Comparison of Two Laboratory/Computer Interfacing Systems for...
2 downloads
103 Views
6MB Size
Comparison of Two LaboratorylComputer Interfacing Systems for Temperature and Absorbance Measurements L. H. ~ e r k .a ?W. J. Clark. and D. C. White Worcester Potyiecnnic lnstute Worcester, MA 01 609
In considering incorporating laboratorylcomputer intcrfacing experiments into our Gcncral Chemistry program, we became intrirued hv what differences there might be between the relatively-simple and inexpensive temperature/colorimeter system available from Project SERAPHIM' and the more ornate and expensive CHEMPAC3 system. The results of our comparison including pedagogical considerations are reported herein. Experimental Method Apparatus Under the supervision of SERAPHIM workshop leader P. A. Girard,4 each co-author constructed a n IBM-PC adapter hox, a thermistor probe, and a Hlocktronic I colorimeter, at a fee of $4O/participant for materials. Our estimate of the cost of the comoonents is included in the Itrsults section below. One of the adapter boxes, two of the thermistor probes, denoted below a s GCO and GC1, and one of the colorimeters were used. Spectronic 20 test tubes were used as sample cells for the SERAPHIM system. These tubes had a measured internal diameter of 1.065 cm. Acomputer interface board, which is included with the CHEMPAC system, was purchaseds for the SERAPHIM system for $61. Mr. Girard also supplied us with a copy of the SERAPHIM 1203 IBM working disk. The cost of the complete IBM-PC CHEMFAC package, which included a n interface module, student experiment set, and teacher's package, was $2090. The CHEMPAC thermocouple, two thermistors-denoted below by TC, TM1, and TM2, respectively-and colorimeter were used. The colorimeter consists of a sealed unit containing three light sources/sensors, denoted a s GREEN, YELLOW, and RED. The CHEMPAC system also includes interfacing devices for measuring emf, pH, and pressure, which were not used. The CHEMPAC student emeriment set orovided the H e thermometer, which served a's the tempeiature standarg the olastic calorimeters. cover and stirrer assemblies. and plas'tic 1.00-cm cells used i n the CHEMPAC absorbance measurements. l b o portable AT&T IBM-compatible PC's were available in our dcoanment: one was dedicated to the SERAPHIM system, tke other tb the CHEMPAC system. Comparison absorbance measurements were carried out using a Shimadzu 2100U W-Vis spectrophutometer having its own computer interfacing system. The total cost of the Shimadzu system including software was $13,572. Research grade matched 1.00-cm cells were used with the Shimadzu instrument. 'Corresponding author. '~epartment of Chemistry, University of Wisconsin-Madison,1101 University Avenue, Madison, WI 53706. 3Manufactured by E&L Instruments, 70 Fulton Terrace, New Haven, CT 06512; purchased from Electronic Marketing Co., 1092 Johnson Road, Woodbridge, CT 06525. 4Director of Science, Salem High School, Salem. MA01970. 'Gamecard Ill Plus game control adapter made by CH Products. 1225 Stone Drive, San Marcos, CA 92069 and purchased from a local vendor. 'E&L Instruments, personal communication, 1991.
Unlike the Shimadzu instrument, both the SERAPHIM and CHEMPAC culorimetcrs involve a sinelc beam measurement. The two devices reviewed here Zffer electronically in two major respects ( I ) .
-
Thc SERAPHIM systpm 12). hcrause 11 uses a gnrneport, senses only changes in elenncal reststance. The rwitance of the thermistor or the photocell, either serving as one arm of a bridge circuit, is determined by comparing the voltage drop across that arm with the drop across a fured resistance.
The system uses the 5-V power line of the computer itself. The CHEMPAC svstem has a more sonhistieated. indeoendently powered m"odule that measures'the resis&ce of the thermistor, but measures the EMF produced by the photocell. B o t h systems use an AID converter,but the SERAPHIM system utilizes the 8-hit converter of the gamepart, while the CHEMPAC module has a 12-bit~onverter.~ This notation refers to the size of a counter on the AID chip. The voltage range that can be measured is determined by the total hardware of the system; this range is divided into Zn intervals to accommodate the n bit counter. Thus an &bit converter divides the range into 256 discrete segments, each representing 0.39% of the total range. The segments for a 12-bit converter represent 0.024% ofthe range. However, since the zem state ofthe munter always represents the starting paint of zero voltage, the AID chip can return only 2" - 1values of measured voltaee. that is. uo to (2" - 1)/2nof the total ranee. The simplest &conve&; is the binary ramp type. H&, the counter is zeroed, then incremented by one as the output voltage is increased by one discrete interval. This voltage is compared to the input voltage, and the process repeated until the output voltage equals or exceeds the input. The number in the counter is communicated to the computer where the software converts it into a physically meaningful quantity. Temperature Measurements
Temperatures given by the thermometer were checked both in ice water and boiling water a t the ambient atmospheric pressure of 746.9 ton. The two SERAPHIM thermistors were calibrated according to thermistor menu item 1of the SERAPHIM software simultaneouslv aeainst the thermometer in a calorimeter of heated o; coked water a t seven temperatures rangingfrom 10.0 to 78.1 T.After similarly calibratingthe CHEMPAC thermocouple a g a i n s t t h e thermometer (CHEMPAC Experiment 3; five temperatures, 0.4-88.6 'C), the two CHEMPAC thermistors were calibrated simultaneouslv against the thermocou~le(CHEMPAC Emeriment 12; five temperatures, 0.1-88.6 'C). In all cases the calibration files were saved to disk. To simulate actual experiments and to compare the results obtained from the five interfaced sensors. three runs were made in which a container of water (R& 1in a calorimeter, Run 2 in a beaker i n air, and Run 3 in a n ice bath) was sllowcd to cool. For each run, after readmg in the prevlously obtamed callbratmn files, snnultaneous tempcraturc mcasurcmcnts werc recorded for the five probes vla computcr (SERAPHIM t h e r m i s t o r menu ircm 3: CIIEMPAC. Ytllltv Dlik menu ltcm C) and bv hand for thc thermometkr a t i0-s intervals. T ~ ~ ' s E R A ~ ' H I probes M were connected to one computer, the CHEMPAC probes to the other. The following temperature ranges were covered: Run number Temperature Range Total Time of Run
('c)
(min)
In making the calibrations and measurements the probe sensing ends were clustered around the thermometer bulb and held in place using a rubber band; constant manual Volume 69 Number 11 November 1992
891
stirring was provided using the CHEMPAC stirrer wire loop. Care had to be taken in usina the SERAPHIM svstem a t lower temperatures. i t was found that if a thermistor sienal rose above the value 252, i.e., exceeded the maximum of measurable values (corresponding to about 4.1 and 1.0 'C with GCO and GC1, respectively), the program aborted and all previously collected data were lost.
Thermometer (.) & GCO (.) vs. Time
Absorbance Measurements Neither the SERAPHIM colorimeter nor the Shimadzu spectrometer require calibration. Since t h e CHEMPAC colorimeter must be calibrated against solutions of copper(I1) nitrate, Cu(NO&, all of our absorbance d a t a were obtained using solutions of this reagent. 0 50 100 150 200 A stock solution of Cu(N0&3Hz0 was prepared Time (rnin) in 0.10 M HN03. The salt was found not sufficientlv soluble in CHEMPAC's suggested Figure I . Thermometer (-) and GCO (0)vs. time 0.01 M HNO-. The stock solution was stored for several days, then filtered. Tenmilliliter aliquots of a 251250 mL 0.10 M HN03 in the reference cell and stored on disk for dilution of this solution were standardized against 0.09975 future reference. M NaaSa02. . . . which had been ~reviouslvstandardized Results against weighed samples of K I (3). ~ ~~h " estock solution containine 2.093 M lCuZ+lwas diluted with 0.10 M HNO. ~" Temperature Measurements using an&tical p r & i s i ~ ito give an additional 10 solutions in the range 0.05233-1.570 M [CuZ+l.These 11soluThe Hg thermometer gave wrred readings at 0.0 and tions were used in SERAPHIM colorimeter Runs 3 and 4, C to within f0.1 72. 99.5 O CHEMPAC Runs 3 (calibration), 4 and 5. and Shimadzu The SERAPHIM calibration data were independently fit Runs 3 and 4, the results of which are reported here. Runs for each thermistor by the SERAPHIM . Droaam with an 1 and 2 were practice runs carried out using solutions of equation of the form Cu2+made by dilutions of lesser analytical precision. T ( K )= ar" + b SERAPHIMRuns where r is the measured signal anda, b, and n are presumably best-fit constants that were unioue to each thermistor ~ ~ - - ~ Each SERAPHIM run was carried out as follows. ARer (n 4 . 2 3 , a 280, and b 200). ~ d i form s was provided reading the solution blank with 0.10 M HNOBin the wlorin screen documentation given a t the beginning of the imeter (Blocktronic menu item 2),10 readings were taken thermistor calibration and was found to fit the observed on each solution (menu item 4)at l-s intervals and averthermometer temperatures to a n averaee of about f0.2 T. aged No sense could be made of a calibration equation of a different form given in the SERAPHIM comDuter outout at CHEMPAC Runs the conclnsi& of the calibration. The CHEMPAC thermowuple calibration data were fit The CHEMPAC sensors were calibrated following the to withink0.1 'C by the CHEMPAC program with anequaStudent Disk menu item Experiment 19 procedure, which tion of the form included a baseline determination with 0.10 M HN03. Solution absorbances were obtained followine --- the ~ - Utilitv -~ - ---. ~ !Tc)=A+Bx+cx~+D~ Disk menu item F instructions, which included loading th;! where X i s the signal and A-D are best fit constants. We ~reviouslvdetermined and stored calibration file. As in the obtained the same wnstants of fitTbetween signal and obLalibration, 10 readings were obtained in succession for served T, with a correlation wefficient of 0.9999953. the baseline and each solution which were then averaged by the computer. 'wherever possible, curve fitting was done independently by the co-authorsusino the method of least sauares ,~~~ emolovino . , ~the ~"CIlRVF -.~. Shimadzu Runs FITTER-PC programobtained from heraciive Microware, Inc., PO. For each Shimadzu run, the baseline was zeroed using Box 139, State College, PA 16804and ProPiot obtained from Cogent Software, 1030 Tine Street, Menlo Park. CA 94025. The latter son0.10 M HNOBin each cuvette. The spectrum of each soluware pachage was used to obtain the Figures for this paper. tion was then recorded over the range 400-800 nm with
.
-
-
-
-
~
~~
892
Journal of Chemical Education
~7
~
-
-
The CHEMPAC thermistor calibration data were fit to within k0.1 ~C by the CHEMPAC program with an equation of the form
Temperature Deviations
l/T(K) = A + BX
AT (Sensors - Hg Thermometer)
Figure 2. Temperature deviations.
I I
I I
SHIMADZU Absorbance & CHEMPAC Signal Data
(€4SHIMADZ", Runs 3&4, Solns. 1-11
(b) GREEN Sensor, Run 3
Wavelength (nm)
[Cul (moledL)
(c) YELLOW Sensor. Run 3
(d) RED Sensor, Run 3 1 1 1 1 3 1 1 1 1 1 1 1 1 1 1 1 1 . 1 1 . 1 1 1 1
+ CX' + D*
whereX = In (signal) and A-D are best fit constants. Our fit of the data gave significantly different values of the constants A-D (both correlation coefficients on the order of 0.999995) but gave calculated values virtually identical to those calculated using the CHEMPAC constants. Plots of thermometer and SERAPHIM thermistor GCO temperatures as a function of time for Runs 1 3 are given in Figure 1. Both SERAPHIM thermistors exhibited the following remarkable and unique behavior, which is most noticeable in the Figure 1expanded Run 1 plot, less so in the Run 2 plot, and absent in the Run 3 plot. As the thermometer readings decreased, the temperature shown by a thermistor remained constant below the thermometer temperatures for several 30-s intervals. Then, when the t e m ~ e r a t u r ereeistered bv the t'hermometerkas reduceh to the temperature registered by the thermistor, the latter temperature dropped to a new constant value. This problem decreased in severity with decreasing temperature and disappeared below about 30 'C. The change in resistance of a thermistor for a given change in temperature is not constant but increases as the temperature decreases ( 4 )(cf. the above calibration eauations). In the cases of unusual behavior a t the higher temperatures, a change of a degree or more causes the change in voltage drop resulting from the change in resistance t~ be much less than the size of the discrete interval determining the resolution of the converter. Thus the counter doesn't change until the temperature changes by a relatively large amount. The sensitivity to changes in temperature increases with decreasine temoerature until the effect disappears when a change in temperature on the order of tenths of degrees
-
I Figure 3. Shimadzu absorbance and CHEMPAC signal data.
Volume 69 Number 11 November 1992
893
causes a change in voltage drop greater than this interval. CHEMPAC with its smaller voltage intervals does not manifest this problem. Deviations of sensor temperatures fmm the thermometer temperatures are shown in Figure 2. Both SERAPHIM thermistors exhibited a large spread of both negative and positive deviations a t higher temperatures, corresponding to the anomaly noted above. This spread narrowed as the temperature decreased. At the lowest temperatures the deviations were all positive. The three CHEMPAC temperature sensors exhibited mutually similar behavior (TM1 and TM2 remarkably similar to each other), giving mainly positive deviations a t the higher temperatures and negative deviations at the lower temperatures. I t might he argued that the temperature change exhibited by a sensor over a range of temperatures is a better criterion ofjudgement than is its actual temperature value t time. The com~arisonof these chanees a t anv ~ o i n in among {he various sensors is &en in Table 1. The relative aereement between of the thermometer and that of the sensors is:
Table 1. Decreases in Temperature Recorded by Thermometer ( 0 6 s ) and Sensors in Runs 1,2, and 3 Over Selected Temperature Rangesa
TRange OBS
Maximum Deviation
Temperature Range
CHEMPAC thermistors
iO.2 'C
70-5.C
SERAPHIM thermistor GCO
M.2 'C
55-5 'C
CHEMPAC thermocouple
M.4 'C
all measured temperatures
SERAPHIM thermistor GC1
M.4 'C
5 5 5 'C
Absorbance Measurements
The absorption spectra obtained with the Shimadzu instrument ofthc 11 solutions ofCutNOv?used in this studv I of these spec& are shown in Figure 3a. The general && is in agreement with a spectrum published elsewhere (5). Signal readings obtained in the calibration of the three CHEMPAC sensors are plotted against [CuZ+lin Figure 3 M . We fit the data points for each sensor with an equation of the form Signal = A + B [ C U ~+ ]C [ C U ~++D[cu2'13 ]~ +E[cu~+]~ to produce the smooth curves shown, with correlation coefficients of 0.999992 (GREEN,. 0.999994 (YELLOW!. and 0.567 (RED). During the calibration no signals were obtained from the RED sensor for solutions 9-11. The final printout also omitted the signal previously given for solntion 8. Apparently, the CHEMPAC calibration system cuts offtreatment of the higher signals obtained with the RED sensor with solutions having concentrations starting somewhere in the range 0.5233-0.7849 M [CuZ+l. These calibration data were also fit for each sensor by the CHEMPAC program with equations in the form Y=A+BX+CX~+DP+EX' whereX= sipnal and Y = absorbance. WithX= the calibration signal readings, the u,-authors calculated values of Y using the CHI.:MIIAC equations. These points are plotted for each sensor in Figure 4a and b. he smooth~urves shown in these plots were obtained by fitting these data with equations of t h e same form a s given above by CHEMPAC (all three correlation coefficients on the order of 0.999999). As indicated by the plots in Figure 4a, the GREEN andYELLOW sensors give similar smooth curves in the region in which the data points overlap. The shapes of the smooth curves a t higher concentrations with the YELLOW and RED sensors are artifact of the polynomial curve fitting function. 894
Journal of Chemical Education
GCO
GCI
TC
TMI
TM2
Run 1
-
Sensor
OBS
4.5 3.9 5.0 4.0 4.4 4.5 5.0 4.7
2.7 4.7 5.9 3.3 5.6 4.3 5.3 5.0 Run 2
2.2 4.6 5.1 2.3
2.0 4.4 5.3 2.4 Run 3
3.7 3.6 4.6 4.6 4.4 0.1
3.5 3.6 4.4 4.6 4.2 0.1
-
a Wherever possloe, cnosen from ooservea thermometer oat8 in the range 01 the s.ccessve 5 C nsrvan 79 9 75 0 74 9 70~0 69 9 65 ~ 0 atc . ~ 'For example, n R m 1 a10 50 ana 3 50 mnms tne remrow lernperatdres were: 79 1-75a 'C (OBS). 784-73.9(GCO,. 78.0753 GCt,. 79.2-756 ,TCJ. 78.3-756 ITMI) ana 78 6-75.8ITMZJ. to gwe tne ,esJ!s snown n t h s raw ~~
~
Our values of the best fit constants A-E are of the same magnitude, but not equal to those, given by CHEMPAC. These discrepancies can be accounted for by assuming that the CHEMPAC program contains absorbance data as a function of [Cu2+1for each sensor (differing somewhat from the fit values) which i t uses to derive these equations. The RED sensor data plotted and fit in Figures 3d and 4b require further comment. As shown in Figure 3d abnormally high signal readings of 1431.5 and 488.1 were obtained with the RED sensor a t [Cu2+1= 0.1570 and 0.2616 M, respectively; a smooth curve placed through the remaining points would resemble the smooth curves given for the GREEN and YELLOW sensors in Figure 3b and c. Iuclusion of these "bad points" gives the calibration equation for the RED sensor plotted in Figure 4b which has the following unsatisfactory characteristics: (1)The order in which the points appear in Figure 3d is not the same as the order in which they appear in Figure 4h-
we expect the same order, since an increase in signal should correspond ta an increase in absorbance. (2) Consequently, the equation gives multiple (usually two) signal values corresponding to a single value of the absorbance. For these reasons the additional results given below for the RED sensor are probably not an accurate reflection of the quality of data that can be obtained using this sensor. Absorbances measured with the Shimadzu instrument a t 588 and 680 nm (uide infm) and with the SERAPHIM (raw absorbances divided by 1.065 to make them compara-
~
~
Im
CHEMPAC Calibrations & SHIMADZU MeasuredIBest Fit Absorbanoes (b) RED Sensor. Run 3
(a) GREEN and YELLOW Sensors, Run 3
.
YELLOW
0.5
0
0
1WO
5W
1500
Signal
(d) SHIMADZU Runs 3 & 4 (680 nm)
(c) SHlMADZU Runs 3 & 4 (588 nm) 1.5
B
S
g 9
1.0
0.5
ble to the others which were measured using 1.00-cm cells) and CHEMPAC sensors are given in the upper portion of Table 2. These data were fit with linear equations in the form of A =slope [CU~+I + intercept The slopes, intercepts, and correlation coefficients of these fits are given i n the lower portion of Table 2. The measured absorbances and straight lines obtained are also plotted in Figure 4c and d and in Figure 5; each graph includes plots of two data sets and two fitted lines. The degree to which each pair of lines in a given graph superpose can be taken as a measure of the precision obtained with the corresponding instrument or sensor. Using this criterion, the order of precision is: Shimadzu (both at 588 and
0
0.5
0
1.0
1.5
2.0
0.1
0.2
0.3
O*
0.5
p u ] (moleqL)
[Cul (molesIL1
-
I Figure 4. CHEMPAC calibrations and Shimadzu measuredmestfit absorbances.
I
Measured & Best Fit Absorbances vs. [Cu] (b) CHEMPAC GREEN Runs 4 & 5
(a) SERAPHIM Runs 3 & 4 0.8
"
$ 0.5
r
-
O.4W
c
e2
0.4
:
0.2 0 0
0.5
1.0
1.5
2.0
0
0.5
1.0
1.5
2.0
[Cul (maleslL)
[cu](moles/L)
(c) CHEMPAC YELLOW Runs 4 & 5
(d) CHEMPAC RED Runs 4 & 5
0
2
1.0 0.5
0.5
1.0
1.5
2.0
[Cul (moieslL1
Figure 5. Measured and best fit absorbances vs. [Cu]
0
0.1
0.2
0.3
[Cul (molerll)
680 nm) > CHEMPAC GREEN > SERAPHIM CHEMPAC YELLOW > CHEMPAC RED.
0.4
0.5
The degree to which Beer's is obeyed is Law, A = E [CuZ+1, obtained by t h e degree to which the intercepts and correlation coefficients given in Table 2 correspond to 0 and 1.000..., respectively. Using these criteria only the CHEMPAC GREEN sensor gives data satisfactorily following Beer's Law. Estimates were obtained of the wavelength associated with eacb interfacing sensor by the following procedure. For each concentration within a sensor run, the sensor absorbance was matched against t h e corresponding stored spectra from Shimadzu Runs 3 and 4 to give two values of 2. that gave this same absorbance. Extension to the second sensor run and additional 11 concentrations yielded 44 values of the matching 2 s for eacb sensor. These values were then averaged to give the values and standard deviations listed in the lower portion of Table 2. At each of these wavelengths t h e Shimadzu absorbances given in Runs 3 and 4 were averaged. Figure 6 shows the comparison between the measured sensor
Volume 69 Number 11 November 1992
895
Table 2. Measured Absorbances and Some Calculated Parameters Shimadzu A, 588 nm
Shimadzu A. 680 nm
SERAPHIM h1.065
[CU~'] Run 3
Run4
Run3
Run 4
Run3
Run4
0.05233 0.031 0.1047 0.064 0.1570 0.097
0.032 0.064 0.098
0.253 0.510 0.774
0.253 0.509 0.775
0.055 0.085 0.162
0.026 0.082 0.090
Run 3
Run4
Shirnadzu:
588 nm 0.7636 0.7656 4.037 4.037 0.998 0.998
h Slope, r Intercept Correl. Coeff.
680 nm Slope, E Intercept Correl. Coeff.
5.277 5.292 -0.045 -0.047 0.99985 0.9998
SERAPHIM: Slope, r Intercept Correl. Coeff. Estimated h
Run 3
Run 4
0.5621 0.034 0.997 588 f
0.5583 0.014 0.997 9 nm
0.026 0.049 0.067
Run 5
Run 4
Run 5
0.021 0.044 0.064
0.032 0.063 0.101
0.030 0.069 0.096
0.314 0.620 0.825
0.554 0.617 0.873
Run 4
Run 5
Run 4
Slope, t Intercept Intercept Estimated h
0.6446 0.5708 0.008 0.035 0.9987 0.988 58723nm
Slope, e Intercept CorreL Coeff. Estimated h
$ 7.80
5.94
iU4 27.02
Also, the w s t to outfit a mom with the necessary tools to vermit 32 students Der lab oeriod to construct and test ihese devices with m/nimal waiting time is estimated to bc to be $513.08. The tools are reusable items. manv of which are already available in chemistry departments, e.g., four multimeters. Given our cost estimates, the Project SERAPHIM charge of $40/participaut for materials is reasonable.
Discussion Based upon our results and our previous experience, neither the SERAPHIM nor CHEMPAC sensors give accuracy
A
Run4
CHEMPAC. YELLOW:
CHEMPAC, RED:
RED
Run 5
0.3909 0.4007 0.004 -0.001 0.99992 0.9996 572 2 4 nm
Using current prices of items purchased in bulk, the cost of the exvendable items to make the comwnents are estimated to-be:
Journal of Chemical Education
Run 4
Slope, r Intercept Correl. Coeff. Estimated h
Cost of SERAPHIM Apparatus
896
CHEMPAC YELLOW A
A
CHEMPAC, GREEN:
ahsorbances and the best fit lines obtained from the averaged Shimadzu data at the correspondmg estimated X's.
thermistor probe wlorimeter adapter box total cast
GREEN
Run 4
Run 5
Run 5
2.082 3.934 0.385 0.195 0.725 0.861 680 2 31 nm
or precision of temperature or wlorimeter data rivaling those obtained using classical devices, e.g., Hg thermometers and Spectronic 20's. The SERAPHIM system is attractive from the standpoint that, rather than be presented with "black boxes", students can construct their own devices. They also then have another stake in the quality of results they obtain using these sensors. Even if students use devices constructed by others, they have a chance to look through the glass sheath of the thermistor probe or open up the colorimeter and adapter box to see how they are constructed. The light intensity of the colorimeter can be varied and the effect noted. On the other hand, the CHEMPAC sensors are sealed commercial units, with the colorimeter even black in color. The fact that the CHEMPAC wlorimeter has to be calibrated against a specific compound, solutions of which are difficult to standardize, is a disadvantage. Furthermore, our Shimadzu results cast doubt a s to the applicability of Beer's Law to solutions of Cu(NO& in HN03 a t higher wncentrations. The bad points obtained in the calibration of the RED sensor may be innate to the CHEMPAC system or due to our experimental error. This erratic behavior cannot be caused by errors in concentration since the data obtained
simultaneously with the other two CHEMPAC sensors are well-behaved. The performance of the RED sensor needs to be checked out by making additional calibrationlabsorbance runs using a greater number of solutions giving signals smaller than the sensor cut-off.
Comparison of Absorbances with SHlMADZU Best Fit Ave. Abs. vs. [Cu] (a) SERAPHIM Runs 3 & 4
(b) CHEMPAC GREEN Runs 4 & 5
Recommendations We feel that exoosure of General ~ h e m i s tstudents j to lawcom~uterinterfacin~is a worthwhile endeavor and useofthe SERAPHIM system (c) CHEMPAC YELLOW Runs 4 & 5 (d) CHEMPAC RED Runs 4 & 5 should be given serious consideration. Use of t h e CHEMPAC system for this 1.25 SHIMAOZC 587 nm purpose is not justified given 1.00 our results and the much greater expense incurred. In h 0.75 large course the cost of havD ? 0.50 ing each student construct all three SERAPHIM comoo0.25 nents is also not trivial. How0 ever, using our cost estimates 0 0.5 1.0 1.5 2.0 above, some compmmise can ICul (maleslL) be made in which some constructed components are passed down from student to Figure 6. Comparison of absorbances with Shimadzu best fit average absorbance vs. [Cu] student, year to year. It would be worthwhile to purchase one complete CHEMPAC unit and use this as an Council of WPI. The AT&T PCs were by WPI's adjunct to the lecture part of a course in which the comOfflce of Academic Computing. puter screen output can be pmjected to a large screen for viewing by students. The additional emf, pH, and pressure Literature Cited components would also be available for use in this mode. 1. Carr, J. J hi~e-mputer ~nterfming~ n ~ d b A~I D k B. DIA; T S ~ BOO^ BI"~
a
Acknowledgment We thank Nicholas K. Kildahl for many helpful comments regarding the manuscript. William D. Hobey, Alfred
Ridge Summtt, PA, 1980. 2. J . Chem Ed=: Soffunm: 1888,IA.(2). 3. Skaog, D . A.: Wesf D. M.Fvndamenfals ofilnolytlml Chemistry, 4th d.; Saundera: NewYork 1 9 8 2 ; 767-768. ~~ Our standardizstionofCu2*waga~adaptatbnofthe method *"en in this reference for the standardhation of NasS20s against Cu metal.
Volume 69 Number 11 November 1992
897
