvement in the detection limit of monolayer and submonolayer coverages on mica surfaces. 

---