- Email: [email protected]
Kinetic detection of the end-point in titrations involving slow reactions
Talanta,Vol. 30, No. 3, Pp. 14>149, 1983 Printedin Great Britain.All rightsreserved
0039-9140/83/030145-05$03.00/O Copyright0 1983PergamonPressLtd
KINETIC DETECTION OF THE END-POINT IN TITRATIONS INVOLVING SLOW REACTIONS DIRECT TITRATION OF POLYHYDROXY-COMPOUNDS WITH PERIODATE C. E. EKSTATHIOU and T. P. HADJIIOANNOU Laboratory of Analytical Chemistry, University of Athens, 104 Solonos St., Athens 144, Greece (Received 3 April 1982. Revised 24 September 1982. Accepted 4 October 1982)
Summary-A new titration technique is described in which the end-point is determined by measuring the relative reaction rate of the titration reaction. This technique is adequate for rather slow reactions where conventional direct titrations are not applicable. The titrations are done automatically under microcomputer control. The efficiency of this technique is demonstrated with direct titrations of certain polyhydroxy-compounds with standard periodate solution. Ethylene glycol and propylene glycol (0.05-0.3 mmole), glycerol (0.06-0.17 mmole) and mannitol (0.01-0.03 mmole) were determined with average relative errors of 0.1-0.3x.
T in the titrated solution. The following sequence of events (illustrated in Fig. 1) will take place:
A common reaction titrators
requirement for direct titration is that the should be fast, especially when automatic are used. If the reaction is slow, it is custom-
ary to add excess of titrant, and back-titrate, but this increases the overall .analysis time and decreases the accuracy. The end-point is generally detected by observing a change in an indicator colour or some other parameter such as potential, polarization current or conductivity. During the last decade titration techniques have been developed in which kinetic parameters are used indirectly to locate end-points. In these techniques, known as catalytic titrations, the titrant and the titrand react stoichiometrically and the first excess of the former catalyses an indicating reaction that is used to locate the end-point of the titration.’ In this paper, we describe a titration technique which is applicable when rather slow but stoichiometric reactions take place between the titrant and titrand. This technique requires microcomputer control and the end-point is automatically detected when certain preselected requirements of a kinetic nature have been fulfilled. As a practical application of this technique, organic polyhydroxy-compounds have been directly titrated with a standard periodate solution. PRINCIPLEOF THE METHOD Compound C reacts rather slowly but stoichiometrically with compound T and is to be titrated with it according to the reaction C+ T2
products
(1)
Suppose there is a transducer, with sufficiently fast response, capable of monitoring the concentration of
(i) An initial volume, V.,,, of solvent or buffer is transferred into the titration cell. (ii) A predetermined volume, V,, of titrant of concentration ml0 is delivered automatically to form a “base” titrant, with concentration [Tl, which is given by
[Tie= vo[TM vw+ f’o)
(2)
(iii) The corresponding “base” signal of the transducer, S,, is measured and stored in the computer memory. (iv) A sample volume, V,, containing the whole amount of C to be titrated, is transferred into the reaction cell, to give an initial concentration [Cl,. (u) The computer continuously receives the signal S from the transducer, and from it calculates and stores in memory the initial reaction rate, R,, (= 1dS/dt I), in arbitrary units, which is proportional to -dm/dt:
-d[Tlldt = k[ClJ%
(3)
(vi) The burette is then actuated, and automatically delivers titrant until the signal from the transducer returns to the base signal S,. Reaction (1) takes place during this addition, consuming T until the delivery rate surpasses the consumption rate and the concentration of T returns to [T],. The total volume of titrant delivered, V,, is stored in memory. To avoid excessive overshoot of S,, the delivery rate becomes progressively smaller as the signal S approaches &. (vii) The computer, as in step (v), calculates the
new reaction rate, R,, which will be less than the 145
146
C. E. EFSTATH~OU and T. P. HADJIIOANNOU
r
I
measurements
I
of reaction rate -,
-v,+i
-addrtlons
of tltrant-
A
:
V
TIME
Fig. 1. Graphical presentation of the sequence of events taking place during a titration with kinetic detection of the end-point. The scale of the time axis is not uniform along its length. VO,V,, . . . V, denote the time periods during which the titrant is delivered. R,,, R,, . R,,denote the time periods during which measurements of reaction rate are taking place.
initial rate &, because part of the titrand
C has been
consumed. Steps (ui) and (vii) are repeated with addition of volumes V2, V,, . . . , V, and calculation of the corresponding reaction rates R,, R3, . . . , R.. The titration is terminated when the measured reaction rate R, after the n th delivery of titrant is equal to or less than a predetermined fraction of R,,, typically O.OOl&, which denotes that 99.9% of the titrand originally present has reacted, i.e., practically quantitative titration. Calculation titrant
of the corrected
(equivalent)
volume of
The total volume of t&rant delivered for the titration of C, VF, is equal to V, + V2+ . . . + V,. V, is not included since it is delivered solely to form the base titrant concentration [T], which is restored after the termination of the titration sequence. The equivalence volume, V,, has to be calculated from VF, allowance being made for the fact that dilution due to the sample volume and the additions of titrant will require addition of further titrant to restore its concentration in the final solution to [Tl,. Hence at the end-point
EXPERIMENTAL Apparatus A microcomputer-controlled potentiometric analysis system consisting of the following commercially available units was used: (a) an ALTAIR 88OOBmicrocomputer equipped with a 32-kbyte RAM and the necessary peripherals (CRT, teletypewriter, cassette recorder), (b) an Orion 801 digital pH/mV-meter and (c) a Sargent-Welch S-11120-12 multispeed burette driven by a stepper motor. The timing-control circuit of the digital pH/mV-meter was modified (by substitution of components of smaller values for a resistor and a capacitor), to make the range of display time 0.09-0.6 set instead of the of original 0.65 sec. The internal square-wave clock of the burette, which controls the delivery rate, was disconnected and softwaregenerated pulses from the computer, through an optoisolator circuit, were directly fed to the driving circuit of the stepper motor, thus allowing direct control of the burette by the microcomputer; 4500 pulses (maximum frequency 120 Hz) corresponded to delivery of 1.000 ml of titrant. Homemade parallel interface circuits were used to transfer potentials to the microcomputer from the BCD output lines of the digital pH/mV-meter. An Orion 92-81 perchlorate ion-selective electrode in conjunction with a double-junction silver/silver chloride reference electrode was used as a periodate-concentration transducer. The characteristics of this electrode have been described elsewhere.* A double-walled SO-ml beaker kept at 33 + 0.1” was used
as titration cell.
[T] = (total equivalents of T added) - (equivalents of T consumed by C) B total volume in the titration cell = 0’0 + VF)[Tl,, - KITI, _~ __ . VW+
Combining
equations
(4)
v, + vs+ VF
J
(2) and (4) gives (5)
Software Control programs were written in BASIC except for the pqtential-sampling routines, which were written in the 8080 nucroprocessor machine language.
147
Titrations involving slow reactions
I
SUBROUTINES
PROGRAM
OVERALL
1
DELIVER v. (ml)
READ POTENTIAL E
YES.
I V=E-E. (mV)
WAIT 30 set
I
J=Jtl
+
CALCULATE RATE R
R, = R VF = 0 I-
*
END-POINT DETECTION/CALCULATE Vc
SELECT INCREMENT Lb15 15>hlO 10&U> 6 6 >U> 3 3 &lb 2 2 X70> 1 1 Zb 0
+ F=80 -t F=50 + F=30 + F=20 + F=lO + F=5 + F=2
YES
I
w b END
Fig. 2. Flow-charts of the overall control program and of two main subroutines “CALCULATE R” (*) and “DELIVER V” (**).
A flowchart of the overall control program is shown in Fig. 2. In the same figure the flow-charts of two main subroutines (CALCULATE RATE R and DELIVER V) are also shown. This program is suited for titrations monitored by potentiometric transducers. The operator tells the computer the sizes of VW, V, and V, and the initial amount V, ml of the titrant is delivered to form the “base” titrant (periodate) concentration. The “base” signal (potential E,) is read after the delay period of 30 set necessary to obtain a reasonably stable potential reading (within kO.1 mV). The sample is inserted manually and after a short mixing period (about 5 set) the initial reaction rate (R,) is measured by the CALCULATE RATE R subroutine and the incremental additions of titrant are started. The criterion for end-point detection is basically the kinetic condition R/R, < 0.001. Occasionally the quantization error of the potential measurement, combined with a low value of the initial rate (slowly reacting compounds--low sample concentrations) never allows this condition to be met. This case is dealt with (see Fig. 2) by examining the size of the last titrant increment and regarding the titration as completed if this is found to be zero. Once the end-point has been detected the corrected volume of titrant (V,) is calculated and printed.
RATE
In the CALCULATE RATE R subroutine at least 15 sequential potential readings are collected. This collection is terminated when 50 potential readings have been collected or the potential exceeds E, by 40 mV, whichever comes first. A least-squares fit is made and the reaction rate is calculated in terms of rate of increase of the perchlorate-electrode potential, in units of mV per display time period. The DELIVER V subroutine undertakes the delivery of titrant in increments of sizes which always depend on the difference, D, between the actual potential, E, and the “base” potential E,. The progressively diminishing size of each increment, as E approaches E,,, prevents an excessive overshoot of the “base” signal, E,. Reagents
Analytical-grade materials and demineralized distilled water were used throughout. Sodium metaperiodate, 0.1544. Sodium periodate (NaIO,, 32 g) is dissolved in water and the solution diluted to 1 litre. The solution is standardized iodometrically weekly. Acetate buffer, pH 4.5. The pH of O.lOM acetic acid is adjusted to 4.5 k 0.1 with 5M sodium hydroxide. Polyhydroxy-compounds. Stock solutions of ethylene glycol, propylene glycol, glycerol and mannitol are dissolved in
C. E. EFSTATHIOU and T. P. HADJIIOANNOU
148
Table 1. Effect of V, on the precision, accuracy and duration of the titration of 2.00 ml of 0.0779M propylene glycol with 0.1471M sodium metaperiodate VO,ml 0.100 0.200 0.500 1.000
Approximate duration, min
Error, %
13 7 5 3
+0.2 +0.1 +0.6 +3.2
1.061, k 0.007, 1.060, & 0.003, 1.065, + 0.002, 1.093, + 0.007,
*Mean and standard deviation of three replicates. water and standardized titrimetrically.’ Working solutions are made by appropriate dilution of the stock solutions. Procedure The display time of the digital pH/mV-meter is adjusted to about 0.2 sec. Acetate buffer (30.00 ml) is transferred into the titration cell and the control program is initiated. The following parameters are given to the microcomputer: VW= 30.00 ml, V, = 2.00 ml, V0= 0.200 ml. The sample solution (2.00 ml) is pipetted into the titration cell and the titration IS started and terminated automatically. The corrected titrant volume is computed and presented by the microcomputer. RESULTS AND DISCUSSION Selection of volume V, Volume V0 determines the base concentration level of periodate, [IO;ls. The duration of the titration will depend on [IO,], and will be short if [IO;]s is large enough. Unfortunately, in that case an overshoot of the base potential by even as little as 0.1-0.2 mV will cause a large positive analytical error. On the other hand, a smaller [IO;]s will increase the duration of the titration and cause unstable potential readings, which may, in turn, cause large analytical errors in either direction. In Table 1 the effect of the selected V. on the precision, accuracy and duration of the titration of 2.00 ml of 0.0779M propylene glycol with 0.147lM sodium periodate is shown.
Analytical results
Table 2 gives typical analytical results for the titration of the polyhydroxy-compounds, with kinetic end-point detection. The duration of these titrations was in the range 7-10 min for ethylene glycol, 5-9 min for propylene glycol, 5-7 min for glycerol and 7-9 min for mannitol. In most cases 5-10 titrant additions were needed. Conclusions
The kinetic detection of the end-point in titrations involving rather slow reactions yields sufficiently accurate results, at least comparable with those obtained by back-titration. The necessity for computer control is no longer a problem since the advent of microcomputers. A prerequisite for this type of titration is constant stoichiometry of the reaction. Titrations of tartrate and certain carbohydrates with periodate under the same conditions failed because of variable reaction stoichiometry. Such titrations took 3&40 min and gave non-reproducible results. Another prerequisite of utmost importance is the reliability of the transducer used to monitor the concentration of the titrant. The technique is in some degree complementary to the well-known technique of titration to a preset
Table 2. Results of the titration of polyhydroxy-compounds detection of the end-point Concentration,
mM
Taken
Found* f s
Ethylene glycol
25.0, 50.1, 100.2 150.4
25.1, * 49.9, * 100.0 * 150.3 f
0.1, 0.1, 0.2 0.4
Propylene glycol
25.9, 51.9, 103.9 155.8
26.1, k 51.9, f 103.6 f 155.5 *
0.0, 0.2” 0.10.2
Polyhydroxy-compound
Glycerol
28.0, 56.1, 84.2,
28.l,fO.l, 55.9, f 0.1, 84.1, f 0.3,
Mannitol
4.8, 9.6, 14.e
4.8, + 0.01 9.6, k 0.02 14.4,_~ + 0.02
*Mean and standard deviation of three replicates.
with kinetic
Error, “/, +0.2 -0.4 -0.2 +0.1 Av. 0.2, +0.6 -0.0 -0.3 -0.2 Av. 0.2, +0.3 -0.3 -0.1 Av. 0.2, +0.3
Av. 0.1,
149
Titrations involving slow reactions end-point, for which slow reactions are rather unsuitable because of premature (false) end-point detection. Some preset end-point titrators overcome this problem by means of a delay circuit which slows down the response time of the electronic switch controlling the burette.4 This delay is adjustable and some experimentation in needed to find the optimum setting. In essence this is a kinetic end-point detection. In the present work, such an adjustment is not necessary and the computer is left to decide when the preselected degree of reaction has been reached. The applicability of this technique to certain Karl Fischer titrations which proceed rather slowly and to
direct titrations of certain easily hydrolysed esters and alkyl halides with ethanolic potassium hydroxide solutions is under investigation.
REFERENCES
1. T. P. Hadjiioannou, Rev. Anal. Chem., 1976, 3, 82. 2. C. E. Efstathiou, T. P. Hadjiioannou and E. McNelis, Anal. Chem., 1917, 49, 410. 3. G. Dryhurst, Periodate Oxidation of Dial and Other Functional Groups, p. 121. Pergamon Press, Oxford, 1970. 4. G. Svehla. Automatic Potentiometric Titrations, p. 178. Pergamon Press, Oxford, 1978.
0039-9140/83/030145-05$03.00/O Copyright0 1983PergamonPressLtd
KINETIC DETECTION OF THE END-POINT IN TITRATIONS INVOLVING SLOW REACTIONS DIRECT TITRATION OF POLYHYDROXY-COMPOUNDS WITH PERIODATE C. E. EKSTATHIOU and T. P. HADJIIOANNOU Laboratory of Analytical Chemistry, University of Athens, 104 Solonos St., Athens 144, Greece (Received 3 April 1982. Revised 24 September 1982. Accepted 4 October 1982)
Summary-A new titration technique is described in which the end-point is determined by measuring the relative reaction rate of the titration reaction. This technique is adequate for rather slow reactions where conventional direct titrations are not applicable. The titrations are done automatically under microcomputer control. The efficiency of this technique is demonstrated with direct titrations of certain polyhydroxy-compounds with standard periodate solution. Ethylene glycol and propylene glycol (0.05-0.3 mmole), glycerol (0.06-0.17 mmole) and mannitol (0.01-0.03 mmole) were determined with average relative errors of 0.1-0.3x.
T in the titrated solution. The following sequence of events (illustrated in Fig. 1) will take place:
A common reaction titrators
requirement for direct titration is that the should be fast, especially when automatic are used. If the reaction is slow, it is custom-
ary to add excess of titrant, and back-titrate, but this increases the overall .analysis time and decreases the accuracy. The end-point is generally detected by observing a change in an indicator colour or some other parameter such as potential, polarization current or conductivity. During the last decade titration techniques have been developed in which kinetic parameters are used indirectly to locate end-points. In these techniques, known as catalytic titrations, the titrant and the titrand react stoichiometrically and the first excess of the former catalyses an indicating reaction that is used to locate the end-point of the titration.’ In this paper, we describe a titration technique which is applicable when rather slow but stoichiometric reactions take place between the titrant and titrand. This technique requires microcomputer control and the end-point is automatically detected when certain preselected requirements of a kinetic nature have been fulfilled. As a practical application of this technique, organic polyhydroxy-compounds have been directly titrated with a standard periodate solution. PRINCIPLEOF THE METHOD Compound C reacts rather slowly but stoichiometrically with compound T and is to be titrated with it according to the reaction C+ T2
products
(1)
Suppose there is a transducer, with sufficiently fast response, capable of monitoring the concentration of
(i) An initial volume, V.,,, of solvent or buffer is transferred into the titration cell. (ii) A predetermined volume, V,, of titrant of concentration ml0 is delivered automatically to form a “base” titrant, with concentration [Tl, which is given by
[Tie= vo[TM vw+ f’o)
(2)
(iii) The corresponding “base” signal of the transducer, S,, is measured and stored in the computer memory. (iv) A sample volume, V,, containing the whole amount of C to be titrated, is transferred into the reaction cell, to give an initial concentration [Cl,. (u) The computer continuously receives the signal S from the transducer, and from it calculates and stores in memory the initial reaction rate, R,, (= 1dS/dt I), in arbitrary units, which is proportional to -dm/dt:
-d[Tlldt = k[ClJ%
(3)
(vi) The burette is then actuated, and automatically delivers titrant until the signal from the transducer returns to the base signal S,. Reaction (1) takes place during this addition, consuming T until the delivery rate surpasses the consumption rate and the concentration of T returns to [T],. The total volume of titrant delivered, V,, is stored in memory. To avoid excessive overshoot of S,, the delivery rate becomes progressively smaller as the signal S approaches &. (vii) The computer, as in step (v), calculates the
new reaction rate, R,, which will be less than the 145
146
C. E. EFSTATH~OU and T. P. HADJIIOANNOU
r
I
measurements
I
of reaction rate -,
-v,+i
-addrtlons
of tltrant-
A
:
V
TIME
Fig. 1. Graphical presentation of the sequence of events taking place during a titration with kinetic detection of the end-point. The scale of the time axis is not uniform along its length. VO,V,, . . . V, denote the time periods during which the titrant is delivered. R,,, R,, . R,,denote the time periods during which measurements of reaction rate are taking place.
initial rate &, because part of the titrand
C has been
consumed. Steps (ui) and (vii) are repeated with addition of volumes V2, V,, . . . , V, and calculation of the corresponding reaction rates R,, R3, . . . , R.. The titration is terminated when the measured reaction rate R, after the n th delivery of titrant is equal to or less than a predetermined fraction of R,,, typically O.OOl&, which denotes that 99.9% of the titrand originally present has reacted, i.e., practically quantitative titration. Calculation titrant
of the corrected
(equivalent)
volume of
The total volume of t&rant delivered for the titration of C, VF, is equal to V, + V2+ . . . + V,. V, is not included since it is delivered solely to form the base titrant concentration [T], which is restored after the termination of the titration sequence. The equivalence volume, V,, has to be calculated from VF, allowance being made for the fact that dilution due to the sample volume and the additions of titrant will require addition of further titrant to restore its concentration in the final solution to [Tl,. Hence at the end-point
EXPERIMENTAL Apparatus A microcomputer-controlled potentiometric analysis system consisting of the following commercially available units was used: (a) an ALTAIR 88OOBmicrocomputer equipped with a 32-kbyte RAM and the necessary peripherals (CRT, teletypewriter, cassette recorder), (b) an Orion 801 digital pH/mV-meter and (c) a Sargent-Welch S-11120-12 multispeed burette driven by a stepper motor. The timing-control circuit of the digital pH/mV-meter was modified (by substitution of components of smaller values for a resistor and a capacitor), to make the range of display time 0.09-0.6 set instead of the of original 0.65 sec. The internal square-wave clock of the burette, which controls the delivery rate, was disconnected and softwaregenerated pulses from the computer, through an optoisolator circuit, were directly fed to the driving circuit of the stepper motor, thus allowing direct control of the burette by the microcomputer; 4500 pulses (maximum frequency 120 Hz) corresponded to delivery of 1.000 ml of titrant. Homemade parallel interface circuits were used to transfer potentials to the microcomputer from the BCD output lines of the digital pH/mV-meter. An Orion 92-81 perchlorate ion-selective electrode in conjunction with a double-junction silver/silver chloride reference electrode was used as a periodate-concentration transducer. The characteristics of this electrode have been described elsewhere.* A double-walled SO-ml beaker kept at 33 + 0.1” was used
as titration cell.
[T] = (total equivalents of T added) - (equivalents of T consumed by C) B total volume in the titration cell = 0’0 + VF)[Tl,, - KITI, _~ __ . VW+
Combining
equations
(4)
v, + vs+ VF
J
(2) and (4) gives (5)
Software Control programs were written in BASIC except for the pqtential-sampling routines, which were written in the 8080 nucroprocessor machine language.
147
Titrations involving slow reactions
I
SUBROUTINES
PROGRAM
OVERALL
1
DELIVER v. (ml)
READ POTENTIAL E
YES.
I V=E-E. (mV)
WAIT 30 set
I
J=Jtl
+
CALCULATE RATE R
R, = R VF = 0 I-
*
END-POINT DETECTION/CALCULATE Vc
SELECT INCREMENT Lb15 15>hlO 10&U> 6 6 >U> 3 3 &lb 2 2 X70> 1 1 Zb 0
+ F=80 -t F=50 + F=30 + F=20 + F=lO + F=5 + F=2
YES
I
w b END
Fig. 2. Flow-charts of the overall control program and of two main subroutines “CALCULATE R” (*) and “DELIVER V” (**).
A flowchart of the overall control program is shown in Fig. 2. In the same figure the flow-charts of two main subroutines (CALCULATE RATE R and DELIVER V) are also shown. This program is suited for titrations monitored by potentiometric transducers. The operator tells the computer the sizes of VW, V, and V, and the initial amount V, ml of the titrant is delivered to form the “base” titrant (periodate) concentration. The “base” signal (potential E,) is read after the delay period of 30 set necessary to obtain a reasonably stable potential reading (within kO.1 mV). The sample is inserted manually and after a short mixing period (about 5 set) the initial reaction rate (R,) is measured by the CALCULATE RATE R subroutine and the incremental additions of titrant are started. The criterion for end-point detection is basically the kinetic condition R/R, < 0.001. Occasionally the quantization error of the potential measurement, combined with a low value of the initial rate (slowly reacting compounds--low sample concentrations) never allows this condition to be met. This case is dealt with (see Fig. 2) by examining the size of the last titrant increment and regarding the titration as completed if this is found to be zero. Once the end-point has been detected the corrected volume of titrant (V,) is calculated and printed.
RATE
In the CALCULATE RATE R subroutine at least 15 sequential potential readings are collected. This collection is terminated when 50 potential readings have been collected or the potential exceeds E, by 40 mV, whichever comes first. A least-squares fit is made and the reaction rate is calculated in terms of rate of increase of the perchlorate-electrode potential, in units of mV per display time period. The DELIVER V subroutine undertakes the delivery of titrant in increments of sizes which always depend on the difference, D, between the actual potential, E, and the “base” potential E,. The progressively diminishing size of each increment, as E approaches E,,, prevents an excessive overshoot of the “base” signal, E,. Reagents
Analytical-grade materials and demineralized distilled water were used throughout. Sodium metaperiodate, 0.1544. Sodium periodate (NaIO,, 32 g) is dissolved in water and the solution diluted to 1 litre. The solution is standardized iodometrically weekly. Acetate buffer, pH 4.5. The pH of O.lOM acetic acid is adjusted to 4.5 k 0.1 with 5M sodium hydroxide. Polyhydroxy-compounds. Stock solutions of ethylene glycol, propylene glycol, glycerol and mannitol are dissolved in
C. E. EFSTATHIOU and T. P. HADJIIOANNOU
148
Table 1. Effect of V, on the precision, accuracy and duration of the titration of 2.00 ml of 0.0779M propylene glycol with 0.1471M sodium metaperiodate VO,ml 0.100 0.200 0.500 1.000
Approximate duration, min
Error, %
13 7 5 3
+0.2 +0.1 +0.6 +3.2
1.061, k 0.007, 1.060, & 0.003, 1.065, + 0.002, 1.093, + 0.007,
*Mean and standard deviation of three replicates. water and standardized titrimetrically.’ Working solutions are made by appropriate dilution of the stock solutions. Procedure The display time of the digital pH/mV-meter is adjusted to about 0.2 sec. Acetate buffer (30.00 ml) is transferred into the titration cell and the control program is initiated. The following parameters are given to the microcomputer: VW= 30.00 ml, V, = 2.00 ml, V0= 0.200 ml. The sample solution (2.00 ml) is pipetted into the titration cell and the titration IS started and terminated automatically. The corrected titrant volume is computed and presented by the microcomputer. RESULTS AND DISCUSSION Selection of volume V, Volume V0 determines the base concentration level of periodate, [IO;ls. The duration of the titration will depend on [IO,], and will be short if [IO;]s is large enough. Unfortunately, in that case an overshoot of the base potential by even as little as 0.1-0.2 mV will cause a large positive analytical error. On the other hand, a smaller [IO;]s will increase the duration of the titration and cause unstable potential readings, which may, in turn, cause large analytical errors in either direction. In Table 1 the effect of the selected V. on the precision, accuracy and duration of the titration of 2.00 ml of 0.0779M propylene glycol with 0.147lM sodium periodate is shown.
Analytical results
Table 2 gives typical analytical results for the titration of the polyhydroxy-compounds, with kinetic end-point detection. The duration of these titrations was in the range 7-10 min for ethylene glycol, 5-9 min for propylene glycol, 5-7 min for glycerol and 7-9 min for mannitol. In most cases 5-10 titrant additions were needed. Conclusions
The kinetic detection of the end-point in titrations involving rather slow reactions yields sufficiently accurate results, at least comparable with those obtained by back-titration. The necessity for computer control is no longer a problem since the advent of microcomputers. A prerequisite for this type of titration is constant stoichiometry of the reaction. Titrations of tartrate and certain carbohydrates with periodate under the same conditions failed because of variable reaction stoichiometry. Such titrations took 3&40 min and gave non-reproducible results. Another prerequisite of utmost importance is the reliability of the transducer used to monitor the concentration of the titrant. The technique is in some degree complementary to the well-known technique of titration to a preset
Table 2. Results of the titration of polyhydroxy-compounds detection of the end-point Concentration,
mM
Taken
Found* f s
Ethylene glycol
25.0, 50.1, 100.2 150.4
25.1, * 49.9, * 100.0 * 150.3 f
0.1, 0.1, 0.2 0.4
Propylene glycol
25.9, 51.9, 103.9 155.8
26.1, k 51.9, f 103.6 f 155.5 *
0.0, 0.2” 0.10.2
Polyhydroxy-compound
Glycerol
28.0, 56.1, 84.2,
28.l,fO.l, 55.9, f 0.1, 84.1, f 0.3,
Mannitol
4.8, 9.6, 14.e
4.8, + 0.01 9.6, k 0.02 14.4,_~ + 0.02
*Mean and standard deviation of three replicates.
with kinetic
Error, “/, +0.2 -0.4 -0.2 +0.1 Av. 0.2, +0.6 -0.0 -0.3 -0.2 Av. 0.2, +0.3 -0.3 -0.1 Av. 0.2, +0.3
Av. 0.1,
149
Titrations involving slow reactions end-point, for which slow reactions are rather unsuitable because of premature (false) end-point detection. Some preset end-point titrators overcome this problem by means of a delay circuit which slows down the response time of the electronic switch controlling the burette.4 This delay is adjustable and some experimentation in needed to find the optimum setting. In essence this is a kinetic end-point detection. In the present work, such an adjustment is not necessary and the computer is left to decide when the preselected degree of reaction has been reached. The applicability of this technique to certain Karl Fischer titrations which proceed rather slowly and to
direct titrations of certain easily hydrolysed esters and alkyl halides with ethanolic potassium hydroxide solutions is under investigation.
REFERENCES
1. T. P. Hadjiioannou, Rev. Anal. Chem., 1976, 3, 82. 2. C. E. Efstathiou, T. P. Hadjiioannou and E. McNelis, Anal. Chem., 1917, 49, 410. 3. G. Dryhurst, Periodate Oxidation of Dial and Other Functional Groups, p. 121. Pergamon Press, Oxford, 1970. 4. G. Svehla. Automatic Potentiometric Titrations, p. 178. Pergamon Press, Oxford, 1978.