880
Am/. Chem. 1992, 64, 886-891
(52) Means, 0. E.; F e e ~ yR. , E. chdmlccrl AWbkaths of Proteins: HOC dm-Day, Inc.: Sen Francisco, CA, 1971;Chapter 6. (53) Croft, L. R. Hendbwlr of Protek, Sequence Ane!~sb;John Wiley & Sons: Chlchester. U.K.. 1980. (54) Lambert. W. J i -ik&ton, D. L. AMI. m m . 1990, 62, 1585-1587. (55) QOSS, E. MeEMymd. 1967, 1 1 , 238-255. (56) oloss, E.; Witkop, J. J . Bld. Chem. 1962, 237, 1856.
(57) Simpson, R. J.; Nice, E. C. Biochem. Int. 1984, 8, 787-791. (58) Meloun, B.;Moravek, L.; Kostka, V. FEBS Lett. 1975, 58, 134-137. RECEmD
for review September 20,1991.Accepted January
9,1992.
Effect of Direct Control of Electroosmosis. on Peptide and Protein Separations in Capillary Electrophoresis Chin-Tiao Wu, Teresa Lopes, Bhisma Patel, and Cheng 5.Lee* Department of Chemical and Biochemical Engineering, University of Maryland Baltimore County Campus, Baltimore, Maryland 21228
The separations of peptide and proteln dxtures In capMary zone ehctrophorda (CZE) at various dutlon condltlow were rtudkd wlth the dlrect control of eMrooanorls. The f potential at the crqmoulcrrplrry hterface and the remHed eiecttoormosk In the presence of an dectrlc fkld were dC redly controlled by using an addttbnal ehctrk Ikld applied from outrld. of the capillary. The contmhd electrmotlc IkWcr(hcledd~mlgraknthtW8ndzom,nroMknol~
and proteln mlxtures. The changes In the magnltude and polarlty of the potential cawed the varlow degrees of peptide and protein aWnpikm onto the ccrplllary through the dectrootatk Interactlons. The wparatbn effkkncks of peptide and protdn mlxtures were enhanced d m to the reductknkrp@kJe(Md~0hla d ” a t d l s a p b y Wall. The direct manlpulatlonr of the separation etflokncy and resoluth of peptide and protdn mixtures In CZE were demonstrated by sbnply controlling the f potential and the doct r m o t l c flow wlth the appllcatkn of an external electric Ikld.
INTRODUCTION With the application of the current monitoring method’ and the UV marker method,2 we have recently proposed and demonstrated the direct control of electroosmosis in capillary zone electrophoresis (CZE) by using an additional electric field applied from outside of the capillary.” The potential gradient between the external and internal electric potentials is perpendicular a c r m the capillary wall and controls the polarity and magnitude of the 5 potential on the interior surface of the capillary wall. Because the direction and flow rate of e1ect”ceis are dependent upon the polarity and magnitude of the ( potential,Bthe electroosmotic flow can therefore be directly manipulated by varying the external electric field. In CZE, the electroosmotic flow affectsthe amount of time a solute resides in the capillary, and in this sense both the separation efficiency and resolution are related to the W o n and flow rate of electmoeis.7 In thisstudy, the separations of peptide and protein mixtures in CZE with the direct control of electroosmosis are described. The effect of such control on the separation efficiency and resolution of peptide and protein mixtures is discueeed. The direct manipulation of the separation efficiency and resolution in CZE can be easily obtained by simply changing the external electric field. *Towhom all correspondenceshould be addressed. 0003-2700/92/0364-0886$03.00/0
EXPERIMENTAL SECTION The experimental setup for directly controlling the electroosmotic flow in capillary electrophoresis by using an additional electric field applied from outside the capillary has been described in detail in the previous study? Peptide and protein samples were introduced into the inner capillary by using the electromigration injection method? No external electric field was used during the injection period. For peptide and protein separations at various applied potential gradients, the cathode end of the inner electric field was always in reservoir 4,the UV detector end. The electrophoretic migration of peptide and protein mixtures at various solution conditions examined in this study was always toward reservoir 4,the cathode end of the inner electric field. A nonionized molecule, dimethyl sulfoxide, in the solution mixture was used as the e l e c t m o t i c flow marker. The change14 in the diredion and flow rate of electrooemosis with the application of an external electric field were measured. The flow rate of electroosmosiswas assigned as positive when the direction of flow was toward the cathode end of the inner electric field. The direction of electroosmosis was toward the anode end when the { potential at the inner capillary/inner solution interface was changed from negative to positive6with the application of strong positive potential gradients across the inner capillary wall. In order to measure the negative value of electroosmotic mobility by using dimethyl sulfoxide as the flow rate marker, the cathode end of the inner electric field was temporarily changed from reservoir 4 to reservoir 1. The sodium phosphate buffer, dimethyl sulfoxide, hydrochloric acid, and all peptides and proteins were purchased from Sigma (St. Louis, MO). The pH of the buffer solution was adjusted with the application of 0.1 N hydrochloric acid. The separations of peptide mixtures in a 23-cm-long inner capillary with a 50-pm i.d. and 150-rm 0.d. were studied. All peptides except the adrenocorticotropichormone fragment 4-10were dissolved in 10 mM phosphate buffer of pH 2.7 at concentrationsof approximately 200 pg/mL. The adrenocorticotropic hormone fragment 4-10was dissolved in the same buffer at a concentration of apPeptides were injected by using 1-kV inner proximately 40 electric potential for 10 s without the application of an external electric field. A constant inner electric field equal to 239 V/cm (5.5 kV over 23-cm-long inner capillary) was then applied for electrophoresis. The separation distance between reservoir 1 and the UV detector was 14.5 cm. RESULTS AND DISCUSSION Peptide Separations. The experimental result for peptide separation in the absence of an external electric field was shown in Figure 1. The elution order was Lys-Trp-Lys, thymopoietin I1 fragment, adrenocorticotropic hormone fragment 4-10,bradykinin, and human angiotensin 11. The theoretical charges on the peptides were calculated by using Skoog and Wichman’s model? The parameter, charge/(mo0 1992 American Chemical Society
ANALYTICAL CHEMISTRY, VOL. 64, NO. 8, APRIL 15, 1992 2
1 2
,
0
2
887
4
I
I
,
0
2
4
6
Time (min)
Time (min)
Flgure 1. Zone electrophoretic separation of peptides in the absence of an extemai electric field. Elution order: 1 = Lys-Trp-Lys, 2 =
thymopoietin I1 fragment, 3 = adrenocciticotropic hormone fragment 4-10, 4 = bradykinin, 5 = human angiotensin 11; buffer, phosphate 10 mM at pH 2.7; capillary, 50-l,tm i.d. and 150-bm o.d., length to
Figure 2. Zone electrophoretic separation of peptides in the presence of a -5-kV potential gradient across the inner capillary. Other condklons are the sameas in Figure I.
detector 14.5 cm; voltages, 1 kV and 10 s for injection, 5.5 kV for electrophoresis.
Table I. Elution Order, Peptide Sequence, and Charge/(Molecular Weight)s/2of Peptides Used in This Study
name
sequence
Lys-TrpLys thymopoietin I1 fragment adrenocorticotropic hormone fragment
Lye-Trp-Lye Arg-Lys-AspVal-Ty Met-Glu-HisPhe-Arg-TrpGly Arg-Pro-ProGly-Phe-SerPro-Phe-Arg Asp-Arg-ValTyr-Ile-HisPro-Phe
bradykinin angiotensin I1
elution order
eight)^/^
2
0.0376 0.0282
3
0.0232
4
0.0203
5
0.0199
1
1 -
charge/ (molecular
lecular eight)^/^ used by Richard et al.l0 for the prediction of electrophoretic migration of peptides was calculated for our model peptides. The elution order, peptide sequence, and charge/(molecular eight)^/^ of our model peptides at pH 2.7 were summarized in Table I. The separation result in the absence of an external electric field was then used to compare with peptide separations in the presence of various potential gradients across the inner capillary at the same solution conditions. With the application of a -5-kV potential gradient, the enhanced electroosmotic flow was in the same direction as electrophoretic migration of peptides from reservoir 1to reservoir 4. The cathode end of the inner electric field was always in reservoir 4 for the peptide separations. As shown in Figure 2, the enhanced electroosmotic flow in the presence of a -5-kV potential gradient resulted in the decrease of analysis time and zone resolution between the adjacent peptide peaks. The flow rate of electroosmosis was further increased in the presence of a -8-kV potential gradient across the inner capillary. As shown in Figure 3, this further enhancement in the electroosmotic flow separated the peptide mixture in less than 3 min, however, at a large expense in separation resolution. Bradykinin and human angiotensin I1 as peak 4 and peak 5 eluted together with the application of a -8-kV potential gradient. In the absence of an external electric field, the outer electric circuit was still connected between the outer capillary and the outer high-voltage power supply. For the application of a 0 potential gradient, the polarity and magnitude of the outer electric field were the same as the inner electric field in the inner capillary between reservoir 2 and reservoir 3.5 The electroosmotic flow was reversed in the presence of 0- and
0
2
4
Time (min)
Figure 3. Zone electrophoretic separation of peptides in the presence of a -8-kV potential gradient across the inner capillary. Other conditions are the same as in Figure 1. L
1
1
u 4
4 C
T
e si
3.002 AU
It
0
i
, I
i! 0
2
4
I 6
Time (min)
m e 4. Zone eleci~ophoresisseparation of peptides In the presence of a OkV potential gadlent across the inner capHlary. Other conditions are the same as in Figure 1.
+3-kV potential gradients across the capillary. The polarity of the t potential at the inner capillarylinner solution interface was changed from negative in the absence of an external electric field to positive in the presence of 0- and +3-kV potential gradients. The direction of electroosmotic flow was therefore toward reaemoir 1, the anode end of the inner electric field. The electroosmotic flow was in a direction against the electrophoretic migration of peptides. As shown in Figures 4 and 5, the analysis time and separation resolution of peptide mixture were all increased significantly with the application of 0- and +3-kV potential gradients. The values of total spatial variances for peptides were computed from peak profiles by using the formula u2 = [ l / ( 8 In 2 ) ] [ ( L w ) / t I 2 (1) where w is the full peak width at the half-maximum points and L is the separation distance between reservoir 1and the UV detector." If molecular diffusion alone was responsible for zone broadening, the values of total spatial variances would
888
ANALYTICAL CHEMISTRY, VOL. 64, NO. 8, APRIL 15, 1992
T 0.W2AU L
a P
c
0
2
h 4
6
Time (min)
~~
I
0
Figure 6. Zone electrophoretic separation of proteins In the absence of an external electric fleM: buffer, phosphate 10 mM at pH 2.5; capillary, 50-pm 1.d. and 150-l.tm o.d., length to detector 16 cm; voltages, 0.5 kV and 5 s for Injection, 4 kV for electrophoresis.
I2
8
4
16
Time (min)
Figun 5. Zone electrophoretic separation of peptides in the presence of a +3-kV potentlal gradient across the inner capillary. Other conditions are the same as In Figure 1.
Table 11. Values of Total Spatial Variances for Separation Peaks of Lye-Trp-Lys and Thymopoietin at Various Applied Potential Gradients amlied - _ Dotentid
gradient, kV
-a -5
no external electric field 0 3
total variances, cmza migration time, minb Lvs-TrDthvmo- Lvs-TrDthvmo- LYS pdietin LYS pdietin 0.0091 0.0074 0.0071
0.0073 0.0064 0.0062
1.80 2.08 2.47
2.12 2.45 3.00
0.0068 0.0053
0.0057 0.0051
2.85 4.74
3.85 7.04
The experimental error in measuring the value of total spatial variances was about 5-10% for various potential gradients for over five runs. bThe experimental error in measuring the migration time of peptide was about 0-3% for various potential gradients for over five runs. then be proportional to the analysis time at a constant inner electric field." However, the values of total spatial variances for peaks 1-2 as summarized in Table I1 and for peaks 3-5 (not shown) were decreased with the increase in migration time. One possible explanation for such an increase in the separation efficiency was the reduction in peptide adsorption onto the capillary wall. With the application of positive potential gradients, the t potential became positive and repelled the positive charged peptides away from the capillary wall. The effect of the reduction in peptide adsorption was more important than the increase in analysis time on zone broadening. The separation efficiency was therefore increased with the application of positive potential gradients. Conversely, the { potential became more negative in the presence of negative potential gradients and attracted peptides more closely to the capillary surface. The separation efficiency was therefore decreased with the increase in peptide adsorption even though the separation of peptides was fast with the application of negative potential gradients. Thus, the zone resolution of peptides with the application of positive potential gradients was enhanced by the increase in both separation efficiency and migration time. The electrophoretic mobilities of peptides, pep,were calculated by using the formula (2) Pep = P - Cleo where p is the net migration mobility of peptide molecule measured from the experiment. The calculated results for the electrophoretic mobilities of peptides were constant for each peptide at various potential gradients across the inner capillary in this study. The electrophoretic velocities of
peptides were also constant for each peptide with a constant inner electric potential of 5.5 kV. Protein Separations. The difficulty in applying CZE to protein separations arises from silanol groups on the surface of fused silica capillaries. Silanols ionize above pH 3 and greatly increase band broadening and peak tailing through adsorption of many proteins onto capillary wall. One strategy to reduce protein adsorption has been to operate CZE with acidic pH buffers.12 Low pH phosphate buffers have been used by McCormick12for the purpose of reducing the negative charge of fused silica surface as well as inducing some protective screening of the silica surface by phosphate groups. Operation at low pH values (pH less than the isoelectric point (PI)of all proteins) also ensures that the proteins in a sample will have a net positive charge and thus will migrate in the same direction (toward the cathode). The major limitation of performing CZE separations at such a low pH is that the peak capacity of the separations is low because the acidic buffer fully protonates the proteins, thereby diminishing charge differences between the species.12 To investigate this limitation, the separations of protein mixtures in a 25-cm-long inner capillary with a 50-pm i.d. and 150-pm 0.d. were studied at 10 mM phosphate buffer and pH 2.5. All proteins were dissolved in phosphate buffer at concentrations of approximately 0.5 mg/mL. Proteins were injected by using 0.5-kV inner electric potential for 5 s without the application of an external electric field. A constant inner electric field equal to 160 V/cm (4 kV over 25-cm-long inner capillary) was then applied for electrophoresis. The cathode end of the inner electric field was always in reservoir 4, the UV detector end. The separation distance between reservoir 1and the UV detector was 16 cm. As shown in Figure 6, the protein mixture was not completely separated and resolved in the absence of an external electric field. The acidic buffer fully protonated the proteins and resulted in the poor separation resolution. The electroosmotic mobility was (0.65 f 0.02) X lo4 cm2/V.s, in the same direction as electrophoretic mobilities of proteins. With the application of 0- and +3-kV potential gradienb across the inner capillary wall, significant improvements on the protein separation were clearly observed in Figures 7 and 8. The electroosmotic mobilities in the presence of 0- and +3-kV potential gradients were (-0.08 f 0.01) X lo-' and (-1.94 f 0.03) X lo4 cm2/V.s, respectively. The direction of electroosmosiswas against the electrophoretic mobilities of proteins. The elution order for the protein mixture was lysozyme, bovine serum albumin, ribonuclease A, a - c h p o trypsin, or-chymotrypsinogen A, and hemoglobin A. The isoelectric point (PI)and molecular weight of proteins used in this study were given elsewhere.13J4Bovine serum albumin was eluted earlier from the capillary tubing even with its small PI and large molecular weight. Denaturation of bovine serum albumin at extreme pH such as pH 2.5 could be one possible
ANALYTICAL CHEMISTRY, VOL. 84, NO. 8, APRIL 15, 1992
I 1
T 0032AU -t
0
2
'
1
I, 4
3
L
4
6
i
5
0.032 AU
8
Time (min)
Flguro 7. Zone electrophoretic separatlon of proteins In the presence of a OkV potential gradlent across the Inner caplllary. Elution order: 1 = lysozyme, 2 = bovlne serum albumin, 3 = ribonuclease A, 4 = ,a5 = achymotrypshogen A, 8 = hmogkMn A. other condltlons are the same as In Figure 8.
I
I
I
II
I1
.
I 2
I 4
I
I
6
8
2
4
6
E
10
h I 10
1 12
Tme (min)
Flgwe 0 . Zone electrophoretic separation of protelns In the absence of an external electrlc field. Elution order: 1 = lysozyme, 2 = cytochrome C, 3 = ribonuclease A, 4 = a-chymott-ypslnogen A, 5 = myogbbln; buffer, phosphete 10 mM at pH 5; capillary, 50-pm 1.d. and 150-pm o.d., length to detector 18 cm; voltages, 0.5 kV and 5 s for Injection, 4 kV for electrophoresls. 1
I
0
889
12
14
4
16
T i e bin)
Flgvr 8. Zone ektrophoretlc separation of proteins in the presence of a +3-kV potential gradient across the Inner capillary. Other condltlons are the same as in Figure 8.
explanation for its earlier elution and double peaks as shown in Figures 7 and 8. To inveatigate the effect of external electric field on protein adsorption at the capillary wall, the separations of the model proteins at 10 mM phosphate buffer and pH 5 were investigated. Experimental results of protein separations in the absence of an external electric field and in the presence of a -3.7-kV potential gradient were shown in Figures 9 and 10, respectively. The elution order for the protein mixture was lysozyme, cytochrome C, ribonuclease A, a-chymotrypsinogen A, and myoglobin. a-Chymotrypsin, bovine serum albumin, and hemoglobin A examined at pH 2.5 were not used as the model proteins in the separations at pH 5. Due to significant adsorption of proteins onto the capillary wall, poor reproducibility of migration time of proteins was observed. Runto-runreproducibility of migration time was improved to 5% relative standard deviation for over five continuous runs by washing the capillary between the analyses to remove the adsorbed proteins. A 20-min wash with 0.1 N NaOH was followed by 30-min flushes with the separation buffer. The f potential at the capillary wall became more negative with the application of a -3.7-kV potential gradient than in the absence of an external electric field. This more negative f potential, however, induced the stronger protein adsorption onto the capillary wall. Due to the stronger adsorption of positively charged proteins at the capillary wall, the apparent 5 potential and the resulted electrooemotic flow were therefore smaller in the presence of a -3.7-kV potential gradient than in the absence of an external electric field. As shown in Figures 9 and 10, the migration time of protein molecule in
t 0
I 2
I 4
I
I
6
8
I 10
I
I
12
14
Time (min)
Flgm 10. Zone electrophoretic separatkm of protelns in the presence of a -3.7-kV potential gradlent across the Inner capillary. Other condltlons are the same as In Figure 9.
the inner capillary was shorter in the absence of an external electric field than in the presence of a -3.7-kV potential gradient. It was well-known that the adsorbed proteins modified the capillary surface, usually decreasing the electroosmotic flow significantly.' With the application of both 0- and +4-kV potential gradients across the inner capillary wall, significant improvements on the separation reaolution of proteins were clearly observed in Figures 11 and 12. The electroosmotic mobility was reduced from (1.03 f 0.10) X lo4 cm2f V-s in the absence of an external electric field to (0.44 f 0.01) X l0-L cm2f V.s and (0.30 f 0.01) X loa cm2f V.s in the presence of 0- and +4-kV potential gradients, respectively. The migration reproducibility was studied without the washing between the analyses. The run-+run reproducibility of migration time was 2% relative standard deviation for over five runs with the application of 0- and +4-kV potential gradients across the capillary. The influence of external electric field on protein adsorption at the capillary wall was evaluated in terms of the separation efficiency. The values of total spatial variances for various protein molecules on 16-cm capillary (from reservoir 1to the UV detector) were calculated by using sq 1 and summarized in Table 111. For lysozyme and cytochrome C, the two pro-
890
ANALYTICAL CHEMISTRY, VOL. 64, NO. 8, APFiIL 15, 1992 2
Table 111. Values of Total Spatial Variances for Separation Peaks at Various Applied Potential Gradients total variances. cmZa
no external proteins
-3.7 kV electric field
lysozyme cytochrome C ribonuclease A a-chymotrypsinogen A
8
0.256 0.051 0.0233 0.0233
0.085 0.037 0.0183 0.0171
0 kV
+4 kV
0.0256 0.0160 0.0128 0.0098
0.0142 0.0102 0.0085 0.0051
The experimental error in measuring the value of total spatial variances was about 5-10% for various potential gradienta for over five runs.
0.W2A U
f
Table IV. Total Spatial Variances for Separation Peaks at Various Inner Electric Field Strengths
t
I 2
0
I
I
4
6
I 10
I
8
I 12
I
I
14
16
Time (min)
FlgtD'O 11. Zone ebctrophoretk separation of proteins in the presence of a OkV potential gradlent across the inner capillary. Other condltkns are the same as In Figure 9. 2 1 1
II
I/ I 0
I
I
I
I
I
4
0
m
10
I 12
1
I
14
I6
Time (min)
FlgtD'O 12. zone ~ o p h o r e t i c separation of proteins In the presence of a + 4 k V potential gradient across the Inner capillary. Other condltlons are the same as in Flgure 9.
teins were not completely resolved in the presence of a -3.7-kV potential gradient and in the absence of an external electric field. The values of total spatial variances for these two proteins were therefore estimated from single component analysis in the presence of a -3.7-kV potential gradient and in the absence of an external electric field. The experimental results shown in Table I11 clearly indicated the strong effect of external electric field on protein adsorption at the capillary wall. The lowest separation efficiency and serious band tailing of proteins appeared in the presence of a -3.7-kV potential gradient. In contrast, the adsorption of proteins onto the capillary wall was significantly reduced with the application of 0- and +4kV potential gradients for obtaining the highest separation efficiency. The reduction of protein adsorption in the presence of 0- and +4kV potentid gradients also contributed to the great improvements in the separation resolution and migration reproducibility of proteins. However, the values of total spatial variancea observed in this study were still considerably higher than those predicted by theory.11 This theory assumed that molecular diffusion was the only cause of peak broadening. The values of total spatial variances obtained in this study indicated that the electrostatic interaction of proteins with the capillary wall has been reduced
total variances. cmZa 220 V/cm 260 V/cm
proteins
160 V/cm
ribonuclease A a-chymotrypsinogen A
0.0085 0.0051
0.0128 0.0061
0.0183 0.0091
"A constant +4-kV potential gradient across the inner capillary was applied at various inner electric field strengths. The experimental error in measuring the value of total spatial variances plates was about 5-10% for various inner electric field strengths for over five runs.
with the application of positive potential gradients, but not completely eliminated. In addition, the hydrophobic-like interaction of proteins with the capillary wall might account for the lower values of separation efficiency. The inner electric field strength was varied from 160 to 220 V/cm and then to 260 V/cm in the presence of a constant +4-kV potential gradient across the inner capillary wall. The corresponding current in the inner electric field was increased from 11PA at 160 V/cm to 22 PA a t 260 V/cm. The values of total spatial variances for ribonuclease A and a-chymotrypsinogen A at three different inner electric field strengths were sulIllIlflllzed * in Table IV. The fact that the value of total spatial variances increased as the inner voltage was increased was consistent with the results observed by Cobb et al.15 Thermal contribution to band broadening at higher inner electric field strength was offered as the possible explanation in Cobb's work.15 In fact, the effect of temperature on separations has been discussed by Nelson et aLI6 McManigill and Swedberg" have discussed the factors affecting the separation efficiency in CZE in the presence of protein adsorption at the capillary wall. The resistance to mass transfer and adsorption-desorption kinetic terms discussed in their work were two important contributing factors to band broadening. The lower inner electric field strength and the resulted longer migration time for proteins in the inner capillary reduced the contribution of these two kinetic terms to band broadening. The relaxation of these two kinetic terms provided another possible explanation for the effect of inner voltage on the separation efficiency. ACKNOWLEDGMENT
Support for this work by the donors of the Petroleum Research Fund, administrated by the American Chemical Society, and the Instrumentation and Instrument Development Program of the National Science Foundation is gratefully acknowledged. We acknowledge Drs. McManigill, Lux, and Swedberg of Hewlett-Packard Laboratories for their stimulating discussions.
Note Added in Proof. Subsequent to our paper: Ghowsi et al.I8 proposed a similar approach for controlling electroosmosis.
Anal. Chem. 1992, 64, 891-895
Regietry No. Lys-Trp-Lys, 38579-27-0;thymopoietin I1 fragment, 69558-55-0;adrenocorticotropic hormone fragment, 4037-01-8; bradykinin, 5882-2;angiotensin II,4474-91-3; lysozyme, 9001-63-2; cytochrome C, 9007-43-6; ribonuclease A,9001-99-4; a-chymotrypsinogen A, 9035-75-0;a-chymotrypsin, 9004-07-3; hemoglobin A, 9034-51-9.
REFERENCES (1) Huang, X.; (krdon, M. J.; a r e , R. N. Anel. Chem. 1988. 60,
1837-1838. (2) Tsuda, T.; Nomura, K.; Nakagawa, G. J . Chromtogr. 1982, 248, 241-247. (3) . . Lee, C. S.;Blanchard. W. C.; Wu. C. T. AMI. Chem. 1990. 62, 1550-1552. (4) Lee, C. S.;McManlglll, D.; Wu, C. T.; Patel, 8. Anal. Chem. 1991, 63, 1519-1523. (5) Lee, C. S.; Wu, C. T.; Lopes, T.; Patel. B. J . Chrometogr. 1991, 559, 133-140. Science: principles and Appll(8) Hunter, R. J. Zeta Potentlel In cW~&i caf&ms; Academic Press: New York. 1981. (7) Jorgenson. J. W.; Lukacs, K. D. Science 1983, 222, 286-272.
89 1
(8) Welllngford. R. A.; Ewing, A. 0. A&. chrome^. 1989, 29, 1-78. Chem. 1966, 5. 82-90. (9) Skoog. 6.; Wlchman, A. T&Anal. (10) Richard, E.; Strohl. M.; Nblwn. R.; Farb, P. Thlrd International Sympossium on Hi@ Performance CaplHary Electrophoresis 1991,Poster Paper PT-19. (11) Jorgenson, J. W.; Lukacs, K. D. Anal. Chem. 1981, 53,1298-1302. McCormlck, R. M. Anal. Chem. 3968, 60, 2322-2328. (12) (13) Lauer. H. H.; MchrbnigUI, D. Anel. (2”.1980, 58. 186-170. (14) Worthlngton, C. C. Worfhlngton Enzymo Manual; Worthlngton Bbchemlcal Corporatkn: Freehold, 1988. (15) Cobb, K. A.; Dolnik, V.; Novotny, M. Anal. Chem. 1990. 62. 2478-2483, (16) Nelson, R. J.; Paulus, A.; Cohen, A. S.; Outtmen, A.; Karger, B. L. J . C h f O I M w . 1989, 480, 111-128. (17) McManiglll, D.; Swedberg, S. A. In Techniques In Protein ChCHnlpby: Hiall, T. E., Ed.; Academic Press: Sen Dlego, 1989 pp 488-478. (18) ahowski. K.; ale,R. In 8kmensw Technokgy; Buck, R. P., Hatfbkl, W. E., Umana, M., Bowden, E. F., Eds.; Marcell Dekkar: New York, 1990 pp 55-82.
R~~~~~ for review October 1,1991. Accepted J~~~ 31, 1992.
Comparison of Methods To Assess Surface Acidic Groups on Activated Carbons Teresa J. Bandosz, Jacek Jagiello,’ and Jamee A. Schwarz* Department of Chemical Engineering and Materials Science, Syracuse University, Syracuse, New York 13244-1190
The effect of oxidation wlth nItrlc acld on activated carbons from dH.rent ortglns has been studled by Inverse gas chromatography at Inflnlte dlutlon, h h m tItratlon, and mass 11tratbn to determlne the polnt of zero charge (PZC) of carbons. l h e ~ a f o x l d a t b Is n to generate acldlcgrouw and thk becomes more pronounced as tho temperature of oxldatbn k Increased. We Ikrd that the Boehm tltratkn results and th.c a m ’ PZCs correlate wtth the severtty of the oxldatkn treatment. These methods “study” the surface of carbons by Its reponu to an aqueous environment. Inverse gas chromatography results show that Increasing acldlty Is reflected by an Increase In the speclflc lnteractlon of r electrons In n-akmm Wlth addlc centers on th8 surface of carbone. This method probes the donor/acceptor nature of the carbon’s surface and provldes lnfonnatlon compltmentary to that wpplied by the other methods.
INTRODUCTION During the past 20 years great demands have been placed on the development of carbon materials for diverse applications of practical importance. In particular, activated carbons are widely used as adsorbenta in gaseous, aqueous, and nonaqueous streams, as electrode materials in fuel cells, as catalyat supporta, and as fibers for structural reinforcement or fiiters. The demands placed on activated carbon based technologies have outpaced fundamental studies of the relationship between the properties of carbons and their performance. If these materials are to find continued successful use, then the properties of existing materials will require better under-
* Author to whom correspondence should be addressed. ‘Permanent address: Institute of Ener ochemistry of Coal and Physicochemistry of Sorbents,University of Mining and Metallurgy, 30-059 Krakcw, Poland. 0003-2700/92/0364-0891$03.00/0
standing so that new generations of these materials can be developed. A particularly desirable property of activated carbons in their use as an adsorbent is their high surface area, which is the result of their microporosity. The adsorption characteristics of activated carbon are affected by the type of carbon, and that is governed by the source of raw material and the preparation procedures used during carbonization and activation.14 Some adsorption properties can be explained by differences in the microporosities of carbons. However, another important consideration is the surface chemistry of the carbons. The carbon matrix is decorated with heteroatoms; the main heteroatom is oxygen. These can be connected with peripheral atoms of carbon (edges and corners of crystallites) as well as find location in intercrystalline spaces or in defected areas between the planes that create the crystallites. Of particular importance are the heteroatoms located on the carbon surface. Different functional groups can be derived from these chemical centers, and it is found that these groups are analogous to typical organic compounds. The most common are carboxyl, lactonic, carbonyl, and phenolic. The presence of these groups engenders an amphoteric property to the carbon when placed in an aqueous environment. The acidlbase properties can be determined using titration techniques employing specific titers. The most common is that described by Boehme5 However, acidlbase potentiometric titration can be also used to yield an index of the carbon acidity denoted its point of zero charge (PZC). Another method which provides an estimate of a carbon’s PZC has been designated mass titration? We find that there are certain advantages to this method. In addition to the results obtained from optical spectroscopies? a very promising technique that can be used to study carbon functional groups without a surrounding aqueous environment is X-ray photoelectron spectroscopy (XPS). This method is very sensitive to the chemical speciation on the 0 1992 American Chemical Society
