外文翻譯--優(yōu)化流動(dòng)注射氫化物誘導(dǎo)電感耦合等離子體光譜法，在電解錳中測(cè)定硒

1、<p>　　Optimisation of flow-injection-hydride generation inductively coupled plasma spectrometric determination of selenium in electrolytic manganese</p><p><b>　　Abstract</b></p><p

2、>　　Flow-injection-hydride generation procedure for Se in electrolytic manganese was optimized by means of the experimental design ap- proach. Instrumental variables like power supplied (P), sample (F) and argon (G) fl

3、ow rates together with chemical variables like NaBH4 and HCl concentrations were studied. In case of the chemical variables, it was concluded that sodium tetrahydridoborate concentrations</p><p>　　higher tha

4、n 1.0% extinguished the plasma while HCl concentration should always be higher than 0.6 mol dm?3 .</p><p>　　The analysis of effects</p><p>　　suggested that all the instrumental variables are sig

5、nificant factors, and the optimum conditions were P = 1550 W, F = 4.75 mL min?1 and</p><p>　　G = 0.6 mL min?1 . The influence of Mn was specially studied and it was concluded that the interferences were neg

6、ligible if Mn is be-</p><p>　　low 2.0 g L?1 . In the same sense, the interferences of antimony(III), arsenic(V) and mercury(II) were also considered negligible. In the same sense, the interferences of antimo

7、ny(III), arsenic(V) and mercury(II) were also considered negligible. </p><p><b>　　Finally,</b></p><p>　　a detection limit of 0.0005% (w/w) was obtained (a repeatability R.S.D. <2.

8、0% for all Se concentrations tried). Some manganese samples were also spiked with different concentrations of Se(IV) and the results demonstrated to be in good statistical agreement with expected values.1. </p>&l

9、t;p>　　1. Introduction</p><p>　　Manganese is essential to metallurgical industries (alu- minium and steel foundries) with important applications. Manganese additions in aluminium are required for food and

10、drinking packings, domestic tools, decoration or covering. Manganese used for aluminium alloys is produced by elec- trolysis wherein selenium additions are purposefully added to improve the electrical current efficient [

11、1].</p><p>　　This leads to the contamination of manganese as a result of co-deposition of selenium at the cathode. It is believed that over 90% of the selenium used for improving the current efficiency of th

12、e electro-winning of electrolytic manganese enters the cathode in the elemental form and, thus, some of the electrolytic grade of manganese currently used can contain variable quantities of selenium (0.03–0.16%)</p>

13、;<p>　　Within the operating aluminium furnaces (700–800 ?C),</p><p>　　selenium is supposed to evaporate out of the molten alu- minium as metal vapour. Therefore, as a consequence of toxicity and envir

14、onmental hazard in high concentrations of selenium and its compounds [3,4], furnace stack emissions and occupational airborne exposures must be monitored. In that sense, total selenium concentration in aluminium al- loyi

15、ng process, raw and waste materials must be also con- trolled to identify the Se environmental fate and exposures in aluminium processing.</p><p>　　Hagelstein [2] studied the environmental management of sele

16、nium in aluminium processing and indicated that the main environmental issues were se- lenium produced in operating aluminium furnaces. The aluminium and steel in- dustries may limit their future environmental liabiliti

17、es due to selenium accumulation in processing facilities and waste streams by avoiding or controlling raw material inputs con- taining selenium. Thus, different confident selenium analyses should be developed for impleme

18、n</p><p>　　Great selectivity and sensibility are obtained in aqueous samples by using different atomic techniques as inductively coupled plasma optical emission spectrometry (ICP–OES), ETAAS or AFS. Besides,

19、 the implementation of cold vapour or hydride generation previous to electrothermal atomic ab- sorption detection increases this selectivity minimising spec- tral interference caused by other matrix component [7]. In t

20、he last decade, on-line flow-injection (FI) separation techniques have become increa</p><p>　　Hydride generation (HG) together with inductively cou- pled plasma optical emission spectrometry (ICP–OES) is a w

21、idely used method for the determination of selenium [9–10]. However, there is lack of information in the literature on this procedure for Se determination in electrolytic manganese. This analytical technique (FI–HG–ICP–O

22、ES) enables sepa- ration of selenium from the major components of the sample but, in fact, is prone to interference from several transition metal ions and other volati</p><p>　　Specifically, manganese ions p

23、resent in the sample solution in hydride generation conditions yield manganese borides, which absorb hydrides and cause their decomposition [11]. Manganese species may catalyse decay of reducing agent as well. This chemi

24、cal inter- ference is well documented in a review article by Nakahara and Kikui [12]. However, the treatment of mutual interference by hydride-forming elements is relatively scarce and only few works discuss the possible

25、 interference mechanism [13,14</p><p>　　As a consequence, the main purpose of the present study is to find the optimal set of operational conditions that allow a simple and sensitive method for the determina

26、tion of Se in electrolytic manganese by FI–HG–ICP–OES. The experi- mental approach to accomplish this aim was the experimental design [15]. As it is already known, experimental design is the most powerful way to make eff

27、icient experiments as we get the information we need with the minimum effort. In this sense, a full factorial des</p><p>　　2.2. Instrumentation</p><p>　　An inductively coupled plasma optical emi

28、ssion spectrometer IRIS advantage (Thermo Corporation) with a CID detector was used. The instrument consists on a multi-channel peristaltic pump and gasliquid separator assembled to the spectrometer (T-PHD). A 196.090 nm

29、 line for Se was used and no background correction system was used. The presence of spectral interference was considered negligible when the contribution of the interfering signal at the selenium wave- length was less th

30、an 10% [16,17].</p><p>　　Detailed operating conditions for plasma excitation and hydride generation are listed in Table 1.</p><p>　　2.3. Digestion of electrolytic manganese</p><p>　

31、　Electrolytic manganese (Xiangxi Autonomous Prefecture, China) was collected in flakes and crushed into a laboratory mill (Retsch S100) and sieved (500 µm). An acid digestion of the manganese powder obtained (2 g) w

32、as immediately per- formed using concentrated HCl without heating [18]. This procedure was performed both with a water cooling (refriger- ant) and without it. Similar efficiency (signals) was obtained for all samples in

33、both procedures (Fig. 1). Thus, analyti- cal determination of Se(I</p><p>　　2.4. Thermodynamic selenium speciation</p><p>　　The two common inorganic oxidation states of selenium are IV and VI, b

34、ut only the lower state (IV) is successfully reduced to hydrogen selenide by NaBH4 [19]. Thus, it is important to verify that the main species of Se after HCl digestion is IV state.</p><p>　　This fact has b

35、een both thermodynamically and experi- mentally confirmed (Fig. 2a and b). Fig. 2a shows the dis- tribution diagram of Se against pH while Fig. 2b shows the predominance diagram of Se against pH and redox potential,using

36、 the medusa program.</p><p>　　2.5. Hydride generation procedure</p><p>　　For selenium hydride generation, the sample (in HCl</p><p>　　0.6 mol dm?3 ) and sodium tetrahydridoborate (1

37、.0%) solu-</p><p>　　tions are introduced into the FI system using a peristaltic pump. The solutions are then pumped through a gas–liquid separator in a continuous stream. The released hydrides are supporte

38、d by the carrier gas flow (Ar flow rate of</p><p>　　0.6 mL min?1 ) to the plasma. The reading time is 30 s and</p><p>　　no background correction is used.</p><p>　　3. Results

39、and discussion</p><p>　　The efficiency of selenide generation depends directly on the NaBH4 concentration. However, NaBH4 concentrations higher than 1.1% destabilised the plasma discharge due to excessive hy

40、drogen evolution and concentrations higher than</p><p>　　2.0% extinguished the plasma. Thus, NaBH4 concentration of 1.0% was considered the highest operating value in all the analysis performed and it was ke

41、pt constant.</p><p>　　Hydrochloric acid has been found to be the most satis- factory medium to generate hydrides. The efficiency of selenide generation was constant at HCl concentrations above</p><

42、;p>　　1 mol dm?3 (Fig. 3). </p><p>　　However, higher HCl concentrations</p><p>　　than 1 mol dm?3 produce several deleterious effects: a decrease in pump winding life, an increase in acid fum

43、e problems and an increase in the violence of by product hydrogen gas evolution. The efficiency of hydride generation increased slightly with the increasing HCl concentration or reached</p><p>　　a plateau ab

44、ove 0.6 mol dm?3 . Lower concentration values</p><p>　　(HCl <0.4 mol dm?3 ) were enough to produce complete inhibition in the selenium reduction.</p><p>　　Effect of the presence of electrolyt

45、ic manganese on the selenium determination (Se: 200 µg L?1 ) was also examined for concentrations ranged from 0 to 6.0 g L?1 Mn.</p><p>　　The hydride generation efficiency (%R) was defined as:where SSei

46、 is the intensity of Se obtained with interferences</p><p>　　and SSeo is the intensity of Se with interferences.</p><p>　　In general, manganese had a negative effect on the selenium resp

47、onse at concentrations higher than 2 g L?1</p><p>　　(Fig. 4a). This inhibition could be overcome increasing the</p><p>　　HCl concentration but two assumptions were taken into ac- count to fix op

48、timum HCl concentration at 0.6 mol dm?3:</p><p>　　low HCl concentrations are less drastic and inexpensive, and</p><p>　　Mn concentrations higher than 2 g L?1 were not expected.</p><p&

49、gt;　　Finally, interference from other hydride forming elements</p><p>　　like Sb(III), As(V), Hg(II) on the determination of Se(IV) were systematically investigated. It has been shown in the literature [14] t

50、hat hydride generation interference does not depend on the analyte-to-interferent ratio but on the inter- fering concentration in the solution for measurement. The relative sensitivities (%R), i.e. the ratio of the signa

51、l obtained with different concentrations of As(V), Sb(III) or Hg(II) and without those interferences, are depicted in Fig. 4b. Only</p><p>　　when concentration levels of As(V) are high (2 mg L?1 ), adecreasi

52、ng Se signal was observed. Therefore, influence of those volatile elements on the selenium response was not considered in this case.</p><p>　　Reaction rates for hydride generation are controlled by several i

53、nterdependent variables: (I) the concentration of NaBH4 as reducing agent, (II) the acid concentration of the sample, (III) the chemical form of the hydride-forming el- ement and (IV) the oxidation state of the hydride-f

54、orming element. On the one hand, as mentioned above, NaBH4 and HCl concentrations were not included in the optimisation due to the experimental limitations. On the other hand, Se(IV) is the main oxidation state in s</

55、p><p>　　trolytic manganese powder spiked with 48 and 80 µg L?1 of</p><p><b>　　Se(IV).</b></p><p>　　3.1. Optimisation of selenium detection: factorial design</p>

56、<p>　　Several variables such as: power supplied (P), sample flow rates (F), and auxiliary Ar flow rate (G) could affect the se- lenide detection. In order to find the main factors affecting the detection procedure,

57、 a two-level full factorial design (23 +4 replicates of central point) was carried out using ‘The Un- scrambler’ program (Camo, As. Norway, v.7.5) [23]. Table 2 lists the experimental design matrix and the response obtai

58、ned (intensity, mean of three determinations). The analysis of the effec</p><p>　　3.2. Orthogonal central composite design (CCD)</p><p>　　Since all the variables were significant, the design was

59、 extended to a central composite design. In this way, it is possible to build the response surface and to obtain the instrumental conditions that define the maximum response. The CCD de- signs developed were carried out

60、using ‘The Unscrambler’ program. Table 3 shows the central composite design together with the response obtained. The response surface was calcu- lated with ‘The Unscrambler’, and the significant effect of G and P variabl

61、es w</p><p>　　3.2 。Orthogonal central composite design （ CCD ）</p><p>　　If the F variable is fixed at the maximum level of the CCD (180 rpm or 4.75 mL min?1 ), it is possible to plot the respons

62、e</p><p>　　surface as a function of the most significative variables, as can be seen in Fig. 5. From that plot we can deduce the optimum conditions within the factor space, i.e. without any extrap- olation.

63、As a consequence, the determination of Se can be carried out with a power of (P) 1550 W and an Ar flow rate</p><p>　　of (G) 0.6 mL min?1 .</p><p>　　（G）的0.6 0.6 mL min?1 。</p><p>　　A

64、s there are no certified reference materials for Se(IV) in electrolytic manganese, synthetic Se(IV) tests solutions prepared in our laboratory and different real samples spiked with 48 and 80 µg L?1 of Se(IV) were

65、used to check the accuracy of the analytical method (Table 4). As it can be seen, the matrix effect is negligible because the Se concentrations obtained are identical in both direct and spiked determination. Finally, the

66、 detection limit, defined as blank signal + 3 S.D., where S.D. i</p><p>　　2.0% for all concentrations tried.</p><p>　　4. Conclusions</p><p>　　The use of experimental designs shows t

67、hat the three variables studied (power supplied, Ar flow rate and pump flow rate) are significant in inorganic selenium determina- tion. Sodium tetrahydridoborate and hydrochloric acid are important variables too. Howeve

68、r, they were not optimise be-</p><p>　　cause higher values than 1.0% for NaBH4 and 0.6 mol dm?3 for HCl extinguished the plasma. No more critical variables were found for the hydride generation of selenium.&

69、lt;/p><p>　　Under the optimum conditions found (P = 1550 W,</p><p>　　F = 180 rpm or 4.75 mL min?1 , and G = 0.6 mL min?1 ), the</p><p>　　developed method shows adequate analytica

70、l performance for the direct selenium determination in electrolytic man- ganese.</p><p>　　優(yōu)化流動(dòng)注射氫化物誘導(dǎo)電感耦合等離子體光譜法，在電解錳中測(cè)定硒</p><p><b>　　摘要</b></p><p>　　在電解錳中以Flow-injectio

71、n-hydride產(chǎn)生Se被實(shí)驗(yàn)設(shè)計(jì)的方法進(jìn)行了優(yōu)化。對(duì)輔助變量如電力供應(yīng)(P)、樣品(F)、氬(G)流量、化學(xué)變量NaBH4和鹽酸濃度進(jìn)行了研究。在化學(xué)變量的情況下,得出的結(jié)論是,tetrahydridoborate鈉濃度</p><p>　　高于1.0%且鹽酸濃度應(yīng)該高于0.6摩爾dm 3撲滅了等離子體。</p><p>　　對(duì)影響的分析表明了所有的輔助變量都是是重要的因素且最優(yōu)條件是

72、P = 1550 W,F = 4.75毫升分鐘?1和G = 0.6毫升分鐘?1。對(duì)錳的影響進(jìn)行了專門研究并得出結(jié)論,如果錳低于2.0 g L?1 , 干擾是可以忽略不計(jì)的。在相同的意義上,銻的干擾(III)、砷(V)和水星(II)也被認(rèn)為是微不足道的。</p><p>　　最后,得到的檢測(cè)極限為0.0005%(w / w)(重復(fù)性R.S.D. < 2.0% Se濃度嘗試)。一些錳樣本也摻入了不同濃度的Se

73、(IV)結(jié)果展示與預(yù)期一致。</p><p><b>　　1．介紹</b></p><p>　　錳對(duì)治金行業(yè)有著非常重要的應(yīng)用。錳和鋁為食物，飲用水包裝，國(guó)內(nèi)工具，裝飾或其他包裝所需。用于鋁合金的錳是通過ELEC- trolysis制備，其中加入硒以改善電流效率。硒在陰極的共沉積導(dǎo)致了錳的污染。據(jù)認(rèn)為，超過90%的硒是用于提高電積電解錳的電流效率，進(jìn)入陰極中的元素形式

74、，因此，一些目前使用的電解級(jí)包含可變數(shù)量的硒（ 0.03-0.16 ％ ）。</p><p>　　在鋁爐的作用（ 700-800 ? C）下,硒從熔融鋁中蒸發(fā)為金屬蒸汽。由于高濃度的硒及其化合物的毒性和環(huán)境危害,煙氣排放和降落必須處于監(jiān)控之下。在這個(gè)意義上,在鋁 al- loying過程中,總硒濃度，原料和廢料必須也加以控制以確定鋁加工中硒環(huán)境歸趨和風(fēng)險(xiǎn)。</p><p>　　Hagels

75、tein研究了在鋁加工過程中硒的環(huán)境管理并指出主要的環(huán)境問題是鋁爐產(chǎn)生的浮渣和排放的顆粒。由于在加工過程中在設(shè)施里和廢物流中的硒集聚,鋁和鋼行業(yè)可以減少環(huán)境污染的途徑是避免或者控制含有硒的原材料使用。因此，不同的層次的硒的分析應(yīng)該在鑄房屋和冶金等行業(yè)實(shí)現(xiàn)發(fā)展 。</p><p>　　各種原子技術(shù)的運(yùn)用如(ICP–OES), ETAAS or AFS使得在水樣達(dá)到高選擇性和靈敏度。除此之外,冷蒸氣或氫化物發(fā)生前，以

76、電熱原子吸收檢測(cè)增加了這種選擇性減少光譜干擾引起的其他基質(zhì)成分。在過去的十年中，在線流動(dòng)注射（FI）的分離技術(shù)已在不同種類的樣品的測(cè)定微量元素愈來愈廣泛地應(yīng)用。</p><p>　　Hydride generation (HG)和inductively cou- pled plasma optical emission spectrometry (ICP–OES) 是一種廣泛應(yīng)用的測(cè)試硒的方法。然而,電解錳中硒測(cè)

77、試著作信息缺乏。此分析技術(shù)（FI- HG- ICP-OES 可以從樣品的主要成分分離硒,但是事實(shí)上,很容易從若干過渡金屬離子和其它揮發(fā)性元素出現(xiàn)干擾。發(fā)生在液相中過渡金屬的干擾對(duì)硒信號(hào)導(dǎo)致顯著的抑制作用。</p><p>　　具體而言，存在于氫化物發(fā)生條件試樣溶液的錳離子得到錳硼化物，其中吸收氫化物和導(dǎo)致其分解[11] 。錳物種可能催化還原劑以及衰減。Nakahara and Kikui的評(píng)論文章中記錄了這一化學(xué)

78、干擾。。然而，相互干擾氫化物形成元素的處理相對(duì)稀缺，只有少數(shù)作品討論可能的干擾機(jī)理[13,14] 。因此，必須要研究研究和不同的因素或影響氫化物發(fā)生過程變量的識(shí)別，以建立對(duì)硒的測(cè)定正確的分析方法。</p><p>　　因此，本研究的主要目的是找到的最佳操作條件，可以在電解錳中由FI- HG- ICP-OES進(jìn)行硒測(cè)定的簡(jiǎn)單和靈敏的方法。實(shí)現(xiàn)這個(gè)目標(biāo)的實(shí)驗(yàn)方法是實(shí)驗(yàn)設(shè)計(jì)。因?yàn)樗且阎?，?shí)驗(yàn)設(shè)計(jì)是最有力的方式以最小

79、的努力獲得我們需要的信息。在這個(gè)意義上說，一個(gè)完全析因設(shè)計(jì)能夠篩選各變量的影響，而復(fù)合材料的設(shè)計(jì)是用來建立響應(yīng)面。</p><p><b>　　2.2.儀器儀表</b></p><p>　　運(yùn)用感應(yīng)耦合等離子體光學(xué)發(fā)射光譜儀IRIS（Thermo公司）與CID檢測(cè)器。該儀器由一個(gè)多通道蠕動(dòng)泵及氣 - 液分離器組裝到光譜儀（T- PHD ）氣液分離器組成。使用196.0

80、90納米線的硒和無(wú)背景校正系統(tǒng)。光譜干擾的情況下被認(rèn)為是可以忽略不計(jì)，當(dāng)干擾信號(hào)的硒波長(zhǎng)的貢獻(xiàn)率為小于10％ [16,17]。</p><p>　　詳細(xì)的操作條件的等離子體激發(fā)和氫化物列于表1中。</p><p>　　2.3.電解錳的消化</p><p>　　電解錳（湘西州，中國(guó)）被收集在薄片并在實(shí)驗(yàn)室粉碎機(jī)里粉碎（ Retsch的S100）并篩分（ 500微米）。

81、得到的錳粉（2克）的酸性物質(zhì)立刻用濃HCl無(wú)需加熱[18]進(jìn)行實(shí)驗(yàn)。進(jìn)行這兩種步驟都需要水冷卻（致冷劑）。沒有它這個(gè)程序,兩個(gè)程序的所有樣品（圖1 ）中都會(huì)獲得相似的效果。因此，硒的分析測(cè)試測(cè)定（四）依賴于在消化步驟使用的的冷卻柱。</p><p>　　2.4.熱力學(xué)硒形態(tài)</p><p>　　硒的兩種常見的無(wú)機(jī)氧化態(tài)為IV和VI ，但只有較低的狀態(tài)（Ⅳ）在NaBH4作用下衰減為硼氫化鈉[

82、19]。因此，以驗(yàn)證鹽酸作用后硒的主要品種為IV 狀態(tài)是很重要的。</p><p>　　這個(gè)問題已經(jīng)有熱力學(xué)和實(shí)驗(yàn)證了（圖2a和2b ） 。圖。2a顯示了硒的分布圖對(duì)PH值的圖。 2b顯示硒對(duì)pH值和氧化還原電位的優(yōu)勢(shì)圖，用水母程序。</p><p>　　2.5.氫化物發(fā)生過程</p><p>　　對(duì)于硒的氫化物發(fā)生，樣品（以鹽酸</p><p&

83、gt;　　0.6摩爾DM- 3）和鈉tetrahydridoborate （1.0％）溶液是使用蠕動(dòng)泵引入到FI系統(tǒng)。然后將溶液通過一個(gè)氣液分離器中的連續(xù)流泵中的氣液分離器。所釋放的氫化物是由氣體載體（氬氣流量0.6毫升分鐘-1）載到液體中。讀取時(shí)間是30秒，沒有使用背景校正。</p><p><b>　　3.結(jié)果與討論</b></p><p>　　硒代的效率直接依賴

84、于硼氫化鈉濃度。然而，由于過度析氫比和濃度高于2.0 ％會(huì)導(dǎo)致對(duì)等離子體的消滅,硼氫化鈉濃度高于1.1％動(dòng)搖了等離子體放電。因此， 1.0％的硼氫化鈉的濃度被認(rèn)為是在所有分析的最高操作值來進(jìn)行。濃度需要保持恒定。鹽酸是已被發(fā)現(xiàn)的最令人滿意的用來生成的氫化物的介質(zhì)。</p><p>　　在鹽酸濃度高于的1 mol dm -3,硒化物的生成效率恒定。 （圖3） </p><p>　　然而，比的

85、1 mol dm - 3高的HCl的濃度產(chǎn)生一些有害的影響：減少泵繞組的生活，增加了。酸霧的毛病和按產(chǎn)品氫氣進(jìn)化增加的暴力事件。增加鹽酸濃度或達(dá)到高于0.6mol dm-3的峰值時(shí),氫化物產(chǎn)生效率略有增加。低濃度值（鹽酸< 0.4摩爾DM- 3 ）就足以產(chǎn)生在減少硒中完全的抑制作用。電解錳對(duì)測(cè)定硒（硒： 200微克的L- 1 ）中下也被探討的濃度范圍為從0 到?6.0克L- 1錳。該氫化物發(fā)生效率（ ％R ）定義為：其中深交所與干

86、擾硒的強(qiáng)度得到和SSeo是硒與干擾的強(qiáng)度。在一般情況下，錳的濃度大于2克L- 1對(duì)于硒反應(yīng)有負(fù)面影響（圖4a） 。這種抑制作用通過增加鹽酸濃度是可以克服的，但兩個(gè)假設(shè)考慮在內(nèi),最佳鹽酸濃度固定為0.6摩爾dm-3：低濃度的HCl不太激烈，價(jià)格低廉，錳含量不會(huì)高于2克L- 1。</p><p>　　最后，其他氫化元素干擾,像銻（Ⅲ），砷（ V） ，汞（二）,在硒（四）的測(cè)定進(jìn)行了系統(tǒng)的研究。它在文獻(xiàn)[14]中已被證

87、明該氫化物發(fā)生干擾不依賴于被分析物對(duì)干擾物的比例,而是在用于測(cè)量的溶液間中的干擾濃度。相對(duì)靈敏度（％R）,即用不同濃度的砷（ V）而獲得的信號(hào)之比，銻（Ⅲ）或汞（II）和沒有這些干擾，被描繪在圖4b 。只有當(dāng)砷（ V）的濃度水平高（2毫克的L -1），能對(duì)硒信號(hào)進(jìn)行觀察。因此，在硒響應(yīng)那些揮發(fā)性元素的影響在本例中沒有被考慮。</p><p>　　反應(yīng)速率為氫化物是由幾個(gè)相互依存變量來控制： （I）中的NaBH 4

88、作為還原劑的濃度，樣品（II）的酸的濃度，（III）該氫化物形成的化學(xué)結(jié)構(gòu)和（IV）該氫化物形成元素的的氧化態(tài)。一方面，如上所述，由于實(shí)驗(yàn)的限制將NaBH 4和HCl濃度沒有包括在最優(yōu)化中。另一方面，硒（IV）是在從電解錳鹽酸消化得到的解決方案的主要氧化態(tài)。因此， FI- HG- ICP-OES測(cè)定硒采用水性標(biāo)準(zhǔn)溶液進(jìn)行了研究和摻入48和80微克的L- 1的硒（ Ⅳ）ELEC-trolytic錳粉 。</p><p&

89、gt;　　3.1. 優(yōu)化硒檢測(cè)：因素設(shè)計(jì)</p><p>　　幾個(gè)變量，如：供電（P ），樣品流速（ F）和輔助氬氣流量（ G）可能會(huì)影響到亞硒酸鹽檢測(cè)。為了找到影響檢測(cè)步驟的主要因素，一個(gè)兩層的完全析因設(shè)計(jì)（23 +4 replicates of central point）通過使用“整序”進(jìn)行了使用（迷彩，作為挪威， v.7.5 ）[23]。表2列出了實(shí)驗(yàn)設(shè)計(jì)矩陣和所得到的響應(yīng)（強(qiáng)度，平均三次測(cè)定）。效應(yīng)的分

90、析，通過使用'整序'方案，得出結(jié)論，所示的所有變量的響應(yīng)（對(duì)水平<0.05）作用顯著 ，特別是G和P。</p><p>　　受變量的影響,這個(gè)實(shí)驗(yàn)的設(shè)計(jì)延伸至中心復(fù)合設(shè)計(jì)。以這種方式，有可能建立響應(yīng)面，并獲能夠定義最大反應(yīng)的儀器條件。成熟的CCD實(shí)驗(yàn)設(shè)計(jì)利用了“The Unscrambler”程序。表3顯示出了中心復(fù)合設(shè)計(jì)連同反應(yīng)同時(shí)獲得。響應(yīng)面以“The Unscrambler”程序進(jìn)行計(jì)

91、算 ，并且當(dāng)F和P2分別在p = 0.07和p = 0.052 時(shí)效果顯著時(shí),G和P變量的效果也被證實(shí)顯著 。</p><p>　　如果F變量被固定在CCD的最高水平（ 180 rpm or 4.75 mL min?1） ，可以將響應(yīng)</p><p>　　表面作為最有意義的變量的作用繪制出來，如圖5所示 。從這個(gè)繪圖上，我們可以推斷出在因素空間內(nèi)最佳條件，換言之沒有任何外推。因此，硒能夠被

92、測(cè)定出來測(cè)定通過電源1550 W和氬流量。</p><p>　　由于在電解錳中的硒（四）沒有任何被證實(shí)的參考文獻(xiàn)資料，在我們的實(shí)驗(yàn)室制備的合成硒（ IV）的測(cè)試解決方案和不同的實(shí)際樣品加標(biāo)48和80微克的L- 1硒（ IV）被用來檢查分析方法的精確度（表4）。如可以看出，基體效應(yīng)是可以忽略的，因?yàn)樗玫降奈鴿舛仁窍嗤闹苯雍图獯膛卸?。最后，檢測(cè)限，定義為空白信號(hào)+3 SD ， SD哪里是五次測(cè)量的空白的標(biāo)準(zhǔn)偏差，

93、估計(jì)為0.0005 ％。方法（ RSD）的五個(gè)重復(fù)測(cè)量精度低于</p><p>　　所有濃度2.0 ％。</p><p><b>　　4.結(jié)論</b></p><p>　　利用實(shí)驗(yàn)設(shè)計(jì)表明，所研究的三個(gè)變量（供電，氬氣流量和泵流量）對(duì)測(cè)定無(wú)機(jī)硒很重要。鈉tetrahydridoborate和鹽酸也是影響明顯的變量。然而，這兩種物質(zhì)并不是絕對(duì)有影