20 Photometers and Spectrophotometers 0.5 p=020 010 02 002 005 T0002001 0 15 20 Fig.7.Effect of stray light on ab. d. sorbance A of magnitude of0.0o5≤p≤0.02，an absorbance of0.6to0.8 should not be exceeded.The graph shows for each individual case the greatest value of ab- sorbance which may be used in order to avoid exceeding a specific relative error in the m ment. The useful light must be removed from the light path when measuring the proport on of stray light. th ent I ca aused sole ly by stray ligh an be sorption r shielding After the proportion of stray light,p,has been established,the correct transmission (T)is given by: T=T-p 1-D In practice,to determine the true transmittance,T in the presence of stray light,p,we can proceed as follows.Cuvettes containing solvent(L),sample solution (P)and a very strongly absorbing solution (K)are used.The ette k has the sam athler gth and contains the substance to be mea ed at such ration that the actual transmission lies below 0.1 in the regi n.We measure ainst L ult T and k against L (resul t p)and calculate the sam ple,using the above equati It follows from the definitions ofT,Tand p that we must use the val as a fraction of 1.0 and not as percentage figures in formulae,i.e.T=0.32 (instead of 32%)and p=0.01 (instead of 1)
20 Photometers and Spectrophotometers t OJ <x: '- <[ <l 0.2 i-f-+-j'-t-=-=-::-I-+-+-l 0.1 1-.,-4-t;-r-7''t-r-l A'- Fig. 7. Effect of stray light on absorbance A of magnitude of 0.005 ~ p ~ 0.02, an absorbance of 0.6 to 0.8 should not be exceeded. The graph shows for each individual case the greatest value of absorbance which may be used in order to avoid exceeding a specific relative error in the measurement. The useful light must be removed from the light path when measuring the proportion of stray light. Then the photoelectric current If caused solely by stray light can be measured. The useful light can be removed by absorption or shielding [8]. After the proportion of stray light, p, has been established, the correct transmission (T) is given by: T'-p T=- 1-p . In practice, to determine the true transmittance, T in the presence of stray light, p, we can proceed as follows. Cuvettes containing solvent (L), sample solution (P) and a very strongly absorbing solution (K) are used. The cuvette K has the same pathlength and contains the substance to be measured at such a concentration that the actual transmission lies below 0.10/0 in the useful-light region. We then measure P against L (result T'), and K against L (result p) and calculate the corrected transmission T of the sample, using the above equation. It follows from the definitions of T, T' and p that we must use the values as a fraction of 1.0 and not as percentage figures in formulae, i.e. T' = 0.32 (instead of 32%) and p = 0.01 (instead of 1 %)
Light Sources for UV-VIS Spectroscopy 21 This method of approximation,when determining p,is adequately ac This with a stro on band since,in this event,the test solution K ay ligh t m more than the sample P or solvent L.However,correction of the stray ligh error is particularly important in this case because the sample is likely to have a relatively high absorption in the useful-light region. Under less favorable conditions,where the sample P,when at sufficiently high concentration for use as test sample K,absorbs the stray light con- siderably more than the actual sample P,another substance must be used as test sample.A suitable substance can easily be found when measuring in the short-wavelength UV region.The assumption that L and P always ab- e extent should be avoided.In this case the T'= 1+Ic Io+Ig. but the stray-light photoelectric current Ir and It,differ from one another.It is obvious from this that two values,p'and p can be me asured with an ideal test sample (p')or by cutting out the useful light (p). 卫=0+1。 g=10+。 In this case,the following applies for the true transmission T=T-p 1-p Whether it improved approximaionto depends upon how accurately the two stray light ratios p'and p"can be measured The work of Luther et al.[and Luck [10]should be consulted with regard to the stray-light error of Uv measurements.Further details and pro cedures for and evaluation of the stray-light error are given by Burgess [11], Cook et al.[12],Poulson [13],Renle [14]and Kaye [15]. Reference may be made to Derkosch and Gauglitz for further details of measurement techniques [16,17]. 3.4 Light Sources for UV-VIS Spectroscopy sone of the most important components of a U spectr t must be ur easure a complete absorption spectrum in the 900m
Light Sources for UV-VIS Spectroscopy 21 This method of approximation, when determining p, is adequately accurate if the useful-light region coincides with a strong absorption band since, in this event, the test solution K will not weaken the stray light much more than the sample P or solvent L. However, correction of the stray light error is particularly important in this case because the sample is likely to have a relatively high absorption in the useful-light region. Under less favorable conditions, where the sample P, when at sufficiently high concentration for use as test sample K, absorbs the stray light considerably more than the actual sample P, another substance must be used as test sample. A suitable substance can easily be found when measuring in the short-wavelength UV region. The assumption that Land P always absorb the stray light to the same extent should be avoided. In this case the following equation applies T' = I+If Io+If ' o but the stray-light photoelectric currents, If and If, differ from one o another. It is obvious from this that two values, p' and p", can be measured with an ideal test sample (P') or by cutting out the useful light (p"). I If p= - -, Io+If o I p"=~ Io+If o In this case, the following applies for the true transmission T'_p' T=-. l-p" Whether it is appropriate to use this improved approximation to correct T' depends upon how accurately the two stray light ratios p' and p" can be measured. The work of Luther et al. [9] and Luck [10] should be consulted with regard to the stray-light error oj UV measurements. Further details and procedures for and evaluation of the stray-light error are given by Burgess [11], Cook et al. [12], Poulson [13], Renle [14] and Kaye [15]. Reference may be made to Derkosch and Gauglitz for further details of measurement techniques [16, 17]. The light source is one of the most important components of a UV-VIS spectrometer; and it must be a continuum source in order to measure a complete absorption spectrum in the UV-VIS region of 190-900 nm
22 Photometers and Spectrophotometers In the visible spectral region,this requirement is met by the tungster lamp which is a black-body source,the spectral energy distribution of which is described by Planck's radiation formula [18].According to Wien's law, the energy maximum of the energy distribution lying in the NIR region moves to shorter wavelengths with increasing temperature [19].The short- wavelength edge of the distribution in the visible region is thereby raised, hence the radiation yield increases in this region.However,the higher tem- perature of a tungsten coil means a shorter lamp life since tungsten evapora from the oil The aporated tungsten precipitates on the cooler lass envel ope The velope then causes a reduction of radiated mitted energy du absorption enof the e radiation,see Sect.3.3.1 This problem is overcome in modern halogen lamps s.The added haloger (e.g.iodine vapor)forms a volatile compou with the evapora ated t ungster peratures the tungsten coil thickness remains almost constant even during prolonged operation of the lamp.At the same time,the condensation of evaporated tungsten on the bulb is considerably reduced.Thus,both effects result in a higher radiation output in the visible spectral region and a longer lamp life As with any other black-body source.the spectral energy distribution of a tungsten lan eases rapidly below 400 nm.Therefore,these lamps employed in the UV region.In the region below 350 nm gas-dis re the efore used a s radiation urces:of these the hydroger mp isthe most mpor tant so ce of light The hydroger continuo een 160 (62500 000 cm n-1).For this reason,below ca. 28500 cm-),hydrogen or deuterium lamps are commonly used as ligh sources in spectrophotometers.Their construction is described briefly in the following [20]: A tungsten coil is fitted as anode on the axis of a cylindrical,thin-walled quartz discharge tube.An activated tungsten double coil mounted laterally is generally used as cathode.The lamp is filled with either hydrogen or deuterium at ca.10 Torr.A heating voltage is applied to the cathode coil so that the ignition potential remains low (200-400 V)because of the av al en ission properties.Hydrogen,the lightest gas,has a er high diffusion erov losses due to thermal ducti e very laige a gain is ndingly low.For that rea radia tion yield by about oas a result of the dou ng of the In order to produce the highest possible rad nsiti es it weight is necessar to restrict the discharge between cathode and anode by me ans an ape ture of small cross-section (ca.1 mm2)formed from a high-melting point metal (molybdenum). Between 160 and ca.400 nm,an emission occurs due to the transition from a stable excited state,which is the lowest triplet state,of the H2
22 Photometers and Spectrophotometers In the visible spectral region, this requirement is met by the tungsten lamp which is a black-body source, the spectral energy distribution of which is described by Planck's radiation formula [18]. According to Wien's law, the energy maximum of the energy distribution lying in the NIR region moves to shorter wavelengths with increasing temperature [19]. The shortwavelength edge of the distribution in the visible region is thereby raised, hence the radiation yield increases in this region. However, the higher temperature of a tungsten coil means a shorter lamp life since tungsten evaporates from the coil. The evaporated tungsten precipitates on the cooler glass envelope. The deposit on the glass envelope then causes a reduction of radiated energy due to absorption of the emitted radiation, see Sect. 3.3.1. This problem is overcome in modern halogen lamps. The added halogen (e.g. iodine vapor) forms a volatile compound with the evaporated tungsten which decomposes on the hot tungsten coil. Therefore, even at higher temperatures the tungsten coil thickness remains almost constant even during prolonged operation of the lamp. At the same time, the condensation of evaporated tungsten on the bulb is considerably reduced. Thus, both effects result in a higher radiation output in the visible spectral region and a longer lamp life. As with any other black-body source, the spectral energy distribution of a tungsten lamp decreases rapidly below 400 nm. Therefore, these lamps cannot be employed in the UV region. In the region below 350 nm gas-discharge lamps are therefore used as radiation sources; of these the hydrogen or deuterium lamp is the most important source of light. The hydrogen discharge provides a continuous spectrum between 160 and 400 nm (62500-25000cm- 1). For this reason, below ca. 350nm (above 28500cm- 1), hydrogen or deuterium lamps are commonly used as light sources in spectrophotometers. Their construction is described briefly in the following [20]: A tungsten coil is fitted as anode on the axis of a cylindrical, thin-walled, quartz discharge tube. An activated tungsten double coil mounted laterally is generally used as cathode. The lamp is filled with either hydrogen or deuterium at ca. 10 Torr. A heating voltage is applied to the cathode coil so that the ignition potential remains low (200-400 V) because of the favorable thermal emission properties. Hydrogen, the lightest gas, has a very high diffusion velocity; consequently, energy losses due to thermal conduction are very large and the radiation gain is correspondingly low. For that reason, deuterium is used in modern lamps which improves the radiation yield by about 300/0 as a result of the doubling of the molecular weight. In order to produce the highest possible radiation intensities it is necessary to restrict the discharge between cathode and anode by means of an aperture of small cross-section (ca. 1 mm2) formed from a high-melting point metal (molybdenum). Between 160 and ca. 400 nm, an emission occurs due to the transition from a stable excited state, which is the lowest triplet state, 3I;, of the H2
Light Sources for UV-VIS Spectroscopy 23 molecule,to the repulsive state21].Thus,the resulting continuous.Dissociation occurs after the system has reverte state and,for that reason,the continuous spectrum emitted is also referrec to as dissociation radiation [21].The hydrogen or deuterium atoms formed recombine to molecules H2 or D2 on the cold surfaces in the dis. charge chamber.In addition to the continuum,the Balmer series of the H or d atoms can be seen in the visible region and the appropriate Ha Da (n=3n=2)or H(n=4-n=2)lines can be used for calibrating the wavelength scale of a UV-VIS spectrometer.If the lamps are filled with a The lime lie mixture the Balmer series of both gases are emitted. t486.12 nm and the D-line at 485.99nm,and the dif- ference of 4 =0.13nmor4=5.5c can be 1 conveniently used to check the resolving pow r of a UV-VIS spectrometer.For manufacturers of deuterium lamps see [22]. In addition to these two most important lamps,others are employed in special applications where,for example,a higher radiati utput VIS and UV region is of particular interest. Noble-gas discharge lamp show a prominent continuum,and of these,the xenon-high pressure lamp is the most commonly used.The continuum of a xenon discharge cor- responds to recombination radiation,i.e.xenon atoms ionized at high gas s recombine with electrons formed during ionization and emit in the UV-VIS and NIR region. Ma s pr roduce xenon lamps with powers from 75 W to several kW [23],and th ey are pr ed in fluorescence and luminescence excitation Sect.I the ell as in photoacoustic spectroscopy (see are e also used as light sources in microscope spectrometer r these application s xenon lam nps of up to 450 W are usually employed The quartz bulb of a xenon-maximum pressur lamp considerably thicker than that of a deuterium lamp for reas ns of safety.Th s,the inter sity of UV radiation which is already diminishing toward sh orte lengths is increasingly absorbed by the quartz below 250 nm.For this rea son,xenon lamps are produced for special applications with bulbs made of a quartz material such as Spectrosil which is highly transparent in the UV region [23]. In contrast to metal-vapor discharge lamps (e.g.mercury lamps),xenon amps re no ing up time be ecause they reach their full power im- y after switching on Hence they can be easily modulated or used as flash In addition to these continuo us light sour metal-vapor discharge lamps are often employed.The m ortan amp o this type e is the me cury vapor lamp which is produced for low pres n pre mum pressure .As line,they are specifically used fo definite spectral lines by means of suitable interference filters or filter com binations (cf.Table 1,page 10),and therefore,they are employed in photo- meters as light sources
Light Sources for UV-VIS Spectroscopy 23 molecule, to the repulsive state 3I:: [21]. Thus, the resulting spectrum is continuous. Dissociation occurs after the system has reverted to the lower state and, for that reason, the continuous spectrum emitted is also referred to as dissociation radiation [21]. The hydrogen or deuterium atoms so formed recombine to molecules H2 or D2 on the cold surfaces in the discharge chamber. In addition to the continuum, the Balmer series of the H or D atoms can be seen in the visible region and the appropriate Ha-, Da- (n = 3 -+ n = 2) or Hp-(n = 4 -+ n = 2) lines can be used for calibrating the wavelength scale of a UV-VIS spectrometer. If the lamps are filled with a hydrogen-deuterium mixture the Balmer series of both gases are emitted. The Hp-line lies at 486.12 nm and the Dp-line at 485.99 nm, and the difference of LtA. = 0.13 nm or Ltv = 5.5 cm- 1 can be conveniently used to check the resolving power of a UV-VIS spectrometer. For manufacturers of deuterium lamps see [22]. In addition to these two most important lamps, others are employed in special applications where, for example, a higher radiation output in the VIS and UV region is of particular interest. Noble-gas discharge lamps show a prominent continuum, and of these, the xenon-high pressure lamp is the most commonly used. The continuum of a xenon discharge corresponds to recombination radiation, i.e. xenon atoms ionized at high gas temperatures recombine with electrons formed during ionization and emit radiation in the UV-VIS and NIR region. Manufacturers produce xenon lamps with powers from 75 W to several kW [23], and they are preferred in fluorescence and luminescence excitation spectroscopy (see Sect. 5.5) as well as in photo acoustic spectroscopy (see Sect. 5.4). In the UV-VIS region, these lamps are also used as light sources in microscope spectrometers. For these applications xenon lamps of up to 450 Ware usually employed. The quartz bulb of a xenon-maximum pressure lamp is considerably thicker than that of a deuterium lamp for reasons of safety. Thus, the intensity of UV radiat~on which is already diminishing toward shorter wavelengths is increasingly absorbed by the quartz below 250 nm. For this reason, xenon lamps are produced for special applications with bulbs made of a quartz material such as Spectrosil which is highly transparent in the UV region [23]. In contrast to metal-vapor discharge lamps (e.g. mercury lamps), xenon lamps require no warming-up time because they reach their full power immediately after switching on. Hence, they can be easily modulated or used as flashlamps. In addition to these continuous light sources, metal-vapor discharge lamps are often employed. The most important lamp of this type is the mercury vapor lamp which is produced for low pressure, high pressure or maximum pressure [24]. As line sources, they are specifically used for isolating definite spectral lines by means of suitable interference filters or filter combinations (cf. Thble 1, page to), and therefore, they are employed in photometers as light sources
24 Photometers and Spectrophotometers The mercury low-pressure lamp,operated at a pressure of 0.006Torr emits almost exclusively an intercombination line at A=253.7 nm and is suitable for use in photochemistry.Depending on its power,a mercury high pressure lamp operates at a pressure of between 10 to 50 atm.In addition to the emission of the characteristic mercury lines,the emission spectrum of a high pressure lamp also shows a continuous background.On account of the high vapor pressure in the discharge,the spectral lines are con- siderably broadrry medium-pressure lamp hasacon relatively clear.Therefore. use of his pre red i the gion and as an excitation lig urce ir inrich ave given a detailed account of the light sources briefly described here [25]. In recent years,lasers have come to the fore as light sources for special applications;and noble-gas and dye lasers in particular should be men tioned.The dye laser has the advantage of being completely tunable over a wavelength range of ca.600-1000 nm depending on the dye used.As a rule,an argon-ion laser is used as a pump light source for a dye laser. W.Demtroder has given a detailed account of the applications of lasers in spectroscopy [26]. References 1.Osram,Druckschrift(1978)Licht fur Kinoprojektion,Technik und Wissenschaft,Ausg 2.Jener-(19)Ulmanns Encvklopadie der technischen Chemie.4.Aufl.Bd Mainz New York.S 279-116 5.Talsky G,Mayring L,Kreuzer H(1978)Angew Chem90:840 6 FP.Sidwell Jr AE (1937)J Phys Chem 41:37 8.Preston IS(1936)J Scient Instr 13:3681 955)Z Elektrochem Ber Bunsenges 59:159 11.Burgess C.Knowles A (1981)Standards in Absorption Spectrometry,Ultraviolet Spec- V AR9即 49405 16.Derkosch(A em53220 tralanalyse im ultravioletten.sichtbaren und infrarote Gebiet,Bd 5 de Methoden der Analyse in der Chemie.Akad Verlagsges,Frankfurt/M M190 kopie.Attempto,Tabingen 19.Wien W(1894)Ann Phys,series 2,52:132
24 Photometers and Spectrophotometers The mercury low-pressure lamp, operated at a pressure of 0.006 Torr, emits almost exclusively an intercombination line at A = 253.7 nm and is suitable for use in photochemistry. Depending on its power, a mercury high pressure lamp operates at a pressure of between 10 to 50 atm. In addition to the emission of the characteristic mercury lines, the emission spectrum of a high pressure lamp also shows a continuous background. On account of the high vapor pressure in the discharge, the spectral lines are considerably broadened. A mercury medium-pressure lamp has a considerably lower operating pressure and the spectral lines are relatively clear. Therefore, the use of this lamp is preferred in the UV-VIS region and as an excitation light source in photochemistry. Schafer and Heinrich have given a detailed account of the light sources briefly described here [25]. In recent years, lasers have come to the fore as light sources for special applications; and noble-gas and dye lasers in particular should be mentioned. The dye laser has the advantage of being completely tunable over a wavelength range of ca. 600-1000 nm depending on the dye used. As a rule, an argon-ion laser is used as a pump light source for a dye laser. w. Demtroder has given a detailed account of the applications of lasers in spectroscopy [26]. References 1. Osram, Druckschrift (1978) Licht fUr Kinoprojektion, Technik und Wissenschaft, Ausg Dez, 869:1 2. Jenaer Glaswerke Schott & Gen, Mainz, Farb- und Filterglas 3. Perkampus, H-H (1980) In: Ullmanns Encyklopadie der technischen Chernie, 4. Auf!, Bd 5. Verlag Chernie, Weinheim, 269ff 4. Perkampus, H-H (1983) In: Analytiker-Taschenbuch, Bd 3. Springer, Berlin Heidelberg New York, S 279-316 5. Talsky G, Mayring L, Kreuzer H (1978) Angew Chern 90:840 6. Shibata S (1976) Angew Chern 88:750 7. Hogness TR, Zscheile Jr FP, Sidwell Jr AE (1937) J Phys Chern 41:379 8. Preston IS (1936) J Scient Instr 13:3681 9. Luther H, Pokkels G (1955) Z Elektrochem Ber Bunsenges 59:159 to. Luck W (1960) ibid 64:676 11. Burgess C, Knowles A (1981) Standards in Absorption Spectrometry, Ultraviolet Spectrometry Group, Vol I. Chapman and Hall 12. Cook RB, Jankow AR (1972) J Chern Ed 49:405 13. Poulson RE (1964) Appl Opt 3:99 14. Renle A (1971) ColI Spectr Int XVI 1:107 15. Kaye W (1981) Anal Chern 53:2201 16. Derkosch J (1967) Absorptionsspektralanalyse im uhravioletten, sichtbaren und infraroten Gebiet, Bd 5 der Methoden der Analyse in der Chemie. Akad Verlagsges, Frankfurt/M 17. Gauglitz G (1983) Prakt Spektroskopie. Attempto, Tiibingen 18. Planck M (1901) Ann Phys 4:553 19. Wien W (1894) Ann Phys, series 2, 52:132