Trace analysis by crystal sorption detector and recovery of dental waste

by Lee, Chau-wing

Abstract

PART ONE: Recovery of Silver and Mercury from Dental Amalgam Waste.

A simple, inexpensive scheme was developed to recover silver and mercury from silver-tin dental amalgam waste which contained on weight bases 50-60% mercury, 28-35% silver and 10-14% tin, with minor levels of copper and zinc. Direct distillation of the amalgam "under reduced pressure yielded (99.22?.68)% pure mercury with (87.7?.6)% recovery. The recovered mercury could possibly be further purified to (100.03?.14)% by re-distillation. Silver was isolated from

the residue as silver chloride which was then converted to silver oxide and diamino-silver(I) complex for chemical reduction to metal by glucose and electrochemical reduction respectively. The former reductive method recovered (92.4?.5)% silver with (99.5?.1)% purity; while the latter yielded silver with slightly lower recovery, (89.9?.l)%, but higher purity,

(99.99?.17)%. Parameters governing the success of the two recovery routes are discussed. An analytical scheme for investigating the composition of the dental amalgam waste and the purity of the recovered metals is also presented.

- ii -

PART TWO: Coated Piezoelectric Crystal Sorption Detectors.

Gelatin and sodium 1,8-dihydroxy-3,6-naphthalenedisulphonate (chromotropic acid) were chosen from a wide range of

compounds examined as the coatings for the construction of

quartz crystal sorption detectors that were selective to

the determination of moisture and formaldehyde vapour

respectively. Both designed detectors were found to have

long life times especiallly when they were stored under N2

in a desiccator. Frequency drifts of no more than a few

hundred hertzs occurred after about 3 months and could readily

be recovered by blowing warm air for a shortwhile before use.

The effects of the amount of coating material and of potential interferents on the response of the designed detectors were

studied.

A detector with 150 ~g of gelatin as the coating had a sensitivity of about 3.8 Hz per ppm v/v moisture; a linear range up to 400 ppm v/v moisture and a detection limit of

0.04 ppm v/v moisture. Only hydroxy organic compounds interfered and such interference could be removed by a Tenax GC pretrap.

The performance of the detector in measuring the moisture content in compressed gases like methane, ethene, acetylene, hydrogen,

- iii -

oxygen, carbon monoxide and freons was evaluated against

two currently employed methods, viz. absorptigravimetry and electrolysis of sorbed moisture by DuPont .303 moisture analyser. Such comparison revealed the advantages of larger working range, wider applicability and better selectivity

of the designed detector.

A crystal with 58 ~g of chromotropic acid as the coating had a working range of 0-2 ppm v/v and a sensitivity of about 70 Hz per ppm v/v of formaldehyde. Despite its high selectivity against water interference, the latter could not be neglected in atmospheric monitoring because moisture was often over Ixl04 fold that of formaldehyde

in concentration. Such interference could be corrected but

the precision would then be unsatisfactory. Other possible remedies are presented. Anyway the designed detector compared favourably with and often excelled the chromotropic acid colorimetry and Drager tube colorimetry in terms of precision, selectivity, sensitivity and continuous operation mode, as supported by the results of formaldehyde measurements in simulated clinical situations. The detector is especially valuable when

the moisture content of the sample stream could be fixed and controlled as illustrated by the results of determining the formaldehyde released from textile fabrics. The possible

- iv -

mechanisms of interaction between the sorbent coatings and the analyte vapour and any interferents are proposed to

complete the picture of the detection. Possible diversifications of the present research work on piezoelectric crystal detectors are also outlined.

An inexpensive, readily assembled and demounted diffusion cell had been constructed for generating the standard gaseous mixture required in the study. The cell

is applicable to generate many different mixtures continuously and reproducibly over a wide range of concentration. The

main feature of the diffusion tube was the plasticine sealant. Mass spectroscopy showed that the volatiles of interest in this study were not liberated by the sealants and that the sealing was satisfactory.

In the course of the study, attempts were made to

generate dry formaldehyde vapours by pyrolysis of formaldehyde polymers and using selective sorbents to remove other products

of pyrolysis. Results on these attempts were discussed.

Moreover during the study, efforts had been directed to remove

xylene and alcohols selectively by porous polymers in formaldehyde determination. Both species interfered with the formaldehyde sensitive

detector and chromotropic acid colorimetry to certain extents.

- v -

Tenax GC was found to be the best sorbent for this purpose. Although in the chromotropic acid procedure, less intereference was observed using water as scrubbing medium than using

0.1% w/v chromotropic acid in concentrated sUlphuric acid,

the many advantages of using the latter scrubbing procedure

as noted in the study would have to be sacrificed. By using the pretrap, chromotropic-sulphuric acid could be retained

as the scrubbing agent. The pretrap is recommended for atmospheric formaldehyde monitoring especially in clinical laboratories where vapours of xylenes and alcohols can build

up to high levels.