MSc Chemistry
Analytical Chemistry
Literature Thesis
Investigating potential modes of detection for
chromatography-on-a-chip devices.
by
Christian Binder
June 2014
Supervisor:
Dr. Michelle Camenzuli
List of tables.
Table 5.1: Part of a table that indicates (with a +) when a colour reaction occurs between an analyte (right) and a prepared reagent (A.1, A.2, etc.). Diazepam and Acetaminophen, for example, give different reaction patterns. Reproduced from ref. [54].
Table 5.2: A number of LOD values of interest for different analytes and tests. Reproduced from ref. [9g].
Table 7.1: Detection limits for UV and C4D detection for a variety of compounds. Reproduced from ref. [13].
Attachment 1: Overview of the Rf for different analytes depending on the solved used in the method. Reproduced from ref. [57][58]. * have an additional spot at Rf 0.68 caused by the reagent.
List of figures.
Figure 1.1: The structures of GHB and GABA. Reproduced from ref. [6a]. Figure 1.2: Heroin and morphine. Reproduced from ref. [6b].
Figure 1.3: Methamphetamine and amphetamine. Reproduced from ref. [6c].
Figure 3.1: Example of a portable mass spectrometer, the Mini 11 mass spectrometer weighing 4kg. Reproduced from ref. [38].
Figure3.2: Schematics for ESI (above) and APCI (below) ionization sources. Reproduced from ref. [44]. Figure 4.1: Structure of the conductivity cell. Reproduced from ref. [17].
Figure 4.2: The working principle of the developed competitive immunoassay method for impedance detection of methamphetamine (MET) concentration. The gold nanoparticles (AuNPs) are introduced into the electro-microchip, forming a bridge that allows for the transport of electrons. Reproduced from ref. [16].
Figure 4.3: Overview of the creation of the immunosandwich for the optical detection of silver films. Reproduced from ref. [50].
Figure 6.1: A detection method making use of a paper interface and Raman spectroscopy. A) shows the structure of the printed paper. B) illustrates the use as a surface swab for sampling while c) shows the ability for concentration of the taken sample by lateral flow. D) is the detection of the dipstick with the Raman probe. Reproduced from ref. [65].
Figure 6.2: A smartphone with accessory that allows fluorometry. Reproduced from ref.[68]. Figure 7.1: A CE and potentiometry detector setup that indicated the potentiometry detector as an ion selective electrode (ISE) . Reproduced from ref.[12].
Figure 7.2: A C4D detector (right) with the sensor on the left. Dimensions (H/L/W) are 105/250/250 mm (3 kg) and 10/28/23 mm respectively. Reproduced from ref. [80].
Figure 7.3: Voltammogram of Morphine [M], Noscapine [N] and Heroin [H], with an overlap of M and H. A graphene nanosheets-glassy carbon modified electrode is used for the acquisition of signal d. Reproduced from ref.[81].
List of abbreviations
6-MAM – Monoacetylmorphine. SERS – Surface enhanced raman spectroscopy. APCI – atmospheric pressure chemical ionisation. SPR – Surface plasmon resonance. AuNPs – Gold nanoparticles. TMB – Tetramethylbenzidine. BSA – Bovine serum albumin. TLC – Thin layer chromatography. CE – Capillary electrophoresis. UV – Ultra violet.
C4D – Capacitive contactless conductivity detection. ELISA – Enzyme-linked immunosorbent assay. ESI – Electrospray ionisation.
FTIR ATR - Fourier-transform infrared attenuated total reflection. GABA - Gamma-aminobutyric acid.
GBL - Gamma-butyrolactone. GC – Gas chromatography.
GHB – Gamma-hydroxybutyric acid.
HPLC – High-performance liquid chromatography. HRP – Horseradish peroxidase.
IS – Internal standard ISE – Ion selective electrode. LC – Liquid chromatography. LIF – Laser induced fluorescence. LOD – Limit of detection.
LOQ – Limit of quantification. M – Molar concentration: mole/l MA/MET – Methamphetamine.
MCE – Microchip capillary electrophoresis. MDA - Methylenedioxyamphetamine MS – Mass Spectrometry.
MTBE - methyl tert-butyl ether.
Summary.
Lab-on-a-chip devices are interesting. They do not just help the world, and scientists, by required less material and chemicals, but also take over tasks such as mixing and diluting. Analysis that previously have been confined to labs can be taken into the field. This can save time, money and lives, especially in countries with poor infrastructure or a general lack of equipment.
In this literature study a variety of detection techniques will be studied with the intention of coupling one with liquid chromatography, ending with a recommendation for the design of a lab-on-a-chip drug testing device. This device has to fulfil certain expectations:
- The size of a mobile phone.
- Portable and able to function without external equipment. - Able to detect different types of drugs and their metabolites.
However for now this device only has to be able to give a positive or negative results. Further testing should still be done in dedicated labs. The detection methods that are reviewed can be grouped into five categories: mass spectrometry (MS), immunoassays, colour tests, spectroscopy and
electroanalytical detection. Things such as separation methods, chemical requirements and chip design, while important, are not covered in the thesis.
Three different drugs are used as keywords, both to narrow down the literature search and for comparison purposes. The three drugs used in the search are gamma-hydroxybutyric acid (GHB), heroin and methamphetamine. On occasion related drugs(such as morphine) are used.
A set of criteria is used to evaluate the found information. The criteria being:
1. Miniaturization potential: Can the method of detection be scaled down to produce a product of the wanted physical dimensions and is it robust enough for field work?
2. Sample versatility: What kind of samples can be used for detection and is the sample pre-treatment something that can be done in the field, either by a person or by the chip. 3. Selectivity: How many different compounds can be detected during one run?
4. Sensitivity: How sensitive is the detection method? What are the parameters such as detection limit, bias and precision?
Criterion two is proven to be of less worth, none of the found techniques are expressly stated to be unusable for a type of sample (blood, urine, hair), as long as it can pass through a liquid
chromatography column.
The first criterion rules out MS and spectroscopy techniques. Neither appears to be available in the desired size or less. Criterion three removes immunoassay and colour tests as options, for both can only detect anticipated compounds. While this removes potentiometric options of electroanalytical detection, amperometry, conductivity and voltammetry detection is still possible. Voltammetry may even be able to identify compounds based on their voltammograms.
Table of contents
Chapter 1: Introduction. ... 1
Chapter 2: Samples and sample preparation. ... 4
Chapter 3: Mass Spectrometry ... 9
Chapter 4: Immunoassay. ... 12
Chapter 5: Colour tests. ... 16
Chapter 6: Spectroscopy ... 19
Chapter 7: Electroanalytical ... 23
Chapter 8: Conclusion. ... 28
References and Literature ... 29
Attachment 1: Overview of Rf values for TLC. ... 35
Chapter 1: Introduction.
In a world that is trying to become greener, lab-on-chip applications are becoming increasingly interesting. Their advantages do not solely lie in a decrease in the costs of materials and chemicals, because they can also be designed to mix and dilute samples and acting as a reactor[1]. Techniques previously confined to labs can, with sufficient research, be brought into the field, saving valuable time. Especially in countries with poor infrastructure or a general lack of equipment, a portable system could give answers that cut costs and save lives.
In this literature study a variety of detection techniques will be studied with the intention of coupling one with liquid chromatography, ending with a recommendation for the design of a lab-on-a-chip drug testing device. This device should be of the same dimensions as a modern mobile phone, able to function without external equipment and ready to detect different types of drugs and their
metabolites. The final product has to be able to, at the very least, give a yes or no when analysing for drugs, further testing should still be done in dedicated labs. The detection methods that will be reviewed can be grouped into five categories: mass spectrometry (MS), immunoassays, colour tests, spectroscopy and electroanalytical detection.
Things such as separation methods, chemical requirements and chip design, while important, are not covered in this thesis.
In the beginning of the study, three different drugs were used as keywords, not only to narrow down the search in the vast sea of information of drug detection, but also to allow for a greater ease in the comparing of those techniques, as long as the found material allows for it. The three drugs used in the search were gamma-hydroxybutyric acid (GHB), heroin and methamphetamine. Sometimes related drugs (such as morphine and amphetamine) will be used if they are relevant and the three initially chosen drugs to not give an equivalent of information.
GHB:
Odourless, colourless and with just a slightly salty taste, it is used as a date-rape drug through addition to beverages, and will lower the victims inhibitions and relax their muscles[2]. Other effects may include short-term amnesia, confusion and respiratory depression.[3]. However that does not mean that GHB is a compound foreign to the human body[4]. GHB itself is a naturally occurring metabolite of gamma-aminobutyric acid (GABA) and can be found in the whole body but primarily the brain, where it affects the rate of serotonin metabolism[4]. Its exact metabolic pathways are not yet understood[4]. Some detection methods may convert the GHB in a sample into
gamma-butyrolactone (GBL)[2][5]. GHB derivatives and precursors can be found performing other functions such as solvents and in the treatment of narcolepsy and alcoholism[4]. See figure 1.1 for the structures of GHB and GABA.
A variety of commercial tests are available for personal safety in addition to laboratory test that more specialised equipment such as UV spectrophotometers or fluorescence detectors[2][4]. However the personal safety tests are intended for drinks while the specialized instruments are not yet in a portable format.
Heroin.
A highly addictive drug that can be metabolized into compounds such as morphine and 6-acetylmorphine, the latter being an interesting marker since it has no natural source[7]. Heroin produces its effect when it arrives in the brain and is converted. Use may create feeling of euphoria, cloud the mind and cause a dry mouth[8]. Regular use will change how the brain functions, creating a tolerance towards heroin[8]. Other than the dependence, chronic users may suffer from issues such as collapsed veins, constipation, liver and kidney diseases in addition to the risks of Hepatitis and HIV infections if needles are shared or toxic contaminants in the heroin[8].
A common screening test is the mixing of a sample with a colour changing reagent, such as Mandelin or Marquis, and observing the colour change[10].However there is the disadvantage that colour tests can react with other compounds, so additional steps are needed for complex samples, such as blood and urine[10]. Other examples of detection methods are chemiluminescence and CE
potentiometry[10][11][12]. As with GHB, there no examples of portable devices for field use. See figure 1.2 for the structures of heroin and morphine.
Figure 1.2: Heroin and morphine. Reproduced from ref. [6b].
Methamphetamine.
Methamphetamine belongs to the amphetamines, a group of synthetic drugs with relatively similar chemical properties[13]. Once taken they will create feelings of pleasure and well being, with the possible side effects of delusions, depression and hallucinations[13]. Other side effects, some after extended use, are: damage the heart, restricted blood flow, involuntary muscle contractions and overheating[14]. Methamphetamine has a hunger suppressing effect and through that may lead to malnourishment. [14]
Potential ways to detect methamphetamine are with GC-MS analysis of nails, CE with conductivity detection, Immunoassays and Immunosensors[13][15][16][17][18]. As with the previous drugs, none of these are stated to exist as a portable device. But at least conductivity measurements and
immunoassays appear to be small enough, MS devices are in general relatively large. See figure 1.3 for the structures of methamphetamine and amphetamine.
Figure 1.3: Methamphetamine and amphetamine. Reproduced from ref. [6c].
To make a sensible recommendation in regards to a detection method for chip-LC, four criteria will be kept in mind. The criteria are:
1. Miniaturization potential: Can the method of detection be scaled down to produce a product of the wanted physical dimensions and is it robust enough for field work? 2. Sample versatility: What kind of samples can be used for detection and is the sample
pre-treatment something that can be done in the field, either by a person or by the chip. 3. Selectivity: How many different compounds can be detected during one run?
4. Sensitivity: How sensitive is the detection method? What are the parameters such as detection limit, bias and precision?
After this introduction a separate chapter will be dedicated towards possible samples (blood, urine, hair, etc.) and the sample preparation required prior to detection. In other words, the second criterion has its own chapter. The other criteria are contained in the detection method chapters, but are not always clear cut paragraphs. Lastly in regards to criterion four, how values such as limit of detection are reported vary depending on the writer and sample. A hair analysis may give drug per weight of hair while a concentration in blood can be weight per volume. For ease of comparison all the values will be translated (based on sample preparation and injection volume) into a molar concentration (M).
Chapter 2: Samples and sample preparation.
There are a wide variety of different sample types that can be encountered, ranging from street samples from drug seizures to the biological, such as hair and urine .Even if narrowed down to those that can be collected with little to no training on the street, at an interview or during a drug bust, we still have a lot of choice. This chapter contains information about various sample types that can be taken from a human body, with varying degrees of required training. Detection methods that start with drugs in powder or tablet form are rarely mentioned in the literature.
Blood.
Blood is a preferred sample for analysis it has a wide variety of analytical methods and a vast amount of reference data, greater than any other type of sample[9a][19]. However the time window for analysis of drugs is fairly narrow, depending on the drugs the window can range from 6-8 hours, for GHB, up to 24-36 hours[9b].
The collection of blood samples requires personnel with at least some medical training[9c]. Both for proper handling and acquisition of the samples, but proper handling if the possibility exists that there are infectious agents present[9c]. This goes for practically all liquid samples taken from a human body[9c]. Chemicals used for disinfection prior to the extraction of blood may have an effect on the samples[9a].
Blood is a complex sample, only a component of it is used for detection with LC-MS methods[20]. The component is serum, blood plasma that that has been rid of clotting proteins. But even then the serum has to be treated for actual injection and detection. The next step is an extraction.
This can be done either with column, such as a Bond Elute Certify that extract (cationic) drugs from the sample that can then be eluted[21].
A different way to extract drugs it to add methyl tert-butyl ether (MTBE) to a volume of samples that is then centrifuged[22]. From the upper phase a fraction is collected and evaporated to dryness using a speed vac[22]. The resulting pellet is dissolved in 3M HCl n-butanol and incubated at 50 °C[22]. The solvent is removed using vacuum concentration after which 1:1 (v/v) acetonitrile:water is added. This mixture is centrifuged and injected into a HPLC-MS[22].
It is also possible to, once a plasma sample has been spiked with internal standard (IS), diluted the sample with a phosphate buffer to load it on a conditioned solid-phase cartridge[23]. Analytes are eluted with a dichloromethane/isopropanol/25% ammonia (80:20:2 v/v/v) mixture[23]. The eluate is then evaporated under a nitrogen stream at 30 °C while HCl/2-propanol (3:1) is added to prevent volatiles (such as amphetamines) from vanishing[23]. Dried residue is dissolved in mobile phase of which a fraction is injected[23].
Urine.
Plentiful and easy to analyse with potentially greater concentrations than can be found in blood and a larger time window from 8-12 hours for GHB, and 2-3 days for other basic drugs[9b].
Urine requires an extraction just as blood does, but not the removal of proteins[21]. A Bond Elute Certify column can be used here as well[21]. The other possible method of extraction is similar to one described for blood[22]. HCl (0.1 M) and ethylacetate are added to a sample that is then shaken vigorously for a few minutes at room temperature and centrifuged[22]. Any IS should be added prior to the extraction[22]. From the upper layer a volume is collected and evaporated until dryness[22]. The residue is dissolved in 3M HCl n-butanol and incubated at 50 °C[22]. Jjust as with the blood, the solvent is removed and the resulting pellet is dissolved in acetonitrile/water of which a fraction is injected[22].
Dilution is the preferred sample pre-treatment for immunoassay methods, and colour tests may just require an adjustment in pH, as far as colour allows it[24][25].
Beverages.
Spiking beverages is a common method of administration for drugs such as GHB and
Flunitrazepam (Rohypnol), this has led to a variety of commercial tests, these however are not as sensitive as laboratory methods[9b]. Drinks that have been tampered with may show changes in colour, cloudiness or opacity, debris either floating or on the bottom of the container, or oily globules suspended in the beverage[9b].
Either liquid samples of the beverages, or residue scraped from the container can be used for detection, depending on the method. [9b]
Saliva.
Compared with blood or urine, saliva is much easier to collect because it only needs an absorptive material, but may still pass on infectious agents and should be handled with care[9c][26]. Ideally the donor shouldn’t smoke, drink or eat for 10 to 15 minutes prior to sampling[9c]. The use of saliva is becoming more interesting because new developments allow the sensitivity that detection in saliva requires[9c]. Unless anxiety, medical conditions or drug use inhibit the production of saliva, an adult may produce between 500 and 1500 ml per day[9c][26]. If needed the production can be stimulated further by chewing gum or rubber, or through acidic sweats or a citric acid solution[26].The latter being much more effective, resulting in a production of 5-10 ml/min while chewing may at best produce 1-3ml/min[26]. Stimulation will give more sample, less variability in the pH of taken samples and lower the intersubject variability in the saliva/plasma ratio[26].
The saliva/plasma (S/P) concentration ratio is a value that is used to extrapolate the concentration of a drug in plasma from the concentration in saliva[9c][26]. The plasma concentration indicates the free drug concentration that can cross the blood-brain barrier and affect the nervous system
drugs[9c][26]. Smoking a drug vs. intravenous injection will give different results, and will also affect how long the drug can be detection in the saliva (up to 24 hours and 30 minutes respectively)[9c]. This also means that rinsing the mouth or brushing teeth may not be able to clear the saliva collected at a later date of drugs.
Samples are taken by using cotton pads to absorb oral fluid that is then squeezed into a glass container as opposed to directing donors to drool into a container[27]. An amount of samples is taken and saturated ammonium carbonate solution (pH 9.3) is added and vortex mixed[27]. Then ethyl acetate:hexane (4:3) is added and again vortexed, followed by stirring. From the organic phase a fraction is taken and is concentrated to dryness under nitrogen at 50 °C[27]. The residue is
resuspended in 250 µl of mobile phase containing IS[27].
Sweat.
Used to a lesser degree than the above mentioned samples. It can collected by wiping a cotton fleece moistened with isopropanol over skin[21].
Hair.
Hair can be used to test for drugs, but not for the same reason that blood, urine or oral fluids would be used. Hair production is far slower, 1 cm per month for head hair, other types of hair can be even slower[9d]. This means that the upper quarter of a 4 cm long hair would give an indication about the drug use 3-4 months ago. A hair taken when a person is suspected of recent drug use may therefor at best function as a baseline for future detection, or show the difference between chronic use and single exposure[9d]. There is however a lot of hair on a head and it means that a practically identical sample can be taken at a later date, depending on how much hair is needed for detection[9d]. It should be noted that different types of hair have different growth patterns and will show different concentrations[9d]. Different individuals can have different drug and metabolite concentration ratios when comparing hair and blood, unfortunately the exact mechanics of how drugs are incorporated in hair are not sufficiently clarified[9d][28].
What makes hair a harder sample to analyse than blood or urine is that while the drugs of interest are stable in the hair, holding the drugs for years under favourable conditions, they have to be released[9d]. The stability of drugs in hair depend on factors such as hair structure, colour and cosmetically treatments[9d]. Releasing the drugs from the hair can be done through extraction with solvents, hydrolysis or by digesting the keratin. The method used to free the drugs for analysis will have to be carefully chosen so it doesn’t influence the concentrations [9d].
The first step in the preparation of a hair sample, is the decontamination. This can be done in several ways:
- With a volume of methylene chloride for 2-5 minutes[29][30].
- With water and acetone (1 minute and two times 1 minute respectively)[31]. - Water, acetone and then hexane, each for 2 minutes[32].
- Washed with 0.1% sodium dodecyl sulphate, deionized water and acetone [33].
- The hair is added to 0.1M phosphate buffer and mixed for 1 minute before the buffer is removed. The hair is then washed with isopropanol and dichloromethane.[23]
- Hair is added to a fritted reservoir that can be fitted with a luer plug, 2-propanol and a disposable stir-bar are added prior to plugging. Samples are incubated at 37 °C while stirring. The plug is removed and the 2-propanol is collected under vacuum in a glass tube. The reservoir is now filled with 0.01M phosphate buffer, plugged and incubated at 37 °C. The buffer is
collected as before with the 2-propanol, the buffer wash is repeated two more times and combined.[34]
After the decontamination, the hairs are dried and cut into fragments, varying in size from <1mm to 2 mm[29][31][32][33].Depending on what is to be analysed, a wash step can be kept for further analysis[32]. The now washed, dried and fragmented hair is prepared further for injection in one of many ways:
- Digestion with NaOH at 75 °C for 40 minutes[29]. After allowing it to H2SO4 and ethyl acetate is added after which the mixture is agitated and centrifuged[29]. The supernatant (organic layer) is dried and reconstituted with an amount of mobile phase, of which a fraction is injected[29]. - Incubated for 18 hours in methanol/acetonitrile/2mM ammonium formate with gentle shaking
at 37 °C, followed by[31]. The liquid phase is collected and the incubation is repeated. Both extracts are united and evaporated under nitrogen to a 0.5 ml volume, after which a part injected[31].
- Enzymatic digestion solution and IS are added and incubated overnight at 37 °C while stirring[34]. After incubation the digest is collected under vacuum in a glass tube and washed with acetate buffer[34]. Additional acetate buffer is added to the tube and the whole is gently stirred for a few seconds[34]. The digest is poured into a SPE cartridge and then washed with water, HCl and methanol before vacuum drying[34]. The analytes are eluted with several times with methylene chloride/2-propanol/ammonium hydroxide[34]. Then MTBSTFA is added and then evaporated to dryness with nitrogen at 55 °C. It is reconstituted with mobile phase, vortexed and centrifuged before being transferred into autosampler microvials for analysis[34]. - Extraction with methanol followed by extraction with 1.4% HCL/methanolic solution, both
under sonication[32]. Each extraction solution is evaporated under nitrogen at 40 °C[32]. If decontamination wash solutions are to be analysed, a volume is evaporated under nitrogen at 40 °C overnight[32]. Residues (both hair and wash) are dissolved in of methanol and diluted with ammonium formate buffer adjusted to pH 3 with formic acid. 10 µl is injected for analysis.[32]. - Hair is extracted with a solution of internal standard in methanol:acetonitrile:20mM ammonium
formate ratio and sonicated at room temperature[33]. The mixture is centrifuged at after which the supernatant is removed and air-dried, the resulting residue is reconstituted in a
H20:Acetonitrile resuspension solvent[33]. The sample is then diluted 100 fold with blank matrix that has the same concentration of IS before injection[33].
Nails.
Comparable to hair, nails are composed of keratin and show the same tendency to incorporate drugs in their structure, however at a slower rate[35].. Drug ratios may be different compared to hair and concentrations lower[35]. And resembling how the type and location of hair has an effect on concentrations in hair, it appears that nail clipping from fingers have higher concentrations than those from toes[9d][36]. Considering the composition, nails should undergo treatments similar to hairs prior to detection.
Chapter 3: Mass Spectrometry
Mass Spectrometry (MS) is a widely used technique that finds us in many area, among them:
environmental monitoring, food regulation, forensics and homeland security[37][38]. It is a label free technique that can be used for the detection of any molecule as long as it can be ionized[39]. It is capable of giving information on the mass, concentration and identity of analytes[39][37]. Optical techniques in comparison require that the analytes expected in the sample, are given detectable properties[39]. For our end product we will focus on Liquid Chromatography coupled to MS (LC-MS). In this section the potential of MS for lab-on-chip applications will be discussed.
Miniaturization of a MS system comes with the usual advantages of portability, less sample use and lowered power consumption. The decrease in the required power also comes from the fact that less power is needed to keep up the vacuum required by the MS. The current disadvantages of
miniaturized MS systems that are available in a portable format, is that they are limited to non-polar and thermally stable organic compounds[40][41].
In a direct comparison of a LC-MS and a LC-chip/MS (for the analysis of steroids), the LOD of the chip system was higher than the conventional system, 10-150*10-12 M and 50-300*10-12 M
respectively[41]. This could indicate that all miniaturized systems suffer from reduced sensitivity. The problem is that there are portable MS systems, but they have not yet reached the dimensions that we prefer (figure 3.1)[38]. There are no references that mention the bottleneck lying either with the ion source or the mass detector, two of the components that together with the mass analyser form the MS system[38][42].
Figure 3.1: Example of a portable mass spectrometer, the Mini 11 mass spectrometer weighing 4kg. Reproduced from ref. [38].
The ion sources commonly used for LC-MS are electrospray ionisation (ESI) and atmospheric-pressure chemical ionisation (APCI), ESI being more sensitive while APCI works better for molecules that do not ionise as readily[9e]. The designs of either ionization source do not seem to require a significant size, if they do then the references remain quiet about them (figure 3.2 for schematics of the designs). The mass detector merely has to be able to record charges that are induced, or produced, when an ionized molecule (or a fragment), passes the detector or strikes it. As before, no mention is made about the detector being an important bottle neck. Neither a faraday cup or an electron
multiplier looks as if a certain degree of bulk is needed[43]. However the third (or middle) part of the MS, the mass analyser has a link between performance and size[42]. In particular quadrupoles are of interest because of their simplicity and innate compactness[42]. Further miniaturization lowers the internal volume and surface area, the lower volume adds robustness to the device, while the reduction in surface area results in less memory effect and through that faster response in concentration changes[42]. It is however stated by S. Boumsellek et al. that “the reduced surface area of the sensor inlet leads to a total signal loss by a factor of n2”, with nbeing the scale down factor of complete charged particle device[42].
All this considered miniaturizing an MS system is possible, and arguably preferable for the MS itself, but not to the degree that the first criteria requires.
Figure3.2: Schematics for ESI (above) and APCI (below) ionization sources. Reproduced from ref. [44]. The selectivity of a Mass Spectrometer is higher than optical-based techniques and is one of the reasons MS has replaced other analytical techniques, although it should be noted that an internal standard is needed to ensure the best accuracy[39][9e]. In many references there is little to no attention spent on the selectivity of MS methods. However additional MS dimensions (such as tandem MS) have an increased selectivity[45].
The possible detection limits depends on the kind of drugs type of sample.
The LOD for GHB found in Serum can range from 2.8*10-7 M to 9*10-6 M[22][46]. Urine has a slightly higher LOD at 1.24*10-5 M[22]. The values found in a method for detection in hair are given in a per hair concentration, having an LOD of 0.5 ng/mg and a linear range from 0.6 to 50ng/mg[29].
Considering the method uses 25 mg of hair and reconstituted the residue in 50 µl mobile phase, the LOD should be about 2.4*10-6 M[29].
Methamphetamines, detected in hair, have an even lower LOD than GHB, around 0.003 ng/mg[31]. With 20 mg of hair in a final volume of 0.5 ml, the inject concentration would be around 8.0*10-8 M. In one instance LOD as low as 0.1-0.75 pg/mg are noted for cocaine, cocaine metabolites and
amphetamines[33].This method uses only 2 mg of hair in 50 µl, and a 100 fold dilution step, translating roughly to an injection LOD concentration of 1.32*10-11 M for cocaine and 2.96*10-11 M for amphetamines at the lowest[33]. Plasma and urine have been reported to have a LOQ (10 times standard deviation of the blank vs. LOD that has 3 times standard deviation) of 20 ng/ml for 3,4-Methylenedioxyamphetamine (MDA) and 10 ng/ml for other amphetamines, 1.12*10-4 M and 7.41*10-5 M respectively[21]. The LOQ for oral fluids and sweat lies around 1-5 ng/ml for several other amphetamines, with no specific being mentioned[21]. Assuming amphetamine, the LOQ in molar concentration would be 7.41*10-6 M to 3.70*10-5 M.
Heroin has been found to have a LOD of 0.26 ng/ml (7,04*10-7 M)in oral fluid, comparable to four amphetamines mentioned in the same article.[27] The metabolic product morphine has a LOD of 0.65 ng/ml (2.28*10-6 M)[27].
Chapter 4: Immunoassay.
In comparison to the mass spectrometers of the previous chapter, Immunoassay based detectors offer a much smaller and cheaper alternative, due to less expensive instrumentation and required time[16][17][18]. There are commercially available portable immunoassay tests, such as pregnancy tests, indicating that a mobile phone sized end product is possible[9f]. Experimental immunoassay methods that are designed with a conductivity measurement cell end up with cell volumes around 0.09 cm3.[17]. Size is less of a problem compared to MS.
The strength and weakness of an Immunoassay lies in the antibodies that are needed for detection. These antibodies must be produced for each analyte and may have poor specificity, sensitivity to the test matrix or a short shelf life[9f]. Immunoassays may also have problems with analytes of low molecular weight[18]. This issue is not just in detection, but in production as well since the target molecule has to be an immunogen and this is not always the case for smaller molecules[18]. A solution to this is the linking of the molecule of interest to a carrier protein[18].
In addition to the right type of antibody, there are applications in need of a reporter. This reporter is the component that results in a detectable signal[9f]. This could be a radioactive label, something that produces chemiluminescence or colloidal particles for an electrochemical detection[9f][16]. It is possible to use an enzyme in the detection. The so called enzyme-linked immunosorbent assay (ELISA) links an enzyme to an antibody (or antigen) that reacts with a colourless substance to
produce a detectable colour[16]. ELISA makes use of a microtitre plate, which may not be the easiest to method for the portable detector[47]. A construction using microchannels however would be more practical. In these kinds of optical methods, the detection can be done with a
spectrophotometer[48].
Electrochemical detection methods may not have the same problems, as can be seen in figure 4.1[17]. The antibodies are immobilized on the electrode and the conductivity only decreases in the presence of the target analyte (in this case methamphetamine) [17]. On the other hand, using the previously mentioned colloidal particles, it is possible to measure the impedance (conductivity-1), where an increase in conductivity is correlated to the amount of colloidal particles that are bound to the electrode[16]. In addition, it is possible to do this in a “competitive assay”, were analyte in the sample and analyte immobilized to the electrode compete for particles (ex. Gold) bound to
antibodies (Figure 4.2) [16]. It is possible to further enhance the sensitivity by increasing the diameter of the colloidal particles. [16]
Figure 4.1: Structure of the conductivity cell. Reproduced from ref. [17].
Figure 4.2: The working principle of the developed competitive immunoassay method for impedance detection of methamphetamine (MET) concentration. The gold nanoparticles (AuNPs) are introduced into the electro-microchip, forming a bridge that allows for the transport of electrons. Reproduced from ref. [16].
The existence of conductivity (impedance) detection suggest that electrochemical detection methods are possible as well[49]. In contrast to the impedance detector, the antibody used for
electrochemical detection is immobilized on the electrode and there is no colloidal metal, but Horseradish peroxidase (HRP) that is coupled to the competitive analyte[49]. When a
tetramethylbenzidine substrate (TMB) is added, the HRP catalyzes the reduction and amplifies the current signal[49].
Slightly resembling the impedance detection, is an optical technique that uses immobilized antibodies and antibodies with gold labelling to form a sandwich around analyte (Figure 4.3).
[50]After the formation silver-ions are reduced onto the gold particles[50].The resulting film of silver can be measured with light-emitting diodes and photodetectors, or even be examined by eye[50].
Figure 4.3: Overview of the creation of the immunosandwich for the optical detection of silver films. Reproduced from ref. [50].
A different optical technique that does not require a reporter component bound to analyte and antibody, is surface plasmon resonance (SPR)[51]. In this method an antibody is immobilized on a transducer surface (gold) and the act of an analyte binding to the antibody changes the biomolecular interactions that take place in the close vicinity of the transducer surface[51].
The detection is done by aiming a polarized beam of light with an appropriate wavelength and at a specific angle, through a prism, at a thin metal film (the transducer)[51][52]. This causes an oscillation in the charge density inside the metal, decreasing the intensity of the reflected
light.[51][52]. The angle of minimum light reflection is called the SPR intensity[51][52]. Once analyte is bound to the antibody connected to the transducer, there will be a change in the properties of the metal, changing the SPR angle[51][52]. In other words, binding analyte to the transducer changes the refractive index[53].
In short, the possible applications of immunoassays are quite varied. But with their selectivity (based on antibody quality) comes the problem that every possible analyte needs its own antibody. And while it is possible to have a method that can detect multiple types, any additional antibody would require time and effort[50]. That the detector is left blind to unexpected drugs could also be considered a disadvantage.
A more environmental concern is that there is little mention in the literature about the regeneration of the immunoassay for multiple use. When it is, it states that analyte immobilization is the better choice (vs. antibody immobilization) and that the antigen-antibody binding determines what kind of
eluent should be used for regeneration[51]. Considering this, it may mean that the regeneration of a system that hold several antigen-antibody bindings becomes more difficult as the number of bindings increases[51]. It is also possible that some regeneration eluents, glycine-HCl and ethanolamine are mentioned, might prove impractical for a field situation[51]. All this will result in either additional maintenance (dedicated to regeneration) in the field or at a different location, or in disposable immunoassays[51].
A last mark against coupling an immunoassay to a LC-chip, is the lack of literature were an immunoassay is preceded by a liquid chromatography step. When immunoassays and LC are mentioned, it is for a comparison of the methods.
As for the detection limits, they are around 1 ng/ml (6.7*10-6 M) for methamphetamine in urine, about 10 pg/ml (3.5*10-11 M) for morphine in urine, 1.3 µg/ml (1.24*10-5 M) for GHB using a method meant for both urine and blood analysis[18][16][48][24]. In the case of methamphetamine also dependant on the concentration of gold-antibody particles that are used in the method[16].
Chapter 5: Colour tests.
Colour tests are analysis methods that make use of a reagent that is added to a sample of interest to produce a colour. Furthermore colour tests (or chemical spot tests) are meant to be observed with the naked eye[54]. This mean that this type of test is inexpensive in instrumentation, use minimal reagents and are simple to perform, meaning that end users need minimal training[9g][54]. The tests can also be done quite quickly, some only taking a few minutes[25]. Although the addition of colour reagents mean that the sample is diluted. In some cases such as GHB, colour tests are available commercially for personal safety[2]. On the other hand this kind of testing lacks accuracy, either because it relies on subjective interpretation of colour that is further influence by other compounds in the sample, drug concentrations, whether the drug is a salt or free base, or an instability in the colour producing complexes[54]. Although there is a system that can be used to describe the results of colour tests by using ten different colours that are combined, if needed, to indicate variations in hue or colour changes over time[9g]. Therefor evaluation using a reference list (with colours and hues) may require checking for more than one colour[9g]. Colour tests may not be precise enough for definitive measurements, but should suffice for screening and preliminary testing[54]. Samples should be handled carefully since parameters such as pH can influence resulting colours[9g]. The tests can be carried out either in clear containers, on white glazed porcelain tiles that allow a better assessment by providing a uniform background or with paper strips[9g][25]. Colour tests designed for drugs work on about 1mg of analyte, either dried or in a solution (usually water)[9g]. A sample, preferably with an identical matrix, that does not contain the compound of interest should be tested as well to confirm that colour changes are due to the compound of interest[9g]. Because reagents can produce colour when in contact with many different analytes, a battery of colour tests should be used (Table 5.1)[54].
Table 5.1: Part of a table that indicates (with a +) when a colour reaction occurs between an analyte (right) and a prepared reagent (A.1, A.2, etc.). Diazepam and Acetaminophen, for example, give different reaction patterns. Reproduced from ref. [54].
Reagent
A.1 A.2 A.3 A.4 A.5 A.6 Acetaminophen - - - + - + Alprazolam - - - - Aspirin - - - + + - Baking soda - - - - Brompheniramine maleate + - - + - - Chlordiazepoxide HCl + - - - - - Chlorpromazine HC. + - - + + + Contac - - - + - - Diazepam - - - - Doxepin HCl + - - + + + Dristan - - - + + + Ephedrine HCl + - - - - - Exedrine - - - + + +
In the cases that there is overlap between analytes resulting from drug use and analytes that can be found naturally in a human body, a test can be designed to compensate for this. For example a test with a detection limit for a GHB that lies at 0.1 mg/ml in urine (9.61*10-4 M), while the physiological concentration that naturally occurs in a human body is around 0.5-2.0 µg/ml (4.8-19.2*10^-6), and therefor will not give a positive result[55]. The limits of detection for compounds of interest depend on the analyte and the colour tests that are used[54]. LODs typically lie between in the µg, with d-methamphetamine HCl (the hydrochloride salt) at 100 µg and Heroin at 20-200 µg with no sample volume stated, see table 5.2 for a small overview[54][9g].
Table 5.2: A number of LOD values of interest for different analytes and tests. Reproduced from ref. [9g].
Colour test Analyte LOD in µg
Mandelin's test Amphetamine-HCL 10 Heroin 20 Metamphetamine 150 Morphine 5 Marquis Metamphetamine-HCl 5 Morphine 5 Amphetamine-HCl 10 Heroin 10
Sodium nitroprusside-acetone Metamphetamine 5-30
But there still is a problem since the compositions of real life samples are unknown, it is possible that other components in the sample influence the test to produce a false result or making it unreadable.
For example a colour test with Froehde reagent would result in a red colouration for amphetamine and a purple-red colour for heroin, this may be difficult to observed[9g].
It is in that case that a chromatography step is of use, and while there are no explicit mentions of a LC-chip with colour testing, thin layer chromatography (TLC) however may prove useful. In some applications TLC separations can occur by essentially the same physical process as LC[9h].
TLC is a low-cost, easy maintenance, analysis technique suitable for the analysis of a large number of samples[56]. The retardation factor (Rf) is a value used in TLC that describes the distance travelled by a compound versus the distance travelled by eluent[57][58]. It can be used to identify unknown analytes in a sample, although standards should be used alongside the sample, in the same system, to improve accuracy, see attachment 1 for an overview of Rf values[9h][57][58]. It should be noted that some reagents used for colour tests and TLC can suffer from poor selectivity or sensitivity[58]. One more reason to perform a battery of tests when analysing samples.
But it is not just the reagents that depend on the analytes, different compounds of interest and the colour reagents require different mobile phases and solvents, adding additional analysis steps that either need to be performed by the technician or by the device used for analysis[9h][57] .
The testing of a sample using TLC usually requires the application of a sample “spot” on the TLC plate, and depending on the method may require evaporation of the sample solvent, but not always[9h][57].The colour reagent can be added before separation or after. A method with after colouration may require a brief stay an oven-like environment, or at least a drying period, before addition of colour[9h][57] [59]. This adds inconvenience for a portable system that is meant for field use. The TLC separations, coloured or not, can be archived for re-evaluation at a different time and place, or even be used to collect fractions of a sample.
There is the possibility to use optical detectors for colour tests and TLC, however that will be discussed in Chapter “Spectroscopy”.
Chapter 6: Spectroscopy
The immediate advantage that spectroscopy has over colour tests (with and without TLC), is that the detector is an objective instrument and not a subjective human. Spectroscopy itself covers many different analytical methods such as UV-Vis absorbance, fluorescence, chemiluminescence and infrared. The use of visible and ultraviolet light for absorbance detection is widespread when coupled to chromatography methods, despite that other methods (such as fluorescence) have better
sensitivity and specificity[9i][60]. That fact is the reason why the other optical techniques will be considered. Commercially available spectrophotometers are also generally bigger than the desired mobile phone size and there is no information on why the dimensions of the device are what they are[61][62]. It should also be noted that miniaturization of a system that makes us of spectroscopy also leads to a reduction in optical path length, which may weaken the method altogether[12][63]. The techniques will be reviewed after the following examples of spectroscopy that are coupled to microchip (viable) components.
X. Weng et al. developed a polydimethylsiloxane (PDMS) microfluidic chip with fluorescence detection using 410 nm light[64]. It can be designed to use a gentle application of force by just pressing a location with a finger to pump necessary substrate towards the analyte[64]. Heat may be needed to speed up the chemical reaction between analyte and substrate[64].
Surface enhanced raman spectroscropy (SERS) on a paper-based microfluidic device, a dipstick, that is manufactured with an inkjet-printer, which presents a useful and rapid application for on-site analysis[65]. It is a low-cost, less sensitive alternative to other SERS methods that provide a highly sensitive analysis at a higher cost[65]. Material used for the other SERS methods may also suffer from a short shelf life[65]. The dipsticks are also able to process much larger volumes of sample in
comparison to conventional SERS substrates, and can be used as surface swabs for the collection of trace analytes because of the flexible nature (Figure 6.1)[65]. A last advantage is that the paper device can improve signal intensity by concentrating the analyte through a lateral-flow
Figure 6.1: A detection method making use of a paper interface and Raman spectroscopy. A) shows the structure of the printed paper. B) illustrates the use as a surface swab for sampling while c) shows the ability for concentration of the taken sample by lateral flow. D) is the detection of the dipstick with the Raman probe. Reproduced from ref. [65].
Fluorescence
Fluorescence detectors are a tool with high sensitivity, fast response time and technical simplicity[2]. The sensitivity is around 10 fold that of UV-absorption, which usually has an LOD in the range of 10-6 - 10-5 M[60][67].However the difficulty for fluorescence lies in that some methods require the
choosing of an appropriate dye, the available choices numbering around 5500 that then need to be screened for fluorescence intensity and optimal working conditions such as solvent[2]. GHB would be a compound that needs a dye, while methamphetamine related compounds do not[2][58]. Certain dyes may even require additional steps in the analysis procedure due to solvent incompatibilities, such as in one case for the detection of gamma-butyrolactone (GBL) in real drinks because the presence of alcohol[2].
Detecting several different compounds with one device may therefor require that a selection of dyes is added to the sample. This adds an additional level of complexity since the dyes, or the dye addition system, must insure that the selected reagents do not negatively affect the analysis procedure. A further leap in the sensitivity of fluorescence can be done through Laser Induced Fluorescence (LIF) spectroscopy[60].This can result in LOD’s that lie in the range of 10-12 - 10-10 M[66][67]. It can still suffer from problems such as scattered excitation light that can interfere with the measurements and deteriorate the S/N (signal/noise) ratio, despite using a laser.[66]. However LIF with a PDMS/Glass chip that uses small-angle optical deflection to minimize interference from all light resulting from scattering, reflection or refraction[66]. Diode-LIF detection systems focus on near-infrared and red
regions of the spectrum where background noise is low, but lacks labelling dyes for wider application[66]. The size of the detector remains a problem, as previously mentioned for spectrophotometers, but while not strictly mobile phone sized or for the detection of drugs, an accessory has been presented by J. Canning et al. that allows a smart phone to act as a fluorometer (figure 6.2)[68].
Figure 6.2: A smartphone with accessory that allows fluorometry. Reproduced from ref.[68]. Chemiluminescence.
An alternative to Fluorescence is Chemiluminescence which has high sensitivity, a simple optical structure, a wide linear working range and low background noise[67][69]. The LOD values for heroin, morphine and codeine lie in the range of 10-8 to 10-9 M, methamphetamine has values of 0.002-0.041 ng/mg in hair (shaft, injection concentration around 8.37*10-8 to 1.7*10^-6 M) and 0.001-0.008 ng/ml (6.7*10-9 M -5.36*10-8 M) in blood[10][70].
It has also successfully analysed a variety of compounds when combined with microchip capillary electrophoresis (MCE)[67]. For signal detection it can use instruments such as a fluorometer or a photomultiplier tube (PMT)[67]. Unlike Fluorescence, it does not need an excitation source, and while not as bright it means that the system is more compact[71].
To produce chemiluminescence, a reagent is needed, for example luminol which it emits light in the wavelength range 425-435 nm[67]. Good contact, and mixing, between analyte and reagent is important for the strength of the luminescent signal[71]. Note that only a limited amount of molecules are chemiluminescent[71].
The choice of reagent for chemiluminescence is just as important as colours and dyes in previous analysis techniques. For example: in the case of heroin detection the chemiluminescence reagents of choice will be either tris(2,2’-bipyridine)ruthenium(III) (|Ru(bipy)3|3+) or potassium
permanganate[10]. The permanganate requires the heroin to be hydrolized into
6-monoacetylmorphine (6-MAM) and/or morphine for a reaction[10]. |Ru(bipy)3|3+ may give false positives when exposed to tertiary amines such as codeine[10].
Drawbacks are rapid reaction time and fast decay of the luminescence. This means that the reactants need to be mixed quickly and immediately analysed[71].
Infrared
Infrared makes use of vibrational (bending and stretching) characteristics of compounds for identification and in cased such as a Fourier-transform infrared(FTIR) attenuated total reflection (ATR) spectroscopy (with a one-step extraction) can be done in the field[72]. ATR instruments further have the ability to dry samples[72]. However the instrumentation used for IR is far too big for a mobile phone sized device[65]. And none of the literature mentions IR-LC.
LOD values for heroin and cocaine are 9 ng and 15 ng respectively if the paper-based method mentioned earlier is used, with a sample volume of 5 µl the molar concentrations will be around 5.95*10-6 M and 9.88*10-6 M respectively[65]. The FTIR ATR spectroscopy method can give LOD values of around 0.74 µg/ml (2.43*10-6 M] cocaine in saliva, 30 minutes after consumption of 40 mg[72].
Chapter 7: Electroanalytical
This chapter is about electroanalytical detection methods that are not coupled to an immunoassay. This means that they do not, in comparison, have the same specificity or complexity. Neither do they have the same size requirements as an MS system, colour reagents, highly specialized
instrumentation or high-intensity light[73.] A last advantage is that electroanalytical techniques allow for the analysis of turbid samples, although this is a debatable advantage because the aim is to have LC prior to the detection[74]. At its most basic, electroanalytical detection requires a sample, two electrodes, a current and a component that detects changes in the current. This means that electroanalytical sensors have simple to integrate structures that allow for easy (and cheap)
miniaturization without sensitivity loss[73][75]. Depending on the type of technique, it may even be many times more sensitive than a UV[76].
The general disadvantage of electroanalytical techniques, which adds inconvenience to a portable device, is the possibility of electrode surface fouling.[76][77].
Electroanalytical methods are not as widely used as an MS or some kind of spectrometry, but there are instances where heroin, methamphetamine and in one case GHB have been tested with electroanalytical techniques[ 88]. This chapter will briefly look at potentiometric, amperometric, conductometric and voltammetric methods. It is unlikely that any of these techniques will prove themselves too large for the desired (mobile phone sized) end product[74].
Potentiometry can be considered one of the most sensitive electrochemical (detection making use of an electrochemical reaction) techniques[12]. The response this method gives is only dependant on concentration and not on parameters such as detection cell dimensions and an in the case of conductivity measurements, the flow rates[12][73].
In the case of Sekula et al. where capillary electrophoresis (CE) is used for separation, an electrode is coated with polymers, resulting in something resembling an ion selective electrode (ISE) Which is not mentioned in the report itself, but in one of the illustration(see figure 7.1)[12]. The coated electrode is able to detect heroin (and related compounds) and a variety of amines, there for the detection of amphetamines would not be farfetched[12]. The lack of full understanding on how the detected potential is generated, is given as a reason as to why there are not more potentiometry/separation applications[12].
Figure 7.1: A CE and potentiometry detector setup that indicated the potentiometry detector as an ion selective electrode (ISE) . Reproduced from ref.[12].
Amperometric detection of morphine is promising for a portable and fast sensor, although less documented[79]. The process of detection is compatible to HPLC and molecularly imprinted
polymers (MIP), the later also happens to be suitable for bio-samples (urine, blood, etc.) because of its high selectivity for morphine[79] .
Conductometric detection measures a change in resistance (or impedance) that results from a change in concentration[73]. However the flow rate can have an effect as well, increasing at a low flow rate and decreasing at a high flow rate[73]. Reasons for this can be streaming potential, heat convection or deformation in the carbon nanotubes used[73]. A conductivity detector can make use of electrodes that are in contact with the sample, comparable to potentiometry and amperometry. However a different take on conductometry is the detection with capacitive contactless conductivity detection (C4D) [13]. C4D uses two radial metallic electrodes that are fitted around a capillary, voltage is applied, the current passes through the electrodes, capillary wall and gap between electrodes before being picked up again[13][80] [13] The differences between the conductivity of the buffer and the conductivity of the sample are converted into a peak[13] C4D is suitable for a variety of analytes including amphetamines, inorganic ions, carboxylic acids and explosives[13]. It also happens to be electrically isolated from the sample and prevents the previously mentioned electrode fouling[80]. See figure 7.2 For the C4D detector.
Figure 7.2: A C4D detector (right) with the sensor on the left. Dimensions (H/L/W) are 105/250/250 mm (3 kg) and 10/28/23 mm respectively. Reproduced from ref. [80].
Lastly we have voltammetric detection, with a drawback that may be an advantage. The drawback for voltammetric techniques, when used for simultaneous detection, is the overlap of
voltammograms (Figure 7.3), the graphene nanosheets also have antifouling properties [81]. With a separation step in the microchip, these voltammograms could be used as an additional tool for identification purposes.
Figure 7.3: Voltammogram of Morphine [M], Noscapine [N] and Heroin [H], with an overlap of M and H. A graphene nanosheets-glassy carbon modified electrode is used for the acquisition of signals c and d. A and b makes use of differential pulse voltammetry. A and c have an absence of drugs and b and d have a presence. Reproduced from ref.[81].
In the case of the voltammetry method, modified electrodes are used because unmodified electrodes used for electroanalysis suffer limitations such as low sensitivity, reproducibility and stability[81].
Specifically mentioned in the case of voltammetry, is the importance of the optimization of the detection potential because it affects the electrochemical behaviour of analytes[75]. A potential that is too low will give a signal that is too small, while a high potential gives an increase in the
background current (Figure 7.4)[75].
Figure 7.4: Voltammograms of Morphine (a) and Codeine (b). Reproduced from ref. [75] .
The LOD values for heroin are around 3*10-7 M for amperometric detection (2*10-7 M for morphine), 1.35*10-7 M for potentiometry and about 5*10-7 M for voltammetry.[12][75][81][79]. LODs can be higher in the case of voltammetry if several compounds are analysed simultaneously[81]. No
detection limits for conductometric detection of heroin were found during the writing of this report. For methamphetamine, amperometry gives LOD values of about 1.3 *10-6 M, potentiometry has a value of 10*10-6 M and voltammetry goes as low as 0.5*10-7 M[77]. The conductometric LOD (with a C4D) is 0.75 ppm for amphetamine and 0.94 ppm for methamphetamine about 5.56*10-9 M and 6.29*10-9 M if in water[13]. Table 7.1 holds the LOD values, in ppm, for several other drugs for C4D and UV detection that were done simultaneously.
Table 7.1: Detection limits for UV (detector of undetermined size) and C4D detection for a variety of compounds. Reproduced from ref. [13].
LOD (ppm) for UV and C4D Sample UV absorbance LOD C4D LOD 2-4-MPEA 0.46 0.38 MBA 0.75 0.83 2-MMA 1.5 2.5 PEA-HCl 1.8 1.3 AM-c-BC 1.7 1.4 MPEA 0.79 0.65 MMBA 0.54 0.58 AMP 0.81 0.75 MA 1.1 0.94 MDMA Peak 1 0.79 2.5 MDMA Peak 2 0.83 2.5 Average 1.0 1.3
Chapter 8: Conclusion.
As was stated in the beginning, four criteria would be used to recommend one type of detection for a portable, mobile phone sized, lab-on-a-chip device. The second criterion, “Sample versatility”, does not currently add (or detract) much to (from) any of the techniques value for the desired end product. There are no methods that are expressly stated to be unsuitable for one or more types of samples (urine, blood, hair, etc.). Even if there were, all samples require some degree of sample preparation to pass through an LC-chip. However some possibilities, such as the electroanalytical detection of GHB, still have gaps in the literature[78].
Criterion one, “Miniaturization potential”, filters out MS, simply because MS instruments have currently not been miniaturized enough, although portable versions exist. The same concerns exist for spectrophotometers and the IR detector mentioned earlier (Figure 6.1) also appears to require a little more space than the desired end product allows. This could mean that fluorescence and chemiluminescence systems may require comparable dimensions. Also because practically all of the literature deals with methods performed in the lab, the robustness of portable field applications remain unknown. Spectroscopy may be an option, but nothing can currently be said about the quality of a spectroscopic detector that fits inside the dimensions of the lab-on-a-chip device and is able to detect on the chip. If colour tests are to be used in combination with a LC-chip, then the volumes will have to be large enough so that the naked eye can observe the results.
This leaves spectroscopy, immunoassays and electroanalytical techniques. The third criterion is “Selectivity”. Immunoassays and some electroanalytical techniques (such as ISE) have to be
predictive, they should not give a signal unless they are designed to detect a specific compound, as to spectroscopy methods that require dyes . That is fine if the end device is only meant to detect a select amount of compounds, however every additional compound that needs to be detected means one more immunoassay or ISE. Since the end product has separation with a LC-chip the other
electroanalytical methods such as amperometry and conductivity detection can be applied as long as the identity of compounds can be confirmed, for example by spiking samples or using standards. Voltammetry can be used to detect several compounds simultaneously without making use of chromatography, however it means that the voltammograms could be used to identify
compounds[81]. Because it measures amperes versus a variable voltage, it may be difficult to apply it to a continuous flow LC-chip. Instead a continual flow with breaks that allow for a voltammetric detection sequence could be applied. If diffusion during the breaks proves a problem, the chip could be designed to add sections (plugs or droplet) of a liquid that does not mix with the mobile phase or extracts compounds of interest[82].
The last criterion for review is “Sensitivity”. A number of the LODs found have a concentration of around 1*10-5 M to 1*10-7. There is no strong reason to favour a different technique than electroanalytical, especially because there is no certainty that any of the LODs found in the lab procedures can be achieved in a portable field device. See attachment 2for an overview of LOD values found in the literature.
References and Literature
[1] A. J. deMello, “Control and detection of chemical reactions in microfluidic systems”, Nature., volume 442, 27 July 2006, 394-402.
[2] D. Zhai, Y.Q.E. Tan, W. Xu, Y. Chang, “Development of a fluorescent sensor for illicit date rape drug GHB”, Chem. Commun, 2014, 50, 2904-2906.
[3] M. P. Elie, M.G. Baron, J.W. Birket, “Enhancement of Microcrystalline Identification of gamma-Hydroxybutyrate”, J Forensic Sci, January 2008, Vol. 53, No. 1, 147-150.
[4] L.A. Baumes, M.B. Sogo, P. Montes-Navajas, A. Corma, H. Garcia. “A Colorimetric Sensor Array for the Detection of the Date-Rape Drug gamma-Hydroxybutyric acid (GHB): A Supramolecular
Approach”, Chem. Eur. J., 2010, 16, 4489-4495.
[5] A.A. Elian, “A novel method for GHB detection in urine and its application in drug-facilitated sexual assaults”, Forensic Science International, 109 (2000) 183-187.
[6a][6b][6c] Illustrations of the structures have been taken from the wikipedia pages of: a) gamma-hydrobutyric acid and gamma-aminobutyric acid, b) heroin and morphine, and c) methamphetamine and amphetamine.
[7] Y. Jong, Y. Ho, W. Ko, S. Wu, ”On-line stacking and sweeping capillary electrophoresis for detection heroin metabolites in human urine”, Journal of Chromatography A, 1216 (2009) 7570-7575.
[8] National Institute on Drug Abuse (NIDA), DrugFacts: Heroin:
http://www.drugabuse.gov/publications/drugfacts/heroin
[9] “Clarke’s Analysis of Drugs and Poisons in pharmaceuticals, body fluids and postmortem material” 4th edition, edited by A.C. Moffat, M.D. Osselton, B. Widdop, J. Watts. Published by Pharmaceutical Press, London (2011). Chapters: 8-Drug-facilitated Sexual Assault [9b], 18-Drugs in Saliva [9c], 19-Hair Analysis [9d], 28-Sampling, Storage and Stability [9a], 30 Colour Tests [9g], 31-Immunoassay [9f], 32-Ultraviolet, Visible and Fluorescence Spectrophotometry [9i], 37-Mass-Spectrometry [9e], 39-Thin-layer Chromatography [9h].
[10] J.M. Terry, Z.M Smith, J.J. Learey, R.A. Shalliker, N.W. Barnett, P.S. Francis, “Chemiluminescence detection of heroin in illicit drug samples”, Talanta, 116 (2013) 619-625.
[11] C. Han, Z. Shang, H. Zhang, Q. Song, ”Detection of hidden drugs with a molecularly imprinted electrochemiluminescence sensor”, Anal. Methods, 2013, 5, 6064-6070.
[12] J. Sekula, J. Everaert, H. Bohets, B. Vissers, M. Pietraszkiewicz, O. Pietraszkiewicz, F. Du Prez, K. Vanhoutte, P. Prus, L.J. Nagels, “Coated Wire Potentiometric Detection for Capillary Electrophoresis Studied Using Organic Amines, Drugs and Biogenic Amines”, Anal. Chem., 2006, 78, 3772-3779
[13] R. Epple, L. Blanes, A. Beavis, C. Roux, P. Doble, “Analysis of amphetamine-type substances by capillary zone electrophoresis using capacitively coupled contactless conductivity detection”, Electrophoresis, 2010, 31, 2608-2613.
[14] www.methproject.org
[15] O. Suzuki, H. Hattori, M. Asano, “Nails as useful materials for detection of methamphetamine or amphetamine abuse”, Forensic Science International, 24 (1984) 9-16.
[16] C. Yeh, W. Wang, Y. Lin, P. Shen, “A developed competitive immunoassay based on impedance measurements for methamphetamine detection”, Microfluidics and Nanofluidics, Septemter 2012, Volume 13, Issue 2, pp 319-329.
[17] K. Yagiuda, A. Hemmi, S. Ito, Y. Asano, “Development of a conductivity-based immunosensor for sensitive detection of methamphetamine (stimulant drug) in human urine”, Biosensors &
Biolectronics Vol.11, No. 8, pp 703-707, 1996.
[18] G. Sakai, S. Nakata, T. Uda, N. Miura, N. Yamazoe, “Highly selective and sensitive SPR immunosensor for detection of methamphetamine”, Electrochimica Acta 44 (1999) 3849-2854. [19] V. Vindenes, H.M.E. Lund, W. Andresen, H. Gjerde, S.E. Ikdahl, A.S. Christophersen, E.L. Øiestad, “Detection of drugs of abuse in simultaneously collected oral fluid, urine and blood from Norwegian drug drivers”, Forensic Science International 219 (2012) 165-171.
[20] P. Horvatovich, N.I. Govorukhina, T.H. Reijmers, A.G.J. van der Zee, F. Suits, R. Bischoff, ”Chip-LC-MS for label free profiling of human serum”, Electrophoresis 2007, 28, 4493-4505.
[21] N. Samyn, G. De Boeck, M. Wood, C.T.J. Lamers, D. De Waard, K.A. Brookhuis, A.G. Verstraete, W.J. Riedel, “Plasma, oral fluid and sweat wipe ecstasy concentration in controlled and real life conditions”, Forensic Science International 128 (2002) 90-97.
[22] E. Kaufmann, A. Alt, “Determination of GHB in urine and serum by LCMS using a simple one-step derivative”, Forensic Science International 168 (2007) 133-137.
[23] A. Wohlfarth, W. Weinmann, S. Dresen, ”LC-MS/MS screening method for designer
amphetamines, tryptamine and piperazines in serum”, Anal. Bioanal. Chem. (2010) 396:2403-2414. [24] V. Grenier, G. Huppé, M. Lamarche, P. Mireault, “Enzymatic Assay for GHB Determination in Forensic Matrices”, Journal of Analytical Toxicology 2012; 36; 523-528
[25] W. Dungchai, O. Chailapakul, C.S. Henry, “Multiple colorimetric indicators for paper-based microfluidic devices”, Analytica Chimica Acta 674 (2010) 227-233.
[26] S. Townsend, L. Fanning, R. O’Kennedy, “Salivary Analysis of Drugs-Potential and Difficulties”, Analytical Letters, 41:6, 925-948.
[27] I. Zancarno, R.P. Limberger, P.O. Bohel, M.K. dos Santos, R.B. De Boni, F. Pechansky, E.D. Caldas, “Prescription and illicit psychoactive drugs in oral fluid-LC-MS/MS method development and analysis of samples from Brazilian drivers”, Forensic Science International 223 (2012) 208-216.
[28] M. Wada, R. Ikeda, N. Kuroda, K. Nakashima, “Analytical methods for abused drugs in hair and their applications”, Anal. Bioanal. Chem. (2010) 397:1039-1067.
[29] E. Bertol, A. Argo, P. Procaccianti, F. Vaiano, M.G. Di Milia, S. Furlanetto, F. Mari, ”Detection of gamma-hydroxybutyrate in hair Validation of GC-MS and LC-MS/MS methods and application to a real case”, Journal of Pharmaceutical and Biomedical Analysis 70 (2012) 518-522.
[30] ,M. Chèze, A. Lenoan, M. Deveaux, G. Pépin, “Determination of ibogaine and noribogaine in biological fluids and hair by LC-MSMS after Tabemanthe iboga abuse Iboga alkaloids distribution in a drowning death case”, Forensic Science International 176 (2008) 58-66.
[31] S. Broecker, S. Herre, F. Pragst, “General unknown screening in hair by liquid chromatography-hybrid quadrupole time-of-flight mass spectrometry (LC-QTOF-MS)”, Forensic Science International 218 (2012) 68-81.
[32] M.M. Madry, K.Y. Rust, R. Guglielmello, M.R. Baumgartner, T. Kraemer, “Metabolite to parent drug concentration ratios in hair for the differentiation of tramadol intake from external
contamination and passive exposure”, Forensic Science International 223 (2013) 330-334.
[33] K.Y. Zhu, W. Leung, A.K. Ting, Z.C.F. Wong, W.Y.Y. Ng, R.C.Y. Choi, T.T.X. Dong, T. Wang, D.T.W. Lau, K.W.K. Tsim, “Microfluidic chip .based nano liquid chromatography couple to tandem
spectrometry for the determination of abused drugs and metabolites in human hair”, Anal. Bioanal. Chem. (2012) 402:2805-2815.
[34] A. Polettini, E.J. Cone, D.A. Gorelick, M.A. Huestis, “Incorporation of methamphetamine and amphetamine in human hair following controlled oral methamphetamine administration”, Analytica Chimica Acta 726 (2012)35-43.
[35] V. Cirimele, P. Kintz, P. Mangin, “Detection of amphetamines in fingernail: an alternative to hair analysis”, Arch. Toxicol. (1995) 70: 68-69.
[36] D.A. Engelhart, A.J. Jenkins, “Detection of Cocaine Analytes and Opiates in Nails form Postmortem Cases”, Journal of Analytical Toxicology, Vol. 26, October 2002.
[37] S. Taylor, N. France, “Miniature and micro mass spectrometry for nanoscale sensing applications”, Journal of Physics: Conference Series 178 (2009) 012003.
[38] W. Xu, N.E. Manicke, G.R. Cooks, Z. Ouyang, “Miniaturization of Mass Spectrometry Analysis Systems”, Journal of Laboratory Automation 2010 15: 433-439.
[39] “Miniaturization and Mass Spectrometry Chapter 1”, edited by S. Le Gac, A. van den Berg, Published by RSC publishing.
[40] C. Janfelt, R. Græsbøll, F.R. Lauritsen, “Portable electrospray ionization mass spectrometry (ESI-MS) for analysis of contaminants in the field”, Intern. J. Environ. Anal. Chem. Vol. 92, No. 4, 15 April 2012, 397-404.
[41] L. Ahonen, P. Keski-Rahkonen, T. Saarelainen, J. Paviala, R.A. Ketola, S. Auriola, M. Poutanen, R. Kostianen, “Comparison of liquid chromatography-microchip/mass spectrometry to conventional
