the effect of temperature on the cation-exchange...

Chem. Anal. (Warsaw), 49, 665 (2004)

* Corresponding author. E-mail: [email protected]; Fax: (0-22) 811 15 32

Key words: ion chromatography, effect of temperature, thermodynamics of ionexchange, inorganic cations, mechanism of zone spreading

The Effect of Temperature on the Cation-ExchangeSeparations in Ion Chromatography and the Mechanism

of Zone Spreading

by Krzysztof Kulisa

Department of Analytical Chemistry, Institute of Nuclear Chemistry and Technology,ul. Dorodna 16, 03-195 Warszawa, POLAND

The effect of column temperature on retention of alkali metal and alkaline earth metalcations, as well as some amines, has been studied using three commercially availablecation-exchange columns of Dionex Ion Pac type (CS10, CS12 and CS12A). Each columnwas packed with the cation-exchange resin containing different functional groups: stronglyacidic sulphonic acid groups in Ion Pac CS10, weakly acidic carboxylic acid groups in IonPac CS12, and a mixture of weakly acidic carboxylic acid and phosphonic acid groups inIon Pac CS12A column. The temperature was being changed from 10°C up to 60°C, whichcaused selectivity changes of the cation exchangers towards the analyzed cations, as well asthe changes of column efficiencies. The changes of thermodynamic functions (free energy,enthalpy and entropy) for the studied cation-exchange processes were calculated. The compa-rison of the temperature effect on the ion-exchange behaviour of several ions on threecation-exchange columns with pellicular-type resin beds has been made. It has been noti-ced, that in some cases the lowering of the column temperature down to 10°C was moreadvantageous to achieve complete and rapid separation of some cations than the increase ofthe column temperature. It has been found that in the case of ions of high retention factorsthe longitudinal diffusion within the resin phase may significantly contribute to the totalplate height.

666 K. Kulisa

W niniejszej pracy zbadano wp³yw temperatury na rozdzielanie kationów metali alkalicznychi ziem alkalicznych oraz niektórych amin na trzech wysokosprawnych kolumnach kationo-wymiennych typu Dionex Ion Pac (CS10, CS12 i CS12A). Ka¿da z kolumn by³a wype³nionakationitem zawieraj¹cym inny rodzaj grup funkcyjnych: silnie kwasowe grupy sulfonowe(kolumna CS10), s³abo kwasowe grupy karboksylowe (kolumna CS12) oraz mieszaninês³abo kwasowych grup karboksylowych i fosfonowych (kolumna CS12A). Zmiany tempe-ratury pracy kolumn w zakresie od 10°C do 60°C wp³ywa³y na selektywnoæ badanychkationitów w stosunku do rozdzielanych kationów, jak równie¿ na sprawnoæ kolumn. Dlaka¿dej z badanych kolumn i ka¿dego z analizowanych kationów obliczono zmiany entalpii,entropii i energii swobodnej reakcji kationowymiennych. Porównano efekty temperaturowedla trzech badanych kolumn. W przypadkach niektórych par kationów obni¿enie temperaturydo 10°C sprzyja³o ich lepszemu rozdzieleniu. Wykazano równie¿, ¿e w przypadku kationówo wysokich wspó³czynnikach retencji dyfuzja pod³u¿na w fazie jonitu mo¿e wp³ywaæw znacz¹cy sposób na wysokoæ pó³ki teoretycznej kolumny.

Column temperature is a factor, which determines the selectivity of ion exchangereactions and the efficiency of the column, and thus influences the quality of the separa-tion of ions. The application of elevated temperatures in HPLC techniques, e.g. rever-sed-phase liquid chromatography (RPLC) [15] and ion-exclusion chromatography[6] has become popular recently. Also in classical ion-exchange chromatography[721] temperature of the column has been evidenced as a factor influencing theselectivity towards inorganic ions. In ion chromatography (IC), however, the role ofthe temperature as the selectivity- and efficiency-controlling factor has been rarelyreported [2226]. Smith et al. studied anion-exchange behaviour of Dionex Ion PacAS4A column at elevated temperatures (up to 80°C) [27]. Hatsis and Lucy reportedthe temperature effect (27°C60°C) on the selectivity of Dionex Ion Pac AS11 andAS14 columns towards anions [28]. Dybczyñski and Kulisa investigated the influ-ence of temperature on thermal stability and performance of anion-exchange columnsDionex Ion Pac AS5 and AS9SC [29]. Rey and Pohl [30] as well as Hatsis and Lucy[31] investigated the influence of elevated temperature (27°C60°C) on the separa-tion of alkali, alkaline earth metals and amines on Dionex Ion Pac CS12A cation-exchange column. Paull and Bashir studied temperature effects on the retention ofalkali, alkaline earth and transition metals on four different cation-exchange columns(sulfonated Dionex Ion Pac CS10, carboxylated Dionex Ion Pac CS14, dicarboxylatedHamilton PRPX800 and silica based iminodiacetic acid functionalized BioChemMack columns) [32]. The authors concluded that the increase of the temperature decrea-sed the retention of alkali metal ions on all types of the studied cation-exchangers. Foralkaline earth metal cations more complex temperature behaviour has been observed.Depending on the acidity of the cation-exchange functional groups and eluents, reten-tion either increased or decreased with the temperature. Moreover, the column effi-ciency towards the studied cations was significantly improved by raising the tempera-ture up to 60°C.

667The effect of temperature on the cation-exchange separations in ion chromatography

The aim of this work was to extend and supplement the data concerning the tem-perature effect on the retention properties, selectivity and efficiency of Dionex IonPac columns towards cations and amines. Different column packings containing vario-us functional groups and different eluents were used. Three cation-exchange columnswere investigated: Ion Pac CS10 with strongly acidic SO

3, Ion Pac CS12 with weakly

acidic CO2, and Ion Pac CS12A with a mixture of weakly acidic CO

2 and PO

32

[33]. The temperature was being changed in the range 10°C60°C. Temperatureeffect on retention times of cations and plate heights depended on the column andthe eluent used (hydrochloric aciddiaminopropionic monohydrochloride acid,(HClDAP), methanesulfonic acid (MSA), or H

2SO

4).

Selectivity coefficients were computed and used for the calculations of the changesof enthalpy, free Gibbs energy and enthropy (∆H, ∆G and ∆S, respectively) of the res-pective cation-exchange processes. The values and signs of enthalpies of ion exchangereactions differed significantly depending on the cation, column and eluent used, andwere indicative for either exothermic or endothermic character of the reaction.

The performed comparison of capabilities of different columns towards the sepa-ration of cations in the presence of various eluents may encourage one to use the tem-perature as a separation aid. In contrast to other papers focused on the separations ofcations at ambient or elevated temperatures (19°C65°C) [3032], in this work alsolow temperature (10°C) has been used. It was found that lowering the temperaturemay occasionaly be advantageous for the successful separation.

EXPERIMENTAL

Apparatus

A Dionex 2000i/SP ion chromatograph (Dionex Corporation, 1228 Titanway, Sunnyvale, Ca, USA)equipped with a CDMII Dionex conductivity detector was used. Separations were performed on the ana-lytical columns: Ion Pac CS10, CS12, and CS12A (with guard columns: Ion Pac CG10, CG12, and CG12Arespectively) [33]. Ion Pac CS10 + CG10 cation-exchange column was used together with a chemicalmicromembrane suppressor for cations CMMS1, which was regenerated with a 100 mmol L-1 TBAOHsolution pumped at a flow rate of 5 mL min-1. Ion Pac CS12 + CG12 and CS12A + CG12A columns wereaccompanied by electrochemical self-regenerating suppressor for cations CSRS1. Injections were per-formed manually with a 50 µL Tefzel sample loop (Dionex). The column temperature was controlled ther-mostatically (UTU4 ultrathermostat, Tarnów, Poland) with the accuracy of ±1°C. Both analytical and guardcolumns were encompassed by a water jacket. When low temperatures of the column (e.g. 10°C) wererequired, the circulating water was previously cooled in a refrigerator and slowly adjusted to the requiredtemperature in the ultrathermostat. The temperature of the mobile phase was adjusted by passing the latterthrough the long spiral Tefzel tube (ca 1 m-in-length, 0.3 mm I.D) placed in a water jacket prior to theapplication onto the column. Before entering the CDMII, mobile phase stepwise adopted was able to reachthe ambient temperature level by passing through 0.5 m length of the 0.3 mm I.D. Tefzel tube and themicromembrane or electrochemical suppressor. Slight fluctuations of the temperature, which might affect

668 K. Kulisa

the conductivity of the solution, were compensated by the Temperature Compensation System installed inthe Dionex 2000i/SP ion chromatograph together with conductivity detector CDMII. Temperature compen-sation coefficient in this study was 1.7%/°C. The hold-up volume was measured at each temperature applied,using the water dip. Experimental data were collected using Dionex AI450 Chromatography SoftwareProgramme, (release 09) [34] and IBM PC.

Chemicals

All reagents were of analytical grade and were used without purification. Deionized water was used forthe preparation of all solutions. It was obtained from a MilliQRG ultra-pure water system (Millipore Co.)and deoxygenated with nitrogen. The following eluents and flow rates were applied:

Dionex Ion Pac CS10 + CG10 columns: eluent: 40 mmol L-1 HCl2 mmol L-1 DAP·HCl, flow rate:1 mL min-1.

Dionex Ion Pac CS12 + CG12 columns: eluents: 18 mmol L-1 MSA and 20 mmol L-1 H2SO4, flow rate:1 mL min-1.

Dionex Ion Pac CS12A + CG12A column: eluents: 18 mmol L-1 MSA and 20 mmol L-1 H2SO

4, flow

rate: 1 mL min-1.Eluents were prepared by weighing out the appropriate amount of acids and their dilution in volumetric

flasks. Methanesulfonic acid (MSA) and H2SO

4 were supplied by Fluka Chemie AG, Switzerland.

Diaminopropionic monohydrochloride (DAP·HCl) acid was purchased from Dionex Co., Sunnyvale, USA.1 g L-1 stock solutions of cations were prepared from the respective chloride salts of the highest avail-

able purity by dissolving them in 18 MΩcm ultra-pure water. Mixed standard solutions of cations and amines, comprising Li+, Na+, NH

4+, K+, Mg2+, Ca2+, Rb+, Cs+,

Sr2+, Ba2+, iso-butylamine (IBuA) and iso-propylamine (IPrA) were prepared by the appropriate dilution andmixing of 1 g L-1 stock solutions.

Procedure and calculations

At each investigated column and for each eluent used the chromatograms were obtained as a function ofcolumn temperature. For several inorganic cations, such as alkali and alkaline earth metals, ammoniumcation and some amines the changes of chromatographic parameters were established in the temperaturerange: 10°C60°C for Ion Pac CS10, CS12 and CS12A columns. The temperature was being graduallyraised in the sequence: 10°C, 20°C, 30°C, 40°C and 60°C. The investigated metal cations were Li+, Na+, K+,Mg2+, Ca2+, Rb+, Cs+, Sr2+, Ba2+, the amines were iso-butylamine (IBuA) and iso-propylamine (IPrA).Before each elution run, the chromatographic system was allowed to equilibrate for the period of ca3045 min in order to stabilize thermal conditions of the experiments. Total retention time, t

R, and the hold-

up time, tM

were estimated from the chromatograms, other chromatographic parameters: column efficiency,N (the number of theoretical plates), peak asymmetry factor, As, and resolution, Rs were calculated usingDionex AI450 Chromatography Software Program, (release 09) [34].

Despite high thermal stability of cation-exchangers at elevated temperatures [35], in this study thetemperature did not exceed 60°C. The stability of Ion Pac cation exchangers was investigated by measuringthe total column exchange capacity, P

c. The column exchange capacity was determined from the break-

through curves of sodium ions. To control the exchange capacity of strongly acidic sulfonic functionalgroups in Ion Pac CS10 + CG10 column, break-through curves were recorded between the heating runs. Forthis purpose, 100 mg L-1 solution of Na+ (pH = 5.85) prepared from pure NaCl was passing through thecolumn instead of the eluent. Previously, the column was washed with 40 mmol L-1 HCl2 mmol L-1 DAP·HCl.Total column exchange capacity, P

c of weakly acidic carboxylic and phosphonic functional groups in Ion


Pac CS12 + CG12 and CS12A + CG12A columns was determined analogously using 500 mg L-1 Na+ solu-tion (pH = 11.85) prepared from pure NaOH. The columns were previously washed with 18 mmol L-1 MSAor 20 mmol L-1 H

2SO

4. The exchange capacities (µeq/col.) were calculated using the following formula:

( )( )+

+−=

Na

NaM0.5R

cM

CttvP (1)

where Pc total column exchange capacity of strongly or weakly acidic groups (µeq/col.), v eluent flow

rate (mL min-1), tR(0.5) retention time (at the 0.5 height of the breakthrough curve, min), tM hold-up time(min), CNa+ Na+ concentration in the solution (mg mL-1), MNa+ gram equivalent of Na+ ion (22.99).

The cation-exchange columns used in the experiments were not the new ones and their real total columnexchange capacities, P

c were slightly lower compared with the nominal values. Nevertheless, determined P

c

values were practically constant during the study (maximum change was 1% of nominal Pc value) for all

three columns examined. This was an evidence for the satisfactory stability of Ion Pac cation-exchangers atelevated temperatures (up to T = 60°C).

Microsoft Excel 97 software was applied for the processing of the obtained data and for the calcula-tions of linear regression coefficients and the temperature changes of thermodynamic functions: ∆H, ∆S and∆G.

RESULTS AND DISCUSSION

The effect of temperature on selectivity and retention

Vant Hoff plots of log vs T-1 were described in many papers. They were usedas basic characteristics of selectivity changes as a function of temperature in classicalion-exchange chromatography [719]. Exothermic or endothermic character of ionexchange process was evaluated from the slope of linear vant Hoff plots also for lowcapacity ion exchange resins in IC [23,28,31,32]. The temperature effect on the selec-tivity of the separation is evident if one considers enthalpy values of ion-exchangeprocesses of different ions. In IC, most of investigators evaluated enthalpy changesaccompanying ion-exchange processes from the temperature changes of retention factor[27,28,31,32]. However, in order to estimate ∆G, ∆H and ∆S in IC and to comparethem with the values obtained for classical ion-exchange chromatography, the equili-brium constant for the respective ion-exchange reaction should be taken into account.For the ion-exchange reaction:

p

qBA

K

q

p

p

q Aq

1BR

p

1B

p

1AR

q

1 +=+ (2)

where R structural unit of the resin, Aq and Bp counter ions from the eluent andsolute of the charge q and p, respectively.

670 K. Kulisa

The selectivity coefficient, p

qBA

K is equivalent to the equilibrium constant (acti-vity coefficients were assumed to be equal to unity in dilute solutions) and is given bythe formula [7,36]:

p

1

r

p

1

q

1

A

p

1

B

p

1

B

q

1

AR

q

1

A

p

1

BRB

A

C

dm

mN

mNK

qp

pq

qpp

q

⋅⋅=

⋅

⋅=

λ(3)

where: N mole fraction in the resin phase, m molality of the solution, d eluentdensity (g cm-3), C

r concentration of the resin phase (mmol g-1 of dry resin), and

λ mass distribution constant (earlier called weight distribution coefficient, i.e. amountper g of dry resin / amount per mL of the solution). The latter one is given by therelation:

r

M

B m

Vkp =λ

where k retention factor, VM

/mr the ratio of the volume of the mobile phase in the

column to the mass of dry ion exchanger.Equation (3) was obtained assuming that the Bp ion occurs in trace amounts.Free energy change, ∆G is given by [10,39]:

∆G = RT ln KBp

Aq

Enthalpy change accompanying the ion-exchange reaction (equation (2)) may becalculated from the temperature dependence of the selectivity coefficients (cf. Eq.(3)), i.e. from the vant Hoffs isochore [10,39]:

(4)

(5)

( )1/Td

dlnKR H

pB

qA−=∆

where ∆H enthalpy change, p

qBA

K selectivity coefficient, R gas constant, 1.987cal K-1 mol-1.

Occassionaly for some ion-exchange reactions, enthalpy change may be constantover a particular temperature range, yet this is not a rule. Thus, it was assumed thatenthalpy can vary with temperature as follows:

∆H = ∆H0 + ∆C

pT

where ∆Cp heat capacity.

(6)

(7)


Selectivity coefficients for the respective cation-exchange reactions in the tempe-rature range of 10°C60°C, as well as the corresponding values of enthalpy changeswere calculated from the experimental data according to [37,38].

Entropy changes can be then obtained from the relationship:

T

GHS

∆∆∆ −= (8)

Temperature changes of selectivity coefficients of Li+, Na+, NH4

+, K+, Mg2+, Ca2+,Rb+, Cs+, Sr2+, Ba2+, IBuA and IPrA observed for the investigated cation-exchangecolumns: Ion Pac CS10 + CG10, Ion Pac CS12 + CG12 and Ion PacCS12A + CG12Aare shown in Figures 1,2,5.

0.36

0.72

1.08

1.44

1.80

0.00

Log

KH

+Cat

ion

60 50 40 30 20 10

0.30 0.31 0.32 0.33 0.34 0.35

Temperature, °C

1/T × 10-2 [1/K]

IBuA

IPrACs+

Rb+

K+

Li+Na+NH4

+

Temperature, °C

60 50 40 30 20 10

1/T × 10-2 [1/K]

0.30 0.31 0.32 0.33 0.34 0.35

0.36

0.72

1.08

1.44

1.80

0.00

Ba2+

Sr2+

Ca2+

Mg2+

Log

KH

+Cat

ion

a b

Figure 1. Vant Hoff plots of several cations and amines. Column: Dionex Ion Pac CS10 + CG10, eluent:40 mmol L-1 HCl 2 mmol L-1 DAP·HCl, flow rate: 1 mL min-1; temperature range: 10°C60°C;IPrA iso-propylamine, IBuA iso-butylamine; a monovalent alkali metal cations and amines,b divalent alkaline earth metal cations

In Figure 1 a,b vant Hoff characteristics of monovalent alkali metal cations, NH4+,

two amines (1a) and divalent alkaline earth metal cations (1b) separated on DionexIon Pac CS10 column are presented.

672K

. Kulisa

Table 1. Exemplary values of enthalpy changes at 25°C calculated from linear regression slopes of the vant Hoff plots and those calculated from the modelwhere ∆H is a function of temperature for two selected cations (Na+ and Ca2+). Cation-exchange columns: Ion Pac CS10 + CG10; eluent: 40 mmol L-1

HCl 2 mmol L-1 DAP·HCl, flow rate: 1 mL min-1; Ion Pac CS12 + CG12 and CS12A + CG12A; eluents: 18 mmol L-1 MSA and 20 mmol L-1 H2SO4,flow rate: 1 mL min-1

CS10 CS12 CS12A

40 mmol L-1 HCl – 2 mmol L-1 DAP·HCl

18 mmol L-1 MSA 20 mmol L-1 H2SO4 18 mmol L-1 MSA 20 mmol L-1 H2SO4

∆H (kcal equiv-1)

∆H (kcal equiv-1)

∆H (kcal equiv-1)

∆H (kcal equiv-1)

∆H (kcal equiv-1)

Linear regression

Non-linear model

Linear regression

Non-linear model

Linear regression

Non-linear model

Linear regression

Non-linear model

Linear regression

Non-linear model

Analyte

r2 ∆H ∆H25ºC r2 ∆H ∆H25ºC r2 ∆H ∆H25ºC r2 ∆H ∆H25ºC r2 ∆H ∆H25ºC

Na+ 0.644 –0.08 –0.32 0.961 –0.83 –2.33 0.450 0.14 –1.54 0.969 –0.66 –1.70 0.842 –0.04 –0.21

Ca2+ 0.994 0.19 0.42 0.997 –0.09 –0.22 0.841 0.13 0.14 0.959 –0.11 –0.31 0.970 0.07 0.08


The increase of the temperature from 10°C up to 60°C resulted in significant changesof the column selectivity towards monovalent cations, amines and divalent cations.Increased temperature caused a decrease in retention factors of monovalent cationsand amines, while retention of divalent cations considerably increased. However, onlyfor some of the investigated analytes (Rb+, Cs+, Mg2+, Ca2+ and Ba2+) the linearity ofvant Hoff plots was satisfactory and the values of correlation coefficients (r2) wereclose to 1. r2 values for Li+, Na+, K+, NH

4+, Sr2+ and IPrA were distinctly lower

(see Tab. 1, examples of Na+ and Ca2+) and thus the corresponding vant Hoff plotsexhibited a slight curvature. Enthalpy changes (∆H) of ion-exchange reactions ofthese analytes were temperature-dependent. Their signs were negative for monova-lent cations and amines (exothermic character of ion exchange) and positive for diva-lent alkaline earth cations (endothermic character of ion exchange). Vant Hoff plotsin Figure 1a exhibit selectivity changes of the columns towards some monovalentcations and amines (e.g. improved separation between IPrA and Cs+ at 60°C). Gene-rally, the separation between alkali metal cations as a group and alkaline earth metalswas improved with the increase of the temperature.

Ion Pac CS12 and CS12A are the cation-exchange packings of different structurethan the CS10 cation exchanger [30]. In Figure 2 a,b respectively, the vant Hoff plotsfor monovalent and divalent cations separated on CS12 column using 18 mmol L-1

MSA eluent are presented. Only for some of the analytes linear vant Hoff relation-ships were obtained. For the others, e.g. Li+, Na+, NH

4+ and Sr2+ these plots exhibited

distinct curvature. The sign of ∆H was negative for all the analytes, which was indica-tive for exothermic character of the ion-exchange reactions. For several analytes sub-stantial changes of selectivity between some analytes followed the changes of thetemperature. A good example is the temperature influence on retention of Rb+, Cs+,K+ and IPrA. At 40°C the reversal of the elution order between Cs+ and IPrA wasobserved (see Fig. 3 a,b), as well as co-elution of IPrA and Rb+ occurred at 10°C. Li+,Na+ and NH

4+ were co-elute at 60°C. The best separation of these cations was achieved

at 10°C (see Fig. 4 a,b).The use of 20 mmol L-1 H

2SO

4 as the eluent considerably changed the tempera-

ture behaviour of the analytes separated on Ion Pac CS12 column. According to thedata in Table 1, linearity of vant Hoff plots (Figure 2 c,d) was significantly poorer(particularly for Na+, NH

4+, K+, Ca2+ and Sr2+) compared with that obtained with the

use of MSA eluent. Some cations (Li+, Na+, Mg2+, Ca2+ and Sr2+) revealed endother-mic temperature behaviour (positive sign of ∆H). Usually this phenomenon is regar-ded as a result of complexation between cations of divalent alkaline earth metals andsulphuric acid eluent at a room temperature [30]. Instability of such complexes atelevated temperatures may cause an increased retention of divalent cations. This hypo-thesis, however, has not been satisfactorily confirmed in the experiment. For example,monovalent Li+ and Na+ exhibit an increased retention with the rise of temperature,

674 K. Kulisa

while retention of divalent Ba2+ slightly decreased. Selectivity changes of CS12 col-umn towards Na+ and NH

4+ led to their co-elution at elevated temperature (60°C), as

well as for Rb+ and K+ pair co-elution occurred at 10°C.

Log

KH

+Cat

ion

Log

KH

+Cat

ion

Temperature, °C

Log

KH

+Cat

ion

Figure 2. Vant Hoff plots of several cations and amines. Column: Dionex Ion Pac CS12 + CG12, eluentflow rate: 1 mL min-1; temperature range: 10°C60°C. IPrA iso-propylamine, IBuA iso-buty-lamine; a monovalent alkali metal cations and amines, eluent: 18 mmol L-1 MSA, b divalentalkaline earth metal cations, eluent: 18 mmol L-1 MSA, c monovalent alkali metal cations andamines, eluent: 20 mmol L-1 H

2SO

4, d divalent alkaline earth metal cations, eluent: 20 mmol L-1

H2SO4

Log

KH

+Cat

ion

1.57

1.29

1.01

0.73

0.45

1.85

60 50 40 30 20 10

0.30 0.31 0.32 0.33 0.34 0.35

Temperature, °C

1/T × 10-2 [1/K]

IBuA

IPrA

Cs+

Rb+

K+

Li+Na+

NH4+

a

1.57

1.29

1.01

0.73

0.45

1.85

60 50 40 30 20 10

0.30 0.31 0.32 0.33 0.34 0.35

1/T × 10-2 [1/K]

Sr2+

Ba2+

Mg2+Ca2+

b

1.57

1.29

1.01

0.73

0.45

1.85

60 50 40 30 20 10

0.30 0.31 0.32 0.33 0.34 0.35

Temperature, °C

1/T × 10-2 [1/K]

IBuA

IPrA

Cs+

Rb+

K+

Li+Na+

NH4+

c

1.57

1.29

1.01

0.73

0.45

1.85

60 50 40 30 20 10

0.30 0.31 0.32 0.33 0.34 0.35

Temperature, °C

1/T × 10-2 [1/K]

Sr2+

Ba2+

Mg2+Ca2+

d


Figure 3. Elution curves of K+, Rb+, Cs+ and IPrA. Ion Pac CS12 + CG12 column, eluent 18 mmol L-1

MSA. Temperature: a 20°C, b 60°C

a

b

Time, min

Con

duct

ivity

, µs

Time, min

Con

duct

ivity

, µs

676 K. Kulisa

a

b

Figure 4. Elution curves of Li+, Na+, NH4+, K+, Mg2+ and Ca2+. Ion Pac CS12 + CG12 column, eluent

18 mmol L-1 MSA. Temperature: a 10°C, b 60°C

Time, min

Con

duct

ivity

, µs

Time, min

Con

duct

ivity

, µs


Figure 5. Vant Hoff plots of several cations and amines. Column: Dionex Ion Pac CS12A + CG12A, eluentflow rate: 1 mL min-1. Temperature range: 10°C60°C. IPrA iso-propylamine, IBuA iso-bu-tylamine; a monovalent alkali metal cations and amines, eluent: 18 mmol L-1 MSA, b divalentalkaline earth metal cations, eluent: 18 mmol L-1 MSA, c monovalent alkali metal cations andamines, eluent: 20 mmol L-1 H

2SO

4, d divalent alkaline earth metal cations, eluent: 20 mmol L-1

H2SO

4

The Ion Pac CS12A cation-exchange column was introduced as the improvedhydrophilic version of Ion Pac CS12 stationary phase. According to the literature data

Log

KH

+Cat

ion

Log

KH

+Cat

ion

Log

KH

+Cat

ion

Log

KH

+Cat

ion

Temperature, °C

1.88

1.66

1.44

1.22

1.00

2.10

60 50 40 30 20 10

0.30 0.31 0.32 0.33 0.34 0.35

Temperature, °C

1/T × 10-2 [1/K]

IBuA

IPrA

Cs+

Rb+

K+

Li+Na+

NH4+

a

60 50 40 30 20 10

1.88

1.66

1.44

1.22

1.00

2.10

b

Ba2+

Mg2+

Ca2+Sr2+

1/T × 10-2 [1/K]

60 50 40 30 20 10Temperature, °C

0.30 0.31 0.32 0.33 0.34 0.351/T × 10-2 [1/K]

1.62

1.44

1.26

1.08

0.90

1.80

IBuA

IPrA

Cs+

Rb+

K+

Li+

Na+

NH4+

c

60 50 40 30 20 10Temperature, °C

Temperature, °C

1.62

1.44

1.26

1.08

0.90

1.80

d

Ba2+

Mg2+

Ca2+

Sr2+

1/T × 10-2 [1/K]

678 K. Kulisa

[30], the introduction of phosphonate functional groups to the stationary phase impro-ves the column efficiency and chromatographic peak shapes by preventing the innercore penetration of the polymer stationary phase and support particles. This is impor-tant in the separation of hydrophobic amines, which may be retained on the stationaryphase via the combination of cation-exchange and adsorption processes.

In Figure 5 a,b the vant Hoff plots for cations and amines eluted from the Ion PacCS12A column using 18 mmol L-1 MSA eluent are shown. The signs of ∆H werenegative, which is typical of exothermic ion exchange. Indeed, retention of the aboveanalytes decreased with the temperature. Linearity of the vant Hoff plots was rathergood (see examples in Tab.1). Substantial changes of the CS12A column selectivitytowards several investigated analytes were noticed. For example, the reversal of theelution order for the IPrA and Rb+ pair at 60°C and IPrAK+ pair at 35°C was obser-ved. The best separation quality for K+, Rb+, Cs+ and IPrA was achieved at 10°C.

Temperature behaviour of the analytes studied was significantly changed when MSAeluent was replaced with 20 mmol L-1 H

2SO

4. Vant Hoff plots obtained for these sepa-

rations are shown in Figure 5 c, d. For some analytes (Li+, Mg2+, Ca2+, Sr2+ and IPrA) thesigns of ∆H were positive, which was indicative for endothermic ion exchange. Linea-rity of vant Hoff plots was improved compared to that observed for CS12 column andH

2SO

4 eluent (see examples in Tab.1). The temperature of 10°C was more convenient

for the separation of Na+ and NH4+, which co-elute at 60°C (see Fig. 6a,b). The reversal

of the elution order for IPrARb+ pair at 60°C and IPrAK+ pair at 35°C occurred.

Figure 6. Elution curves of Li+, Na+, NH4+, K+, Mg2+ and Ca2+. Ion Pac CS12A + CG12A column, eluent

18 mmol L-1 H2SO4. Temperature: a) 10°C, b) 60°C

Time, min

Con

duct

ivity

, µS

Time, min

Con

duct

ivity

, µS

a b

679T

he effect of temperature on the cation-exchange separations in ion chrom

atographyTable 2. Exemplary values of the changes of thermodynamic functions calculated from elution curves for two selected cations (Na+ and Ca2+) as a function of

temperature. Columns: Ion Pac CS10 + CG10; eluent: 40 mmol L-1 HCl 2 mmol L-1 DAP·HCl, flow rate: 1 mL min-1; Ion Pac CS12 + CG12 andCS12A + CG12A; eluent: 18 mmol L-1 MSA, flow rate: 1 mL min-1

Temperature, K

283 293 303 313 333 Ion

Exchange Reaction

Thermodynamic Function

CS10 CS12 CS12

A CS10 CS12

CS12A

CS10 CS12 CS12

A CS10 CS12

CS12A

CS10 CS12 CS12

A ∆G

(kcal equiv–1) –0.32 2.04 2.21 –0.32 2.21 2.35 –0.32 2.36 2.49 –0.32 2.50 2.62 –0.35 2.73 2.87

∆H (kcal equiv–1)

–0.54 –3.05 –2.00 –0.39 –2.57 –1.80 –0.25 –2.09 –1.61 –0.10 –1.61 –1.41 0.20 –0.66 –1.02 R[H+]

→R[Na+] ∆S

(cal K–1 equiv–1) –0.76 –18.0 –14.9 –0.25 –16.3 –14.2 0.24 –14.7 –13.5 0.72 –13.2 –12.9 1.63 –10.2 –11.7

∆G (kcal equiv–1)

–0.04 0.89 0.88 –0.05 0.92 0.93 –0.07 0.96 0.97 –0.09 1.00 1.00 –0.13 1.08 1.08

∆H (kcal equiv–1)

0.35 –0.18 –0.40 0.40 –0.20 –0.34 0.45 –0.22 –0.27 0.49 –0.25 –0.20 0.59 –0.29 –0.07 R[H+]

→R[Ca2+] ∆S

(cal K–1 equiv–1) 1.4 –3.77 –4.53 1.55 –3.84 –4.31 1.71 –3.91 –4.09 1.85 –3.98 –3.88 2.14 –4.11 –3.47

680 K. Kulisa

The calculation of ∆H from the averaged slopes of vant Hoff plots, using theequation: slope = ∆H/R, is the most popular approach in ion chromatography repor-ted in the literature [23,32]. However, the obtained values of ∆H are only approxi-mate values, averaged for the entire temperature range applied. They presumablydiffer from the values estimated separately for individual temperatures, especiallywhen the linearity of vant Hoff plot is poor (r2 < 1). In order to check this prediction,∆H values calculated from the averaged slopes of vant Hoff plots were compared tothose estimated from elution curves of Na+ and Ca2+ according to non-linear model[37,38] (averaged for T = 25°C). The results are presented in Table 1. The observeddistinctions between ∆H values can be assigned to the differences in shape and linea-rity of individual vant Hoff plots. An extreme case is a reversal of ∆H sign for ion-exchange reaction of Na+ on the Ion Pac CS12 column with H

2SO

4 eluent (cf. Fig. 2 c

and Tab. 1).∆G, ∆H and ∆S values for Na+ and Ca2+ are presented in Table 2. They occurred to

be strongly temperature-dependent. The comparison of ∆H values in Tables 1 and 2reveals explicitly the differences between ∆H values estimated from the averagedslopes of vant Hoff plots and those calculated taking into account their non-linearity[37,38]. The former are presumably only approximate, while the latter provide moreprecise information on the influence of the temperature on ion-exchange processes.

The effect of temperature on column efficiency and quality of separations

As was pointed out by Dybczyñski [10], substantial changes in ion-exchange equili-brium and in the values of separation factors of several pairs of ions in numerous ionexchange systems can be caused by variations of the temperature. Also their separa-tion quality is often additionally improved as a result of better column performance(efficiency and peak shape). Exemplary temperature dependencies of chromatographicparameters, i.e. retention factor (k), number of theoretical plates (N), asymmetry fac-tor (A

s), separation factor (α), and resolution (R

s) for selected cations and amines

(Na+, K+, Ca2+, Sr2+, Ba2+, IPrA and IBuA) are shown in Table 3 A,B. From the ob-tained data one can conclude that the number of theoretical plates (N) usually, increa-ses with the temperature. However, in several situations the maximum of N occurredat 40°C and at 60°C N slightly decreased. Only for IPrA and IBuA the values of Nicreased monotonically up to 60°C. The improvement of column efficiency with thetemperature rise was more pronounced for strongly acidic sulfonated Ion Pac CS10column than for weakly acidic Ion Pac CS12A and CS12 columns.


IPrA (values of á and Rs – for ion pair IPrA - K+)

IBuA (values of á and Rs – for ion pair IBuA - IPrA) T, °C

k N As á Rs k N As

á Rs

10 4.18 1315 4.22 1.43 2.19 6.72 156 3.12 1.61 0.97

20 4.18 1366 4.20 1.63 2.84 6.69 182 3.05 1.60 1.05

30 4.18 1526 4.00 1.65 3.06 6.52 190 3.00 1.56 1.08

40 4.16 1865 3.38 1.65 3.38 6.40 197 2.66 1.54 1.09

60 4.11 2240 3.17 1.70 3.80 6.02 294 2.72 1.46 1.17

Ca2+

(values of á and Rs – for ion pair Ca2+- IBuA) Sr2+

(values of á and Rs – for ion pair Sr2+- Ca2+)

T, °C k N As á Rs k N As á Rs

10 29.2 5148 1.00 4.34 13.2 43.3 4535 1.64 1.48 21.1

20 32.1 7359 1.01 4.79 16.6 44.2 6136 1.65 1.38 20.7

30 35.2 7450 1.50 5.39 16.8 44.2 6326 1.65 1.25 15.5

40 36.6 7811 1.92 5.72 18.0 44.7 6881 1.71 1.22 14.6

60 47.3 7980 1.80 7.85 18.2 50.7 6880 1.70 1.07 5.60

(Continuation on the next page)

Na+ K+

(values of á and Rs – for ion pair K+- Na+) T, °C k N As k N As á Rs

10 1.86 1539 1.79 2.91 1890 2.84 1.56 11.6

20 1.78 1731 1.77 2.56 1971 2.32 1.44 9.60

30 1.78 1760 1.70 2.53 1990 2.20 1.42 9.26

40 1.76 1802 1.60 2.52 1970 2.05 1.44 9.85

60 1.76 1380 1.60 2.40 1606 2.00 1.36 7.35

Table 3. Exemplary variations of chromatographic parameters observed for selected cations and aminesas a function of temperature

A. Column: Dionex Ion Pac CS10; eluent: 40 mmol L-1 HCl 2 mmol L-1 DAP·HCl, flow rate: 1 mL min-1

682 K. Kulisa

Table 3. (Continuation)

Ba2+

(values of á and Rs – for ion pair Ba2+- Sr2+)

T, °C k N As á Rs

10 50.7 4033 1.66 1.17 9.02

20 52.5 5686 1.63 1.18 11.2

30 53.2 5712 1.58 1.20 12.3

40 54.2 6138 1.52 1.21 13.3

60 58.5 6190 1.54 1.15 10.1

Na+ IPrA

(values of á and Rs – for ion pair IPrA/K+- Na+)

T, °C k N As k N As á Rs

10 0.67 1243 0.95 1.05 848 2.12 1.57 5.36

20 0.55 1255 0.93 0.97 1230 2.34 1.76 7.98

30 0.53 1260 0.93 20.96 1362 2.00 21.81 27.94

40 0.51 1273 0.93 20.94 1527 1.90 21.76 28.45

60 0.46 1280 0.94 0.90 1407 1.95 1.60 6.90

K+

(values of á and Rs – for ion pair K+- IPrA)

IBuA (values of á and Rs – for ion pair IBuA -

K+/IPrA) T, °C

k N As á Rs k N As á Rs

10 1.44 1454 1.55 1.37 1.35 2.79 423 4.97 1.94 1.83

20 1.05 1404 2.26 1.08 0.34 2.38 443 4.00 2.07 2.31

30 20.98 1666 2.01 21.02 20.09 2.26 806 4.00 2.31 2.77

40 20.90 1725 1.68 21.04 20.09 2.12 815 3.65 2.25 2.67

60 0.74 1873 1.70 1.21 0.79 1.85 941 2.70 2.05 2.55

2 Change of elution order of the pair IPrAK+.

(Continuation on the next page)

B. Column: Dionex Ion Pac CS12A; eluent: 18 mmol L-1 MSA, flow rate: 1 mL min-1


Table 3. (Continuation)

Ba2+

(values of á and Rs – for ion pair Ba2+- Sr2+)

T, °C k N As á Rs

10 8.26 2355 2.77 1.79 19.0

20 7.05 2417 2.68 1.85 19.4

30 5.54 2688 2.42 1.51 14.8

40 5.20 2918 2.35 1.46 14.3

60 3.93 2645 1.98 1.32 9.83

The shape of chromatographic peaks has been generally improved with the tem-perature rise, particularly for the amines. Column efficiencies towards the investi-gated amines (especially for IBuA) were considerably lower than for metal cations,which is obvious from the data presented in Table 3 A, B. This phenomenon has beenalready reported in the literature and explained in terms of reversed-phase hydropho-bic nature of interactions of amines with the resin phase [30]. The increased tempera-ture reduces this effect, and leads to the significant improvement of column efficiencyand peak shape for amines similarly as organic modifier (e.g. acetonitrile) [31].

For several pairs of ions the changes of temperature substantially improved theresolution. For example, a significant increase of R

s values for the Ca2+ IBuA pair

separated on the Ion Pac CS10 column using 40 mmol L-1 HCl 2 mmol L-1 DAP·HCleluent was observed at elevated temperatures. R

s values increased consequently as

separation factors, α monotonically rose with the temperature. However, the effect ofthe enhanced column efficiency for Ca2+ and IBuA at elevated temperatures is alsoevident (see Tab. 3A). Similar observations were made for Ion Pac CS12A columnusing MSA as the eluent for the separation of the same pair of analytes. This timea slight increase of α values was accompanied by a significant increase of column

Ca2+

(values of á and Rs – for ion pair Ca2+- IBuA) Sr2+

(values of á and Rs – for ion pair Sr2+- Ca2+)

T, °C k N As á Rs k N As á Rs

10 3.71 2253 2.40 1.33 2.24 4.61 2295 2.44 1.24 7.47

20 3.16 2771 2.66 1.32 2.40 3.80 2634 2.03 1.20 6,47

30 3.12 2990 2.55 1.38 2.84 3.68 2765 2.00 1.18 6.28

40 3.11 3006 2.53 1.37 2.90 3.57 2900 1.99 1.15 5.46

60 2.74 3204 2.22 1.48 3.32 2.97 2990 1.93 1.08 3.03

684 K. Kulisa

efficiency for both Ca2+ and IBuA (e.g. N was doubled at 10°C and 60°C for IBuA)(see Tab. 3B). The improvement of column efficiency towards amines at elevatedtemperatures was of great importance also for their mutual separation (e.g. separationof IBuA and IPrA on Ion Pac CS10). Despite a decrease of α with the temperaturerise, R

s increased by 20% within the temperature range 10°C60°C (see Tab. 3A).

In contrast to Ion Pac anion-exchange columns [29], no undesirable phenomenaoriginating from thermal degradation of resin bed appeared were observed duringexperiments. Chromatographic parameters, such as retention factor, peak asymmetryand the number of theoretical plates were not affected on cation-exchange columns atelevated temperatures (up to T = 60°C), as well as no changes in the resin phase(decrease of the total column exchange capacity) were observed. Some chromatogra-phists recommend columns of smaller diameter (i.e. 2 mm or less) when non ambienttemperatures are used in HPLC. The columns of smaller diameter are easier to equili-brate at a given temperature [40]. However, the procedure of thermal equilibrationestablished for Ion Pac anion-exchange columns [29], was successful also for Ion Pac4 mm cation exchangers. It allowed one to stabilise chromatographic parameters, too.The effect of longer tube-connection between the column outlet and conductometricdetector could be also neglected.

Possible mechanisms of zone spreading

The plate heights, H were calculated from elution curves of selected monovalentalkali metal and divalent alkaline earth metal cations (Na+, K+, Ca2+, Sr2+ and Ba2+),obtained at 100C and 600C on three cation-exchange columns: Ion Pac CS10 + CG10,CS12 + CG12 and CS12A + CG12A and using 40 mmol L-1 HCl / 2 mmol L-1 DAP·HCl,18 mmol L-1 MSA and 18 mmol L-1 MSA solutions, respectively as the eluents.

In Figure 7 a, b, c H values are plotted versus reciprocal of the weight distributioncoefficient, 1/λ. For Ion Pac CS10 cation-exchanger and 40 mmol L-1 HCl / 2 mmol L-1

DAP·HCl eluent, the column efficiency for divalent cations Sr2+ and Ba2+ showinghigh affinity to the ion exchanger (λ > 22) was slightly poorer than for Ca2+ (λ = 15).The H vs 1/λ plots, after reaching a minimum, increased both at low (10°C) andelevated (60°C) temperatures. Similar column performance behaviour was observedduring the separations of cations on Ion Pac CS12 and CS12A columns with somesubtle differences however. In the case of CS12A column, distinct decrease in thecolumn efficiency for divalent cations Sr2+ and Ba2+, showing high affinity to the ionexchanger (λ > 5) was observed at elevated temperature (60°C) as compared with thatfor Ca2+ (λ = 4.3) (see Fig. 7 c). This phenomenon did not occur at 10°C. In the caseof ion Pac CS12 cation-exchanger, at 10°C, column performance for Sr2+ and Ba2+

(λ > 4.5) was poorer than that for Ca2+ (λ = 3.8). At 60°C the column efficiency for alldivalent cations (Ca2+, Sr2+ and Ba2+) was poorer than that observed for monovalentalkali metal cations (see Fig. 7 b).


1/λ

H, m

m

1/λ

H, m

m

1/λ

H, m

m

Figure 7. Plate heights calculated from elution curves of Na+, K+, Ca2+, Sr2+ and Ba2+ at 10°C and 60°Cobtained on Ion Pac CS10 + CG10 column and 40 mmol L-1 HCl 2 mmol L-1 DAP·HCl eluent,flow rate: 1 mL min-1 (a), Ion Pac CS12 + CG12 column and 18 mmol L-1 MSA eluent, flow rate:1 mL min-1 (b), and Ion Pac CS12A + CG12A column and 18 mmol L-1 MSA eluent, flow rate:1 mL min-1 (c), as a function of reciprocal of the distribution coefficient. Open circles ():T = 10°C, full circles (): T = 60°C

For a given value of distribution coefficient, the plate height generally decreaseswith the increase in temperature for three columns studied especially for ions show-ing higher affinity to the resin. For cations showing low distribution coefficients theplate height sometimes first decreased and then increased with the further rise oftemperature (see Fig. 8 a, b, c).

a b

c

686 K. Kulisa

Figure 8. Temperature dependence of the plate height (normalized with respect to individual values of theweight distribution coefficient values) as a function of temperature: Ion Pac CS10 + CG10column and 40 mmol L-1 HCl 2 mmol L-1 DAP·HCl eluent, flow rate: 1 mL min-1; λ = 1;¢ λ = 20 (a); Ion Pac CS12 + CG12 column and 18 mmol L-1 MSA eluent, flow rate: 1 mLmin-1; λ = 1; ¢ λ = 5 (b); Ion Pac CS12A + CG12A column and 18 mmol L-1 MSA eluent,flow rate: 1 mL min-1; λ = 1; ¢ λ = 5 (c). Continuation on the next page

The above observations are consistent with those reported by Dybczyñski andKulisa and referring to the temperature effect on the performance of Ion Pac AS9SCanion-exchanger column [29]. In classical ion-exchange chromatography, the mecha-nism of increase of the plate height with the increase of retention factor and of tempe-

a b

c


( )( )

( ) ( )( ) ui

D2

u

Di2

u70r1i11D

ur0.266

Di

ur0.142rùH

'S

0

2'

20

2'

2'

20

'

0

ëã

ë

ë

ë

ë +++−+

++

+= (9)

rature has been difficult to explain using Van Deemter, Glueckauf [41] or Hamilton[21] equations. These equations are usually used for the description of the zone spread-ing in HPLC. The above mentioned equations contain mostly terms in which diffu-sion coefficients are in denominator, hence the plate height should rather decreasewith the rise in temperature (diffusion coefficients are expected to increase with therise of temperature) [4245].

Adapting to expression for diffusion derived by Giddings [46], Dybczyñski[4749] proposed to incorporate to the Hamilton equation the term describing thecontribution of longitudinal diffusion in the stationary phase to the total plate height:

where H = L/N plate height, (L column length), r0 particle radius, D and D

diffusion coefficients in the resin phase and in the solution, respectively, λ= λ dz

bed distribution coefficient (dz bed density), u linear flow rate, i fractional void

volume of the bed, ω and γs constants.

In equation (9), the terms from the left to right represent the contribution of Eddydiffusion, diffusion within resin particles, diffusion within the liquid film adherent tothe resin particles, longitudinal diffusion in the mobile phase, and longitudinal diffu-sion in the resin phase, respectively.

Equation (9) was applied to explain the dependence of the plate height upon theion diffusion coefficient in the stationary phase in classical ion-exchange chroma-tography, e.g. in the case of separation of alkaline earth metals [48] on variouslycross-linked ion exchange columns. The formula (9) is also able to explain the shapeof H vs 1/λ plots under conditions when longitudinal diffusion in the stationary phase,conrtributes significantly to the total plate height. Equation (9) allows one to correctlyinterpret the changes of plate heights for cations of high retention factors, similarilyas it was the case for anion-exchange columns [29].

CONCLUSIONS

Similarly to classical ion-exchange chromatography, also in ion chromatographycolumn temperature occurred to be an important factor modifying the selectivity ofseparations of alkali metal, alkaline earth metal cations and amines. The changes ofenthalpy for cation-exchange reactions of several cations differ significantly and usu-ally are temperature-dependent. ∆H values should be calculated individually for eachtemperature, and not from the slopes of vant Hoff plots. A priori assumption of the

688 K. Kulisa

linearity of vant Hoff plots [31,32] may be erroneous. The increased temperatureusually improved column efficiency by decreasing the plate height. However, in somecases a reverse tendency was observed at high temperatures (60°C). Another conclu-sion is that longitudinal diffusion in the resin phase may contribute to the total col-umn plate height for some ions of high retention factors, which has not been yet takeninto account in the literature. The three investigated columns exhibited good thermalstability over the temperature range applied (10°C60°C). Advantageous changes inselectivity of separations and column efficiency resulted in the improvement of separa-tion factors of several pairs of cations and amines at elevated and, in some cases, atlowered temperature (10°C). In general, the column temperature should be conside-red as an important parameter improving the separations of cations and amines in ionchromatography.

Acknowledgements

The author is grateful to Prof. dr hab. R. Dybczyñski for his advice and helpful discussions. Mr. E.Romanowski M.Sc. and the staff of A.G.A. Analytical, Warsaw are acknowledged for the technical sup-port with Dionex chromatographic system.

REFERENCES

1. Snyder L.R., J. Chromatogr. B, 689, 105 (1997).2. Dolan J.W., Snyder L.R., Blanc T and van Heukelem L., J. Chromatogr. A, 897, 37 (2000).3. Dolan J.W., Snyder L.R. and Blanc T., J. Chromatogr. A, 897, 51 (2000).4. Yan B., Zhao J., Brown J.S., Blackwell J. and Carr P.W., Anal. Chem., 72, 1253 (2000).5. Ranatunga R.P.J. and Carr P.W., Anal. Chem., 72, 5679 (2000).6. G³ód B.K., Alexander P.W., Chen Zu.L. and Haddad P.R., Anal. Chim. Acta, 30, 267 (1995).7. Dybczyñski R., Anal. Chim. Acta., 29, 369 (1963).8. Dybczyñski R., J. Chromatogr.,14, 79 (1964).9. Dybczyñski R., Nukleonika,12, 927 (1967).

10. Dybczyñski R., J. Chromatogr., 31, 155 (1967).11. Wódkiewicz L. and Dybczyñski R., J. Chromatogr., 32, 394 (1968).12. Wódkiewicz L. and Dybczyñski R., Chem. Anal., (Warsaw) 14, 437 (1969).13. Dybczyñski R. and Maleszewska H., Analyst, 94, 527 (1969).14. Dybczyñski R., Proceedings of the 3rd. Analytical Conference, Budapest, August, 1970, pp.

3543.15. Dybczyñski R. and Maleszewska H., J. Radioanal. Chem., 21, 229 (1974).16. Wódkiewicz L. and Dybczyñski R., J. Chromatogr., 102, 277 (1974).17. Dybczyñski R., Polkowska-Motrenko H. and Shabana R.M., J.Chromatogr., 134, 285 (1977).18. Polkowska-Motrenko H. and Dybczyñski R., J. Radioanal. Chem., 59, 31 (1980).19. Wódkiewicz and Dybczyñski R., Chem. Anal. (Warsaw), 26, 419 (1981).20. Baba Y., Yoza N. and Ohashi S., J. Chromatogr., 348, 27 (1985).21. Hamilton P.B., Bogue D.C. and Anderson R.A., Anal. Chem., 32, 1782 (1960).


22. Pohl C.A., Stillian J.R. and Jackson P.E., J. Chromatogr. A, 789, 29 (1997).23. Fortier N.E. and Fritz J.S., Talanta, 34, 415 (1987).24. Busch M. and Seubert A., Anal. Chim. Acta, 399, 223 (1999).25. Rey M.A., J. Chromatogr. A, 920, 61 (2001).26. Bashir W. and Paull B., J. Chromatogr. A, 942, 73 (2002).27. Smith R.G., Drake P.A. and Lamb J.D., J. Chromatogr., 546, 139 (1991).28. Hatsis P. and Lucy C.L., J. Chromatogr. A, 920, 3 (2001).29. Dybczyñski R. and Kulisa K., Chromatographia, 57, 475 (2003).30. Rey M.A. and Pohl C.A., J. Chromatogr. A, 739, 87 (1996).31. Hatsis P. and Lucy C.L., Analyst, 126, 2113 (2001).32. Paull B. and Bashir W., Analyst, 128, 335 (2003).33. Dionex Product Selection Guide, Dionex Corporation, Sunnyvale, CA 1997-98.34. AI450 Chromatography Users Guide, Doc.No 034039, Release 09, March 1993.35. Harland C.E., Ion Exchange: Theory and Practice, 2nd. edn., The RSC, Cambridge, 1994.36. Bonner O.D., J.Phys.Chem., 58, 318 (1954).37. Kraus K.A., and Raridon R.J., J. Phys. Chem., 63, 1901 (1959).38. Dybczyñski R., and Wódkiewicz L., J. Inorg. Nucl. Chem., 31, 1495 (1969).39. Dybczyñski R., Roczn. Chem., 41, 1689 (1967).40. Thompson J.D., Brown J.S., and Carr P.W., Anal. Chem., 73, 3340 (2001).41. Glueckauf E., Ion Exchange and its Applications, Society of Chemical Industry, London, 1955,

pp. 34.42. Ravindranath B., Principles and Practice of Chromatography, Horwood Chichester, 1989.43. Neue U.D., HPLC columns: Theory, Technology, and Practice, Wiley-VCH, New York, 1997.44. Ahuja S., Trace and ultratrace analysis by HPLC, Wiley, New York, 1992.45. Snyder L.R., Kirkland J.J., and Glajch J.L., Practical HPLC method development, Wiley,

New York 1997.46. Giddings J.C., Dynamics of Chromatography, Part I, Dekker, New York, 1965.47. Dybczyñski R., J. Chromatogr., 50, 487 (1970).48. Dybczyñski R., J. Chromatogr.,71, 507 (1972).49. Minczewski J., Chwastowska J. and Dybczyñski R., Separation and Preconcentration Methods

in Inorganic Trace Analysis, Horwood, Chichester, 1982.

Received February 2004Accepted July 2004

the effect of temperature on the cation-exchange...

Documents